Anda di halaman 1dari 100

Comprehensive Summaries of Uppsala Dissertations

from the Faculty of Medicine 943

_____________________________ _____________________________

Controlling Iodine
Deficiency Disorders
Studies for Program Management in Sub-Saharan Africa




Dissertation for the Degree of Doctor of Medical Science in Pediatrics presented at
Uppsala University in 2000.


Peterson, S. 2000. Controlling Iodine Deficiency Disorders - Studies for

Program Management in Sub-Saharan Africa. Acta Universitatis Upsaliensis,
Comprehensive Summaries of Uppsala Dissertations from the Faculty of
Medicine 943. 100 pp. Uppsala. ISBN. 91-554-4774-0.

Studies were performed to improve iodine deficiency control programs. Goitre

rates and cassava processing practices were compared in three Central African
Republic (CAR) populations. Short-cuts in cassava processing were associated
with elevated urinary thiocyanate and increased goitre rates, suggesting a
goitrogenic effect in one population. While improved cassava processing may be
beneficial, the priority is to correct the iodine deficiency.
The use of the urinary iodine/tiocyanate ratio as indicator of goitrogenic
effects was explored using data from Tanzania and CAR. As the ratio can be
calculated in four mathematically different ways and has physiological
shortcomings, its use is discouraged.
Biannual iodised oil capsule (IOC) distribution in a Tanzanian population
of 7 million during nine years was studied. Mean distribution coverage was 64%,
mean delay of subsequent distribution 1.25 years, and only 43% of targeted
person-time was covered. The cost of capsules constituted more than 90% of
total program costs. It is cost-effective to invest more funds in communication,
support of peripheral staff and supervision.
In a highland Tanzanian village, salt iodine content was highly variable
compared to national standards. While school-children had adequate urinary
iodine, women at delivery and newborns showed signs of inadequate iodine
status. Salt iodine concentrations should be monitored during production and
distribution down to household level, and iodine status assessed in all vulnerable
groups before adjusting recommended salt iodization levels at production.
WHO’s 1994 change in palpation goitre definition considerably lowered
specificity and increased measured goitre rates by 25% in Tanzanian school-
children compared to the previous system. Ultrasound estimation of thyroid
volume under rugged field conditions requires considerable human and material
resources yet had a precision only slightly better than palpation. In resource
poor settings appropriately trained palpators using the 1960 WHO definition of
goitre remain optimal for estimating thyroid size until precision and cost of
ultrasound has improved.
Monitoring of process indicators needs to be an ongoing priority activity,
separate from periodic evaluations of impact.

Key words: Iodine deficiency, iodized oil, cassava, thiocyanate, iodized salt, cost-
effectiveness, epidemiology, monitoring.

Stefan Peterson, Dept of Women’s and Children’s Health, Section for International
Maternal and Child Health, Uppsala University, SE-751 85 Uppsala, Sweden, and
the Dept of Public Health Sciences, Karolinska Institutet, Stockholm, Sweden

© Stefan Peterson 2000

ISSN 0282-7476
ISBN 91-554-4774-0
Printed in Sweden by University Printers, Uppsala 2000
“Iodine deficiency is so easy to prevent
that it is a crime to let a single child be born mentally
handicapped for that reason.”

Labouisse, UNICEF

“We were striving for the best, but it turned out as usual.”

Victor Chernomyrdin, Former Prime Minister of Russia

This study is dedicated to the millions of people still affected by

IDD and the people working to reduce these disorders
Stefan Peterson


This thesis is based on the following papers, which will be referred to by their
Roman numerals:

I. Peterson S, Légué F, Tylleskär T, Kpizingui E, Rosling H. Improved cassava-

processing can help reduce iodine deficiency disorders in the Central African
Republic. Nutrition Research 1995;15(6):803-812.

II. Peterson S, Rosling H, Tylleskär T, Gebre-Medhin M, Taube A. Endemic

Goitre in Guinea [Limitations of urinary iodine / thiocyanate ratios as
indicator of antithyroid effect in population studies]. Lancet 1995; 345
(February 25):513-514.

III. Peterson S, Assey V, Forsberg B, Greiner T, Kavishe F, Mduma B, Rosling H,

Sanga A, Gebre-Medhin M. Coverage and cost of iodised oil capsule
distribution in Tanzania. Health Policy and Planning, 1999; 14: 390-399,

IV. Sundqvist J, Wijetunga M, Assey V, Gebre-Medhin M, Peterson S. Salt

iodation and risk of neonatal brain damage. Lancet 1998; 352 (July 4):34-35.

V. Peterson S, Sanga A, Eklöf H, Bunga B, Taube A, Gebre-Medhin M,

Rosling H. Thyroid Size Classification by Palpation and Ultrasound in
Field Surveys. Lancet 2000; 355 (January 8): 106-110. Addendum in
Lancet 2000; 355, (June 3):1995-1997

Offprints were made with permission from the journals.

Controlling IDD – Studies for Program Management


IODINE PHYSIOLOGY ................................................................................................................7
IODINE DEFICIENCY DISORDERS ...............................................................................................9
Goitrogens ......................................................................................................................... 11
IODINE STATUS INDICATORS ................................................................................................... 15
Thyroid size ....................................................................................................................... 15
Urinary iodine ................................................................................................................... 18
Hormones .......................................................................................................................... 20
IODINE SUPPLEMENTATION ..................................................................................................... 22
Iodized oil .......................................................................................................................... 23
Iodized salt ........................................................................................................................ 26
Iodization of water, sugar, and other vehicles ..................................................................... 31
Excess iodine intake ........................................................................................................... 33
CONTROLLING IODINE DEFICIENCY DISORDERS ....................................................................... 34
Management structure and strategy ................................................................................... 34
Assessing IDD .................................................................................................................... 36
Communication.................................................................................................................. 37
Plan of action & Achievement of political will...................................................................... 38
Implementation.................................................................................................................. 38
Monitoring and Evaluation................................................................................................. 38
IDD IN SUB-SAHARAN AFRICA................................................................................................. 39
Central African Republic..................................................................................................... 39
Mainland Tanzania ............................................................................................................ 40
AIM OF THE STUDIES........................................................................................................... 44

MATERIALS, METHODS, RESULTS, AND COMMENTS .................................................... 45

DISTRIBUTING IODINE (PAPER III)............................................................................................ 51
EVALUATING PROGRAM IMPACT (PAPER IV,V) .......................................................................... 57
GENERAL DISCUSSION ....................................................................................................... 65
ELIMINATION, NOT ERADICATION ............................................................................................ 65
MONITORING DOES NOT EQUAL EVALUATION .......................................................................... 68
Defining the implementation process................................................................................... 69
Monitoring......................................................................................................................... 71
Evaluation ......................................................................................................................... 71
Modifying the “Social Wheel” .............................................................................................. 73
REASSESSING THE SITUATION ................................................................................................. 76
The strategy....................................................................................................................... 76
The impact on IDD............................................................................................................. 77
The management structure................................................................................................. 78
CONCLUSIONS...................................................................................................................... 81

ACKNOWLEDGEMENTS ....................................................................................................... 84

REFERENCES ........................................................................................................................ 86
Stefan Peterson

CAR Central African Republic
ICCIDD International Council for the Control of Iodine
Deficiency Disorders
ID Iodine deficiency
IDD Iodine deficiency disorders
IIH Iodine induced hyperthyroidism
I/SCN Iodine / Thiocyanate ratio
IOC Iodized oil capsule
I-salt Iodized salt
MIS Management information system
NCCIDD National Council for the Control of Iodine Deficiency
NFCC National Food Control Commission
PHC Primary Health Care
PPM Parts per million, corresponds to a concentration of 1
mg per kg
SCN Thiocyanate
T3 Triiodothyronine
T4 Thyroxine
TBG Thyroxine-binding globulin
TFNC Tanzania Food and Nutrition Centre
Tg Thyroglobulin
TRH Thyrotropin-releasing hormone
TSH Thyroid stimulating hormone
UNICEF United Nations Children’s Fund
USD United States dollar
WHO World Health Organization

Iodine terminology:
Iodize To add iodine, iodide or iodate to e.g. salt
Iodine The elemental atom form of iodine
Iodide Ion form of iodine, added as potassium iodide (KI)
Iodate Oxidized ion form of iodine, added as potassium
iodate (KIO3)

Controlling IDD – Studies for Program Management

Iodine Deficiency Disorders (IDD) are a range of health problems
and goiter is the visible "marker condition" for more severe
disorders. IDD control program activity has increased since the
World Health Assembly’s endorsement of the objective of
eliminating IDD by the year 2000 (WHO 1992) and the World
Development Report’s conclusion that iodine supplementation is
one of the interventions "...that would have the highest cost-
effectiveness of any health intervention available in the world
today" (World Bank 1993). Iodine supplementation programs
have started in 90 out of 94 IDD affected developing countries
(UNICEF 1995). Several interventions with proven high efficacy
are being used. Iodized salt is the main method to combat IDD
for long-term control and iodized oil capsules (IOC) a
complementary method in areas where iodization of salt for
different reasons cannot be implemented or fails to succeed.
Intensive implementation activities have revealed the
limited knowledge of how to translate efficacious interventions
into effective programs with relevant mechanisms for program
monitoring, quality assurance, and impact evaluation. This
thesis reports studies of the iodine supplementation program in
Tanzania −one of the most developed control programs in Africa
− and the early stage of the program in the Central African
Republic with goitrogens as a policy issue. Collaborating
institutions are the Tanzania Food and Nutrition Center (TFNC)
and the Ministère de Santé Publique, Service de Lutte contre les
Grandes Éndemies, Central African Republic.

Iodine Physiology
Iodine is essential for normal growth, development, and
functioning of the body. Since only minute amounts are required
each day, it is known as a micronutrient. Iodine is required for
synthesis of the thyroid hormones, thyroxine (T4) and
triiodothyronine (T3), which are necessary for the regulation of
body metabolism (Hetzel 1989).

Stefan Peterson

The recommended daily iodine intake is 50 µg for infants

(<12 months), 90 µg for children up to the age of 6 years, 120 µg
for children 7-12 years old, 150 µg for adolescents and adults, and
200 µg for pregnant and lactating women (WHO 1996). The adult
human body contains 20-50 mg of iodine, of which 70-80% are
stored in the thyroid gland, a remarkable concentration (Hetzel
In stable iodine intake situations, the amount of iodine
excreted in the urine correlates well with the iodine intake and
serves as an estimate of iodine intake. Less than 10% of human
iodine losses are excreted in the faeces (Choufoer et al. 1963),
sweat (Mao et al. 1990), and milk (Vermiglio et al. 1992).
The stomach and the upper small intestine rapidly absorb
iodine as either one of two chemical forms: iodide or iodate.
Iodate is reduced to iodide, which is transported in the blood to
the thyroid gland, where an active transport mechanism pumps it
into the thyroid cell. About 60 µg of iodine needs to be trapped
per day to maintain an adequate thyroxin level; the efficiency of
the trapping mechanism is regulated with the help of TSH
depending on the availability of iodine and the gland’s activity. In
the gland, iodide is oxidized to iodine, which is bound to tyrosine
to form mono- and diiodotyrosine. These are coupled to form
trioodothyronine (T3) and thyroxin (T4) in the thyroid epithelial
cells. Thyroxin is then stored in colloid follicles bound to
thyroglobulin. When needed, the thyroid stimulating hormone
(TSH) will stimulate the proteolysis of thyroglobulin in the
thyroid cells to release thyroxin into the blood stream (Hetzel
1989; Passmore & Eastwood 1986).
A feedback system regulates iodine metabolism by using the
hypothalamic hormone and the thyrotropin-releasing hormone
(TRH), which modulates the secretion of TSH from the
hypothalamus. The level of thyroxine (T4) in the blood regulates
metabolism. Decreasing T4 levels lead to increasing TSH levels,
which serve to increase thyroid iodine uptake as well as the
production and release of more T4 and triiodothyronine (T3) into
the bloodstream. At high TSH levels the thyroid will
preferentially produce the biologically more active T3.
Conversely, as thyroid hormone levels rise, TSH secretion falls.
Sustained high TSH levels stimulate an increase in the size and
number of follicular cells, an increase in vascularization, and

Controlling IDD – Studies for Program Management

consequently thyroid hypertrophy which leads to better iodine

capture. At some stage of hypertrophy the thyroid is regarded a
“goitre”. In addition, persistent stimulation may also cause the
formation of thyroid nodules (Hetzel 1989).
Thyroid hormones regulate the metabolism of target organs
by entering the cells of the peripheral tissues and bindngi to the
nuclear chromatin via a thyroid hormone receptor protein, which
in turn affects transcription (Baniahmad et al. 1992). T3 is much
more potent than T4. Thyroxin is therefore generally transformed
to T3 prior to biological action with the help of the selenium
containing enzyme 5’deiodinase (Kohrle 1999). This could
explain links between iodine and selenium deficiency
(Anonymous 1993; Thilly et al. 1991; Untoro et al. 1999). A recent
paper suggests that concurrent iron deficiency anaemia impairs
the therapeutic response to iodine supplementation, possibly
mediated via decreased T4 to T3 conversion or through decreased
thyroxiperidase activity impairing iodide organification
(Zimmerman et al. 2000).

Iodine Deficiency Disorders

Iodine was originally present in all soils. However, rainfall,
glaciation, exposure to wind and floods has leached the soils of
iodine, especially in mountainous areas and in flood plains.
Although some iodine is returned to the soil by rain, this is
insufficient and soils remain iodine deficient (Hetzel & Pandav
1996). Foods and animals grown on such soils will be iodine-
deficient. The only rich source of iodine is seafood. If seafood is
eaten one to two times per week it provides sufficient iodine
intake (Passmore & Eastwood 1986). Iodine-containing animal
feeds and antiseptics commonly used in the dairy industry result
in iodine rich milk which is an important source of iodine in
many countries (Phillips 1997). However, food alone is generally
not sufficient to give an adequate iodine intake.
The body’s first reaction to low iodine intake is to increase
the uptake efficiency up to four times the normal rate in order to
maintain the output of thyroxine. This is evidenced by a
reduction in urinary iodine concentrations. If this adaptation is

Stefan Peterson

insufficient, then serum TSH will rise and T4 will fall. The
preferential production of the more potent T3 over T4 increases
the efficiency of the use of the available iodine. Only when these
compensatory mechanisms are exhausted will hypothyroidism
occur as thyroxine levels fall below acceptable levels (Hetzel
The term iodine deficiency disorders (IDD) covers a wide
spectrum of clinical conditions (Hetzel 1983). Some disorders are
the result of compensation mechanisms, such as thyroid
enlargement, goiter. Early abortions, stillbirths, increased
perinatal and infant mortality, and mental and growth
retardation are some of the other manifestations of iodine
deficiency (Lamberg 1991). Some disorders are associated with
permanent brain damage that occurs during the first two
trimesters of pregnancy (Cao et al. 1994a; Hetzel & Mano 1989;
Pharoah & Connolly 1987). The resulting clinical condition,
cretinism, is irreversible and it is the most severe and dramatic
manifestation of IDD. However, IDD also include a range of more
subtle brain damages (Stanbury 1994). Iodine supplementation
has been shown to increase iodine deficient children’s
intelligence quotient considerably, indicating the suppressive
effect of continued iodine deficiency (Bleichrodt & Born 1994).
Global estimates of the number of subjects affected by some of
the IDD conditions are given in Table 1, (UNICEF 1994b; WHO

Table 1. Global burden of Iodine Deficiency Disorders.

Total population at risk 2.200 million

Subjects with goiter 740 million
Subjects with lowered mental ability 300 million
Cretinism 5.7 million

The effects of iodine deficiency on the foetus and neonate

also include increased perinatal and infant mortality (Cobra et al.
1997; DeLong et al. 1997; Hetzel 1989; Pharoah et al. 1976). A
sign of iodine deficiency in this period is neonatal
hypothyroidism. The incidence rate in iodine deficient areas may

Controlling IDD – Studies for Program Management

reach up to 1/10, compared to 1/4000 in non-iodine deficient

areas (Ermans et al. 1980a; Kochupillai 1992; Kochupillai &
Pandav 1987). Even transient hypothyroidism in the neonatal
period has been shown to lead to permanent detrimental effects
on a child’s intelligence (Calaciura et al. 1995).
During a child’s school-age years, iodine deficiency is an
impediment to proper brain functioning, estimated to be 13.5
intelligence quotient points (Bleichrodt & Born 1994). However,
the most obvious sign of IDD in this period is thyroid
enlargement, goiter. Goiter prevalence rates of 85% and higher
have been found in severely iodine deficient areas (Clugston et al.
1987; Eltom et al. 1984). As a sign of hypothyroidism, TSH levels
are elevated in large portions of the school aged group (Moreno-
Reyes et al. 1993). In adults widespread goiter, hypothyroidism,
and reproductive failures contribute to significant human
suffering and economic underdevelopment (Hetzel 1989; Hetzel &
Pandav 1996).
All the effects of iodine deficiency, encompassed by the
concept of IDD, are preventable by increasing the intake of iodine
(Burgi et al. 1990). This not only reduces human suffering, but it
also brings considerable benefits in terms of improved human
and economic development (Li & Wang 1987) and reductions in
medical care costs (Gutekunst 1993).

While Iodine Deficiency Disorders (IDD) are primarily caused by
insufficient dietary intake of iodine, other substances referred to
as goitrogens have been suggested to interfere with the proper
functioning of thyroid hormone synthesis and utilization. Many
substances have been proposed as potential goitrogens of public
health importance but it has been difficult to confirm or reject
these hypotheses or to estimate the importance of a goitrogenic
effect in any one population.
The pseudo-halide ion thiocyanate (SCNΓ) interferes with
the thyroid gland’s uptake and metabolism of iodine through
competitive inhibition (Gaitan 1990; Thilly 1992). SCN is formed
in the body from glycosides in cabbage, or when cyanide from
tobacco smoking or from cyanogenic substances in insufficiently
processed cassava is detoxified to SCN during metabolism

Stefan Peterson

(Bourdoux et al. 1978; Chanoine et al. 1991; Hegedüs et al. 1985).

Cassava, a staple for 500 million people, contains naturally
occurring cyanogenic glycosides. To avoid cyanide exposure from
consumption of roots from bitter cassava varieties, effective
processing methods were developed thousands of years ago
(Rosling 1988). In Africa, the main steps in the most common
processing methods are soaking the root in water (O'Brien et al.
1992), fermenting the grated pulp of the root in sacks, or
prolonged drying of the root in the sun (Hahn 1989; Mlingi et al.
1992). Cell disintegration during processing brings an
endogenous enzyme in contact with cyanogenic glycosides that
break down into cyanohydrins which rapidly yield volatile
hydrogen cyanide at pH levels above 6 (Cooke & Maduagwu
1978). However, insufficiently processed bitter roots may contain
considerable amounts of cyanogenic glycosides and cyanohydrins
that after ingestion may break down and result in significant
cyanide exposure (Casadei et al. 1990; Mlingi et al. 1992; O'Brien
et al. 1992; Tylleskär 1994; Tylleskär et al. 1992). In the human
body most of the cyanide will be enzymatically converted to
thiocyanate, which at physiological levels is slowly excreted in
the urine. The implication of thiocyanate in IDD has led to calls
to reduce cassava consumption (Biassoni et al. 1991; Biassoni et
al. 1990a; Biassoni et al. 1990b; Biassoni et al. 1998). Yet, other
observations suggest that with adequate iodine intake overt
hypothyroidism or goiter will not develop in the presence of high
thiocyanate loads (Cliff et al. 1986b; Delange et al. 1994).
The urinary I/SCN ratio (calculated as µg/mg) of a population
sample has been recommended as an indicator of the combined
effect on the thyroid of low iodine intake and high thiocyanate
exposure. Endemic goiter has been found to occur in cassava-
eating populations in Zaire when the mean ratio is less than 3,
and cretinism has been found to occur when the mean ratio is
less than 2 (Hennart et al. 1982). However, this ratio has not
been validated with stronger study designs, nor have the
programmatic implications for IDD control programs of elevated
thiocyanate loads been drawn beyond the drastic proposal to
replace cassava as a staple crop.
Another foodstuff for which goitrogenic effects have been
postulated is millet (Anonymous 1983; Elnour et al. 1998;
Moreno-Reyes et al. 1993). The goitrogenic effects of millet

Controlling IDD – Studies for Program Management

cannot be overcome by increased intake of iodine (Gaitan & Dunn

1992). High mineral content and bacterial contamination of
water may also be goitrogenic. Prolonged boiling inactivates this
effect (El Mahdi et al. 1986; Gaitan 1990). An association has also
been found between malnutrition and development of goitre
which may indicate an enhancement of the effect of goitrogens
(Gaitan 1990).

Analysing urinary thiocyanate

To assess the possible goitrogenic effects of thiocyanate (SCN)
casual urine samples collected for iodine concentration analysis
are typically analyzed also for their thiocyanate concentrations.
For this procedure a conservant (10% thymol in isopropanol) is
added and refrigeration is recommended. Larger urine volumes
are also necessary, as another 0.5 ml are needed for the analysis
(Lundquist et al. 1979). Thiocyanate values in Swedish non-
cassava eating non-smokers is around 40 µmol/l (0.23 mg/dl)
(Lundquist et al. 1979) and SCN values of 52 µmol/l (0.3 mg/dl)
were found in a Brussels population (Lagasse et al. 1982). Unit
conversions scales with indicators of levels of abnormality are
given in Figure 1.

Stefan Peterson

20 1000

1,4 18 900

16 800
Adequate Very
14 700
4 high
I µmol/l

I µg/dl

12 600

SCN µmol/l

SCN mg/dl
,8 3
10 500

8 Mild 400
deficiency 2
6 300


4 Moderate 200
2 100
deficiency Normal
0 0 0 0

Figure 1. Unit conversion scales for Iodine (I) and Thiocyanate (SCN)
concentrations in urine with indications of different levels of abnormality.

Controlling IDD – Studies for Program Management

Iodine Status Indicators

Thyroid size
The size of the thyroid gland in a school-aged child reflects the
severity of iodine deficiency the child has experienced in the
previous years. The goiter rate in school-aged children has
therefore been the most frequently used indicator to assess the
degree of iodine-deficiency in a population (Delange et al. 1997;
WHO 1994a). School-aged children also reflect a new cohort’s
exposure to iodine deficiency in the last few years. Schoolchildren
aged 6-12 are often used to approximate the goiter rate in their
entire age group. While a school-based survey excludes children
suffering from many of the severe complications of IDD such as
cretinism and deafness, the goiter rate is deemed representative
as long as school attendance in the community is more than 50%
(WHO 1994a).
When using the goitre rate to decide whether a population is
to be considered iodine deficient or not the decision depends on:
(1) the cut-off size at which the thyroid is considered to be a
(2) the precision of thyroid size estimates around that cut-off, and
(3) the prevalence of goiter in school-aged children at which
iodine deficiency disorders (IDD) are considered a public
health problem.
Over the years several definitions of goiter have been used and
the acceptable prevalence has changed. In 1960, WHO defined a
goiter as “a thyroid gland whose lateral lobes have a volume
greater than the terminal phalanxes of the thumbs of the person
examined” (Delange et al. 1986; DeMaeyer et al. 1979; Perez et al.
1960). This allowed for a thyroid to be palpable, yet not to be
considered a goiter. It was not clear whether it was the combined
thyroid volume of both thyroid lobes that had to be larger than
the combined volume of both thumbs, or whether it was sufficient
that one thyroid lobe was larger than that side’s thumb. Iodine
deficiency was defined as a public health problem when more
than 10% of school children had goiter (Delange et al. 1986;
Hetzel 1988; Perez et al. 1960; Querido et al. 1974).

Stefan Peterson

In 1994 WHO simplified goiter grading by reducing the

number of goiter grades from four to two. However, the definition
of goiter was simultaneously changed to “an enlarged thyroid
that is palpable but not visible”. Although it may not have been a
conscious decision the 1994 definition effectively caused all
palpable thyroids to be regarded as goiters, see Table 2. In
addition, the goiter rate at which IDD is considered a public
health problem was lowered from 10% to 5% (WHO 1994a).

Table 2. Simplified classification of goiter according to WHO 1994a.

Grade 0: No palpable or visible goitre.

A mass in the neck that is consistent with an enlarged

Grade 1: thyroid that is palpable but not visible when the neck is
in the normal position. It moves upward in the neck as
the subject swallows. Nodular alteration(s) can occur
even when the thyroid is not visibly enlarged.
A swelling in the neck that is visible when the neck is in a
Grade 2: normal position and is consistent with an enlarged
thyroid when the neck is palpated

The two WHO classification systems and the effect of the

revised definition of goiter are illustrated in Table 3. It is obvious
that the new goiter criterion entails a lowering of the threshold at
which an individual thyroid is considered a goiter. Furthermore,
at the population level the lowering of the acceptable goiter rate
to 5% will label new populations as “iodine deficient.” No
comparative study on the effect of these changes nor on the
precision obtained in goiter surveys using WHO’s 1960 and 1994
criteria for goiter diagnosis was found in the literature.

Controlling IDD – Studies for Program Management

Table 3. WHO thyroid size classification systems of 1960 with 1986

amendments (Delange et al. 1986; Perez et al. 1960) and the 1994 system
(WHO 1994a). Shaded areas indicate the thyroid sizes diagnosed as “goiter”
with each system.

Criteria for thyroid size estimation Goiter grades

WHO1960 WHO1994
Not palpable 0 0
Palpable lobes ≤ terminal phalanxes of 0 1
subject’s thumbs
Palpable lobes > terminal phalanxes of 1A 1
subject’s thumbs
Visible with neck extended 1B 1
Visible with head in normal position 2 2
Visible at a distance 3 2

The severity of iodine deficiency is judged by the Total Goiter

Rate, which is calculated as the number of children with goiters
divided by the total number of children examined. Table 4 gives
the benchmarks for grading iodine deficiency. Although goiter
rates can also be calculated for Visible Goiter (VGR: children with
grade 2 / No of children examined) or, in the old classification, for
goiter visible with the neck extended (VGR1B: 1B+2+3 / No of
children examined) (Mutamba 1993), these indices have found
less widespread use in IDD program management.

Table 4. Epidemiological criteria for assessing the severity of iodine deficiency

disorders (IDD) based on the prevalence of goiter in school-age children
according to WHO’s 1994 criteria (WHO 1994a).

Mild IDD Moderate Severe IDD

Total Goitre Rate 5.0 - 19.9% 20.0 – 29.9% ∃30.0%

Stefan Peterson

The intra-observer (MacLennan et al. 1969) and inter-

observer (Tonglet et al. 1994) variation of goiter classification by
palpation has been studied and found to be substantial. Some
interpretations of this has led to calls to discard palpation in
favor of the more costly ultrasound estimation of thyroid volume
(Gutekunst et al. 1986; Vitti et al. 1994; WHO 1994a; WHO &
ICCIDD 1997). Others have argued that the variations observed
reflect random misclassification and have little or no effect at the
population level, thus allowing continued reliance on palpation,
(Anonymous 1994). The justification given by WHO for the
change from palpation to ultrasound is a personal
communication by Dr R Gutekunst (WHO 1994a). However, the
presented supporting data do not allow determination of the
concordance between ultrasound and palpation, nor are they
based on the subsequently recommended thyroid volume cut-off
values (WHO & ICCIDD 1997).
Studies on the precision of ultrasound estimation of thyroid
volume are generally conducted on only a few subjects, and
report intra and inter-measurer errors of 2-13% from not clearly
deducible statistical procedures (Foo et al. 1999; Özgen et al.
1999; Vitti et al. 1994). The measurement errors reported from
ultrasound estimation of thyroid volume are considerably lower
than errors for determining the volume of other organs such as
the kidney, where the standard deviation of the difference
between repeated measurements divided by the mean volume
ranges from 14-17% (Emamian et al. 1995). This suggests that
different statistical techniques may have been used in the
calculation of the errors.

Urinary iodine
Most of the iodine absorbed is excreted in the urine making
urinary iodine content a good marker of current iodine status.
Although 24-hour urine collection accurately determines the
amount of iodine excreted, measuring urinary iodine
concentrations in casual morning urine samples yields an
adequate assessment of whether a population is iodine deficient
if at least 40-50 individuals are sampled (Dunn & van der Haar
1990). In a school where 300 children may be examined for
goiter, a systematic sub-sample can be sampled for urine; that is,

Controlling IDD – Studies for Program Management

every 6th child provides a urine sample. To more accurately

measure a population’s iodine status in an area a total sample of
200-300 samples is desirable. Collection of casual urine samples
is widely practiced because the sampling is acceptable to the
research subjects, the volume required for most analysis methods
is small (1 ml), the iodine content remains stable, and the
samples do not require refrigeration (Benmiloud et al. 1994;
Dunn & van der Haar 1990; WHO 1994a). An obvious weakness
of using casual urine samples is the error introduced from
variations in urinary volumes between individual subjects and
between different populations.
Using a ratio of urinary iodine and creatinine to compensate
for the lack of 24 hour urine collection has been found
unnecessary and potentially misleading, especially for
individuals who consume very little protein (Dunn et al. 1993;
Dunn & van der Haar 1990; Furnée et al. 1994; WHO 1994a).
Historically iodine concentrations are given as µg/liter
instead of µmol/liter (100 µg/l = 0.79 µmol/l), see Figure 1. The
analytic procedure that examines iodine concentrations was
established by Sandell and Koltoff in 1937 (Sandell & Kolthoff
1937). The main methods used are still based on the reduction of
ceric ammonium sulphate using arsenic containing acids and
timed readings in a colorimeter (Dunn et al. 1993; May et al.
1997). Simpler manual methods for urinary iodine analysis are
relatively inexpensive at USD 0.50 - 1.00 per sample (WHO
1994a). However, great care must be taken to avoid
contamination with iodine, which is found in household items
such as hair shampoo. The method requires special rooms,
glassware, reagents, and quality control procedures in reference
laboratories (Dunn et al. 1993). While there have been attempts
to develop analytical methods based on other reactions (Rendl et
al. 1998), these have so far not been validated in the iodine
deficiency concentration range. Because a rapid field test is not
yet available, transportation of samples to the laboratory and
waiting for the results is necessary.
The severity of iodine deficiency is determined using median
urinary iodine concentration values, see Table 5. Recently a
suggestion has been made to define the severity of IDD using
also the proportion of iodine concentrations below certain cut-off
values, and not just from the median value (Karmarkar &

Stefan Peterson

Pandav 1999). This would allow for the identification of

situations where a subgroup of the population remains iodine
deficient even though the median iodine concentration is
sufficient. “Elimination” of iodine deficiency is defined as
achieving a median urinary iodine concentration above 100 µg/l
with no more than 20% of the specimens registering below 50
µg/l. Assuming a design effect of 3 due to cluster sampling, a
sample size of 750 is required to achieve a confidence interval of
∀5% at a prevalence rate of 20% (WHO 1994a).

Table 5. Epidemiological criteria for assessing the severity of iodine deficiency

disorders (IDD) based on median urinary iodine levels (WHO 1994a).

Median value (µg/l) Severity of IDD

< 20 Severe
20 – 49 Moderate
50 – 99 Mild
∃ 100 No deficiency

Thyroid stimulating hormone
Increasing TSH levels is a compensation mechanism for iodine
deficiency. At population level the TSH levels thus indicate
iodine deficiency and hypothyroidism. However, use of TSH
levels as an indicator beyond the neonatal period has been
discouraged because of inconsistent relationships between
population TSH levels and iodine status (WHO 1994b).
THS levels in the foetus and the newborn are of special
concern since elevated TSH levels signal insufficient levels of the
thyroid hormone. This could cause abnormal brain development
(WHO 1994b). In gestational week 11 TSH is first detected in
fetal blood. TSH rises sharply at birth up to its maximum at 24
hours after birth and then it falls gradually back over the first 3-4
days as the infant’s T4 levels rise (Thorpe-Beeston et al. 1991).
Therefore, it is recommended to sample neonates either from cord
blood at birth or only after 72 hours, and then by heel prick
(WHO 1994b). TSH levels in neonates can be used to identify
congenital hypothyroidism in relatively iodine sufficient
environments. Screening all newborns for TSH will identify the

Controlling IDD – Studies for Program Management

approximately 1 per 4000 neonates (0.275%) that suffer from this

condition in populations with adequate iodine intake. This
procedure is usually done by blotting whole blood on filter paper,
drying it, and sending the sample to a reference laboratory.
Children with high TSH levels are then subjected to further
testing to confirm their deficiency and if necessary thyroid
hormone medicationis prescribed (Burrow 1980; Dunn & van der
Haar 1990). In countries with universal screening such as Italy
an incidence exceeding 1:2000 has been suggested as an
indication of mild iodine deficiency (Sorcini et al. 1995).
Measurement of the distribution of TSH levels in a sample of
newborns can give a picture of the iodine status in the population
(Kochupillai & Pandav 1987; Wächter et al. 1985; WHO 1994a).
Table 6 gives the benchmarks for judging the severity of the IDD
public health problem based on neonatal hypothyroidism.

Table 6. Epidemiological criteria for assessing the severity of iodine deficiency

disorders (IDD) based on prevalence of neonatal hypothyroidism (WHO 1994a).

Mild Moderate Severe

TSH > 5mU/l 3.0-19.9% 20.0-39.9% ∃40%
whole blood in

Thyroxine (T4) is the primary output of the thyroid gland.
Thyroxin is de-iodinated in the peripheral cells to T3 which is
thought to be the active hormone (Hetzel 1988). In severe iodine
deficiency the T4 levels may decrease, but de-iodination under
the influence of increased levels of TSH maintains adequate
levels of T3 (Kochupillai et al. 1973). The total T4 level depends
on the amount of carrier proteins in the blood. It is therefore
preferred to measure the levels of thyroxin that is unbound in the
serum, so called free T4. T4 can be used to characterize a finding
of an abnormal TSH value in a patient (Todd 1998). However, the
use of T4 is not considered as a suitable indicator for population
level assessments of iodine deficiency (Benmiloud et al. 1994;
WHO 1994a).

Stefan Peterson

Thyroglobulin (Tg) is the storage form of iodine and it is a thyroid
hormone precursor. Tg is a glycoprotein found in the follicular
colloid of the thyroid gland. Normally small amounts of
thyroglobulin leave the follicular cell. Tg is released into the
blood when there is an increased turnover of thyroid cells such as
in goiter (Todd 1998). Tg is more sensitive than TSH as a
measure of iodine status, because Tg changes earlier during
iodine deficiency and normalizes sooner during iodine
supplementation. Together with thyroid volume Tg has been
recommended as a marker of iodine status that is more sensitive
than TSH or T4 (Benmiloud et al. 1994; Eltom et al. 2000a; Eltom
et al. 2000b).

Iodine supplementation
Although iodine deficiency started appearing as a theory on
the cause of goiter as early as the 19th century, many other
causative theories persisted into the 20th century. In France,
Courtois isolated iodine in 1813 and in 1820 Coindet suggested
that iodine could treat goiter. Early attempts to use iodine in this
way often produced hyperthyroid effects; its use was therefore not
widely adopted. In 1825, Bossingault found that salt used in
Guaca, Colombia contained iodine. In 1833, he suggested that
iodized salt could prevent goiter. Controlled experiments in
France produced Iodine Induced Hyperthyroidism (IIH) which
again prevented widespread iodine supplementation (Hetzel
1989). In 1915, the use of sodium iodide tablets in Swiss school
children was successful; this led to the introduction of iodized salt
in Switzerland in 1922 (Hetzel & Pandav 1996). Marine and
Kimball carried out large scale controlled supplementation trials
in the United States from 1916 to 1920 that more conclusively
demonstrated both the curative and preventive aspects of iodine
supplementation. This led up to the community scale use of
iodized salt in Michigan in 1924 (Hetzel 1989; Marine & Kimball
1922). In Sweden a pioneering national survey of goiter and
cretinism preceded the adoption of salt iodization on a national
scale (Höjer 1931). Other pioneer efforts demonstrated that
cretinism is caused by iodine deficiency and that
supplementation with iodized oil can prevent new cases

Controlling IDD – Studies for Program Management

(McCullagh 1963; Pharoah et al. 1971; Pharoah & Connolly 1987;

Pharoah et al. 1976).
In several countries tremendous social and economic gains
have resulted from the control of IDD. In China iodine
supplementation led to increased educational achievement,
productivity, and quality of life (Li & Wang 1987). A meta-
analysis of iodine supplementation trials in areas of mild to
moderate deficiency concluded that iodine supplementation
raised the intelligence quotient by 13.5 points (Bleichrodt & Born
1994). A 100-year follow up in Switzerland of an initial survey
demonstrated the disappearance of cretinism and endemic goiter,
as well as reductions in isolated deafness, mental deficiency, and
short stature, making iodine supplementation a highly cost-
effective intervention for the country (Burgi et al. 1990).
Similarly, the World Bank concluded that iodine
supplementation is one of the interventions "that would have the
highest cost-effectiveness of any health intervention available in
the world today" (World Bank 1993). The World Health Assembly
endorsed the objective of eliminating IDD by the year 2000
(WHO 1992).
There is overwhelming evidence and resolve to prevent IDD
by increasing iodine intake in populations at large. Attempting to
reduce IDD through changes in diet is often not practical since
most foodstuffs, with the exception of seafood, are poor sources of
iodine (Passmore & Eastwood 1986). However, use of iodine-
containing animal feeds and disinfectants in the dairy industry
have made milk products an important source of iodine in some
places (Phillips 1997). The approach generally used to increase
the iodine intake countries is either through medicinal
supplements or through fortification of widely used foodstuffs
such as salt.

Iodized oil
Iodized oil was first used as a contrast medium in radiology and
later found, in Papua New Guinea, to be an effective control of
IDD when injected (McCullagh 1963). Subsequently, it was
shown that iodized oil can also be given orally (Eltom 1984).
Many countries have since used large amounts of iodized oil in
control programs. Outside of China the supply is dominated by

Stefan Peterson

Laboratoire Guerbet, Paris, France. Their Lipiodol Ultra-Fluid is

a x-ray contrast medium conventionally used in radiology; e.g., for
lymphography. It is a clear yellow poppyseed-oil based substance
containing 480 mg iodine/ml. It is available in 10 ml ampoules for
intra-muscular injection as well as in 200 mg capsules and as
Oriodol in oral pump-dispensers.
Iodized oil capsules cost about 14 US cents per 200 mg of
iodine and the pump-dispensed Oriodol about 4 cents per 200 mg
(Hetzel & Pandav 1996). This price is rather substantial
compared to e.g. Vitamin A capsules at 2-3 cents each.
Subsequently developed cheaper iodised oils have not found their
way into the market (Inglenbleek et al. 1997), nor has, to my
knowledge, any other producer bid against Guerbet in a
tendering process. The Chinese formulation is based on walnut
or soybean oil. This formulation has not been formally evaluated
for e.g. efficacy, safety and cost against other supplements such
as Lipiodol.

Intra-muscular administration
In the highland villages of Papua New Guinea, it was
demonstrated that a single injection of 2400 mg of iodized oil can
maintain normal thyroid function for 3 to 5 years (Buttfield &
Hetzel 1967; McCullagh 1963). The long lasting effect is
explained by the slow release of the large iodine load from the
muscle and its storage in fat tissue. Subsequently, the dose was
lowered to 1 ml (480 mg) giving an equally long protective period
(Dunn & van der Haar 1990) and large scale injection programs
have been implemented in countries such as Papua New Guinea,
Nepal, and Democratic Republic of Congo (Zaire). In addition,
these injections were reported to be safe for pregnant women
(Delange 1996; Pharoah & Connolly 1991). However, the high
cost, logistical difficulties, and risks of transmission of diseases
such as HIV and hepatitis have switched attention to oral
administration of iodized oil.

Oral administration
Taking iodised oil orally eliminates many of the disadvantages of
injections. There is no risk of transmitting infectious agents via
syringes, it requires less skilled staff, and it takes less time.
However, since there is no intra-muscular store of iodized oil,

Controlling IDD – Studies for Program Management

most is excreted within the first few days and the duration of the
effect is shorter. An initial study recommended 400 mg every two
years (Eltom et al. 1985). In 1987, a review recommended 1 ml
(480 mg) for one year’s duration (Dunn 1987b) and in 1990 this
was extended to 1-2 years (Dunn & van der Haar 1990).
Subsequent prophylactic studies have recommended 240 mg of
iodine for a 6-month coverage or 480 mg for 12 months
(Benmiloud et al. 1994); a study in Zaire found that 47 mg and
118 mg was sufficient for up to one year (Tonglet et al. 1992). A
study in the Sudan found that 200 mg was sufficient to control
iodine deficiency in goitrous adults for a 12 month period
(Elnagar et al. 1995). Presumably these discrepancies could be
explained by different underlying severity of the iodine
deficiency, different sampling or evaluation methodologies,
influences on iodine absorption/utilization by goitrogens,
helminths (Furnée 1994), or iron status (Zimmerman et al. 2000).
It seems clear, however, that the physiologically effective
duration of supplementation lasts closer to one year than to two
years. This conclusion is an important programmatic
The drawbacks of iodized oil supplementation include the
need to make direct contact with every targeted person at the
appropriate distribution interval. Intermittent bolus doses are
less physiologically effective than a constant dietary supply of
iodine, and that, although less than for injections, widespread
distribution poses significant logistical challenges. On the
positive side, oral iodized, oil can be accurately targeted to
population groups and individuals. A campaign can also be
initiated quickly in a severely iodine deficient area while other
control methods are being prepared. Oral iodized oil will thus be
targeted for areas with severe to moderate IDD where adequate
coverage with iodized salt is unlikely to be achieved in the near
future. Depending on the feasibility of introducing iodized salt in
an area, oral iodized oil may have to be considered as a medium
term control measure when iodine deficiency is severe enough to
warrant an intervention (Dunn 1996b).
The target groups for supplementation are, in order of
importance: 1) women of childbearing age; 2) children 0-5 years;
3) older children; 4) adult men 15-45 years of age (Dunn 1987b).
In people older than 45 years of age, iodized oil is not encouraged

Stefan Peterson

due to increased risks of hyperthyroidism. The same applies to

persons with nodular goiter; however, a reduced dose of iodized
oil could be considered. A recent review concluded that in areas
where iodine deficiency is moderate or severe, iodized oil should
also be given during pregnancy (Delange 1996).
The cost of providing iodized oil varies according to the
distribution dose, interval, and delivery strategy adopted. The
total annual cost of an oil supplementation program has been
estimated at USD 0.50 per person and per year (Jamison et al.
1993). However, in spite of the control efforts undertaken using
oral iodized oil in several countries, the literature contains no
account of the real distribution costs, nor of the actual
distribution coverage achieved, and timeliness of repeat

Iodized salt
The first large-scale iodine supplementation programs in the
United States and Switzerland used iodized salt. Subsequently
other countries in Europe started similar programs. Salt has
many advantages. It may be the only foodstuff imported by
subsistence agriculture communities. It is consumed in similar
amounts by all - even the socioeconomic level of a household has
very little influence on its consumption of salt. Furthermore,
iodization does not alter the taste or color. In most settings, this
makes salt an “ideal” vehicle to deliver a micronutrient like
iodine to all individuals in relatively similar daily doses, without
even having to contact them individually. An added advantage is
that livestock consume salt, benefiting themselves and the
consumers of animal products (Mannar 1996).

Iodizing the salt

The first countries to embark on salt fortification used potassium
iodide (KI). This has been the approach replicated by most
countries in temperate climates. However, in tropical humid
conditions the evaporation of iodide is substantial. In such
settings potassium iodate, KIO3, provides a less soluble, more
stable alternative. Iodate also performs better in salt with
impurities. Strictly speaking this is “iodated” salt but it is often
referred to as “iodized” as well (Dunn & van der Haar 1990;
Hetzel 1989).

Controlling IDD – Studies for Program Management

In many settings salt production is fairly centralized to a few

centers using techniques such as mining, evaporation of salt
water or underground brine. This makes it potentially feasible to
introduce and control the use of large-scale iodization technology
for dry mixing, drip feeding, or spraying the salt with iodine.
Thorough mixing of salt is necessary using drum mixers or screw
conveyors. The iodization machines most used in sub-Saharan
Africa are made in the Indian subcontinent and generally quite
simple and rugged. The cost of salt iodization is usually around
USD 0.02-0.10 per person per year. However, expressed as a
percentage of the retail price of salt, it amounts to about 5% of
the total cost of the product (range 2-32%) (Mannar 1996; World
Bank 1993).
Iodization levels in salt depend on the average daily salt
consumption, the severity of iodine deficiency, and expected
iodine evaporation losses during transport, storage, and cooking.
Humans usually consume between 5 to 20 grams of salt daily
and need a daily iodine intake of 100-200 micrograms. In a
warm, moist climate where salt is packaged in large sacks, the
recommended iodization levels for the factory were therefore set
at 50-100 mg of iodine per kilo of salt, often expressed as part per
million (ppm), see Table 7 (WHO 1994a).

Table 7. Salt iodization levels in parts per million recommended at different

stages of the distribution chain by WHO in 1994.

Daily Factory outside Factory inside Retail level House

salt the country the country hold
consum Bulk Small Bulk Small Bulk Small
ption pack pack pack
5 gr. 100 80 90 70 80 60 50
10 gr. 50 40 45 35 40 30 25

Iodine induced hyperthyroidism (IIH) was observed in

Zimbabwe (Todd et al. 1995) where salt was distributed in small
plastic consumer packs minimizing iodine losses. IIH was also
reported from Zaire (Bourdoux et al. 1996). This resulted in the
recommendation that iodization at the production level should be
reduced to 20-40 ppm (Clugston et al. 1996; WHO 1996).

Stefan Peterson

Legislation, quality control, and monitoring

Salt iodization raises the cost of large-scale salt production by
about 5%. This may create a situation where both iodized and
non-iodized salt can be sold at different prices, alternatively non-
iodized salt could falsely be sold as iodized with a larger profit
margin. This necessitates legislation requiring universal salt
iodization. Moreover, enforcement measures need to be in place to
prevent cost competition or fraudulent declarations concerning
iodization of salt. In addition it is recommended to increase the
popular demand for iodized salt (Dunn & van der Haar 1990).
Quality control of the salt iodization production process is
necessary to ensure adherence to the recommended iodization
levels. Although government quality control inspections are an
essential part of the quality control system, the main
responsibility for quality control is left to the producers. Sampling
procedures for such quality control have been suggested based on
the relatively accurate standard titration method of assessing salt
iodine content with intermittent tests using rapid test-kits
(Sullivan et al. 1995; WHO 1994a).
In addition, the salt iodine levels need to be monitored along
the distribution chain: at point of importation (where applicable),
in wholesale warehouses, in retail stores, and in households. One
objective of such monitoring is to identify, impound, and discard
non-iodized salt. Such activities require a titration assessment of
iodine content. Other objectives of iodine content monitoring is to
verify assumptions regarding iodine losses due to the effectiveness
of packaging materials, storage methods, and retail conditions and
develop ways to prevent such losses as well as to study household
salt storage. In addition, if rapid test kits are used, peripheral
monitoring can serve as a way to educate the pubic about salt
iodization. These kits use the starch-iodine color reaction to detect
the presence of either iodide or iodatein a relatively crude way.
Where more accurate determination of iodine content is needed,
e.g. to quantify losses or as a basis for taking legal action, titration
should be used (Dustin & Ecoffey 1978; Sullivan et al. 1995; WHO

Quality control and monitoring responsibilities need to be

clearly defined. Monitoring criteria for judging the adequacy of the
process have been proposed, Table 8. Sampling procedures have

Controlling IDD – Studies for Program Management

been suggested by Sullivan et al (1995).

Table 8. Criteria for assessing adequacy of salt iodization programs, their

quality control and monitoring (WHO 1994a)

Process Indicator Criterion of adequacy

A.Factory or importer level
1. Percent of food grade salt 100%
claimed to be iodized
2. Percent of food-grade salt >90%
effectively iodized
3. Adequacy of internal >90%
monitoring process*
4. Adequacy of external 10-12 monthly checks per
monitoring process producer/importer per year
B.Consumer and District Level
1. Percent of monitoring sites
with adequately iodized salt Adequate in 90% of samples
i) households (or schools)
ii) district headquarters
(including major markets)

2. Adequacy of monitoring 90% or more

* Corrective action systematically taken within 3 hours in 90% of cases using
Lot Quality Assurance Methodology
** Monitoring undertaken in 90% of districts in each province at both household
and district levels

While salt iodization appears to be the main candidate for iodine
fortification there are several potential challenges related to
getting the iodine onto salt, making iodine stay on salt, ensuring
the use of iodized salt, and knowing whether these things occur.
At the iodization stage, challenges include installing
iodization machinery, finding a reliable supply mechanism for
potassium iodate, and ensuring that production machinery is
used and maintained in an effective and consistent manner. This
requires a significant initial investment in equipment and
training as well as a commitment to continuos external
monitoring and quality control. A special difficulty arises in
situations where salt originates from many small-scale

Stefan Peterson

producers. Indonesia is one example and Tanzania with its

estimated 197 salt producers another (DailyNews 1994). It is
even more difficult to monitor quality in the salt produced by an
estimated 7,500 women who extract salt from salty soils using
evaporation processes in certain rural areas of Tanzania.
(ICCIDD 1997b; Momburi 1993). Although small-scale producers
only produce about 10% of the total world production (UNICEF
1994a), artisanal salt production is of large importance in
geographic areas of severe iodine deficiency. As any one producer
is too small to sustain an iodation machine organizing salt
iodation viably is difficult. UNICEF has suggested alternative
strategies to deal with this situation: organizing small producers
into cooperatives that share an iodization machine; sending a
mobile salt iodization machine periodically; providing each
producer with small scale iodization equipment; and establishing
commercial iodization refineries that purchase and iodise the salt
from the small producers (UNICEF 1994a). The relative cost of
iodizing salt in these ways is likely to be higher than in a large
scale operation since more transport and handling of the salt is
required. Monitoring salt iodization among small-scale producers
is also a logistic challenge.
In countries where the prevailing salt type is crude or coarse
salt that has not been crushed and homogenised the large, often
impure, salt crystals pose technical challenges to iodization.
Iodizing wet salt from salt pans adds further technical
The difficulties caused by iodine evaporation require
adequate wholesale packaging, storage, retail packaging and
display mechanisms. Large packaging used for bulk storage and
transportation should be lined with polyethylene and consumer
packaging should preferably be in sealed plastic bags (Mannar &
Dunn 1995). However, often any type of second hand bag is used
to pack salt. The salt may not be dry, which allows the iodine to
migrate to the bottom of 50 and 100 kg sacks causing iodine
levels to be unevenly distributed. Storage may expose the bags to
humidity or heat which accelerates iodine loss (Chauhan et al.
1992). At the retail level, it is customary in many places to pour
salt out in small heaps left uncovered in the sun causing losses of
iodine up to 42% (Kapil et al. 1996). Finally, household storage of
the iodized salt may be inadequate and practices such as

Controlling IDD – Studies for Program Management

washing the salt can further decrease the iodine content before
cooking, which results in as much as a 30% iodine loss (WHO
Public acceptance and demand for iodized salt is considered a
necessary prerequisite of a successful iodization program. This
can only be achieved through a massive education effort
involving decision makers, salt producers, salt retailers, health
workers, citizen groups, and the general public (Dunn 1996a;
Dunn & van der Haar 1990). In Germany and other European
countries, consumer groups often oppose iodine fortification
programs advocating the consumer’s right to choose between
iodized and non-iodized products. Similar arguments have been
raised in India. In other countries, lack of demand for iodized salt
may maintain a market for non-iodized salt, or make the
population prefer the cheaper non-iodized salt where both types
are available.
Several examples from South Asia and Latin America
demonstrate the need for effective monitoring and enforcement
mechanisms to achieve and sustain good coverage with
adequately iodized salt; (Murdoch et al. 1999; Hetzel 1989;
Passmore & Eastwood 1986). Often salt legislation is not
enforced. Monitoring and enforcement mechanisms need to be in
place at each stage of production and distribution. This will
involve several Government agencies and necessitates the setting
up of monitoring mechanisms, titration laboratories, and a steady
supply of rapid test kits (Dunn 1996a; Sullivan et al. 1995; WHO
The combined effect of these difficulties often leads to delays,
sometimes for decades, in achieving an effective salt iodization
program. Spain is such an example of the long delay in
implementing a salt iodization program (Fernandez 1990). Even
in the best case it will take a number of years to establish a
program. However, even after a successful program has been
established there may still be populations without effective
access or utilisation of iodised salt, e.g. because of non-iodating
small salt producers or local production of salt from salt lakes or
“foothill salt” (Momburi 1993). Therefore countries may need to
consider complementary supplementation methods where salt
iodization is not likely to succeed (Solomons 1998).

Stefan Peterson

Iodization of water, sugar, and other vehicles

Iodized water
Since water has to be consumed daily, it is another vehicle for
iodine supplementation. In 1923, the city of Rochester, in New
York state, was the first municipality to add iodine to its drinking
water. Since then, cities in Thailand, Malaysia, and Sicily have
added iodine to their central drinking water. These experiences
were largely successful in correcting iodine deficiency and also
served to disinfect the water supply provided chemical iodine is
used, not iodide or iodate. In Northern Thailand drops of iodine
solution are added to school water supplies with good results but
requiring constant supervision. However, in many settings
drinking water comes from many sources making effective
iodization difficult to achieve (Dunn 1987a; Dunn 1996b).
In semi-arid settings people often drink from one or from
very few water sources. A system using slow release iodine in
silicone matrices has been developed and used in Mali and in the
Central African Republic (Fisch et al. 1993; Pichard et al. 1992;
Yazipo et al. 1995a; Yazipo et al. 1995b). This method of
iodization in hand-pump wells increases water iodine, urinary
iodine, and reduces goiter rates. A study from the Sudan
confirmed these findings but demonstrated limitations in the
effectiveness of the system in traditional wells (Elnagar et al.
1997). The matrices cost around USD 100 each. They need to be
changed every year and the per capita cost will depend on how
many people draw their drinking water from the same well.
Small villages are more expensive to supplement per capita.
In a semi-arid area of Xinjiang, China, iodine absorbed by
the crops from iodized irrigation water found its way through the
food chain and into people. This method not only normalized
iodine deficiency but also reduced infant mortality at an
estimated cost of USD 0.05 per capita per year (Cao et al. 1994b;
DeLong et al. 1997; DeLong 1998).

Iodized Sugar and Bread

In the Sudan, sugar is a food that is consumed by a large number
of people and it is subsidized and rationed. A pilot trial to fortify
sugar with iodine was successful in increasing urinary iodine
concentrations, in reducing goiter rates, and in improving

Controlling IDD – Studies for Program Management

thyroid hormone status. Sugar, therefore, is another way to

introduce iodine into a population (Eltom et al. 1995).
Iodized bread has been used successfully in Holland, Russia,
Australia, and Tasmania but the variations in bread
consumption and the unavailability of bread in remote rural
areas limit the practicality of this method (Clements et al. 1970;
Gandz 1974; Gerasimov et al. 1995). Iodized tea has been
suggested as a means of food fortification in Tibet.

Direct supplementation
Direct supplementation can also be done using iodide tablets or
solutions. Some of the early supplementation research was done
using sodium iodide tablets (Gibson & Backman 1924 cited in
Sjöberg & Sundlöf 1971). This supplementation method was used
in the Soviet Union and sodium iodide tablets continue to be
used in Germany. Iodine solutions such as Lugol’s (6 mg iodine
per drop) have been used successfully in controlling iodine
deficiency in Bolivia. Direct supplementation, which requires
daily, weekly, or perhaps monthly doses, has limited overall
effectiveness since its effectiveness depends on individual and
systemic responsibility and motivation (Dunn 1996b; Todd &
Dunn 1998).

Excess iodine intake

Food, medication, and supplements are sources of iodine. In
general, consumption of iodine that exceeds the recommended
dosage by as much as ten times is well tolerated by most people.
However, some people respond adversely to levels close to the
recommended intake (Todd 1998). A dramatic increase in the
plasma iodide load causes the thyroid gland to block iodine
organification. This transient phenomenon is known as the
Wolff-Chaikoff effect (Wolff et al. 1949). Other manifestations of
excess iodine include thyroiditis, goiter, hypothyroidism, hyper-
thyroidism, and sensitivity reactions.
The phenomenon of most relevance for iodine
supplementation programs is iodine-induced-hyperthyroidism
(IIH). Symptoms are the usual of thyrotoxicosis: anxiety,
palpitations, weight loss, muscle weakness, fatigue, diarrhea,
sweating, and hot sensations. More serious cardiac effects such
as: atrial fibrillation, angina, and heart failure may also occur. In
most cases IIH is caused by toxic nodular goiter (Todd 1998).

Stefan Peterson

Following the introduction of iodine supplementation into a

previously iodine-deficient population susceptible individuals
may develop IIH. In most cases this occurs in women over the age
of 40 with nodular goitre (Todd 1998). The most well-known
example of this followed the introduction of iodized bread into
Tasmania in the 1960’s (Connolly 1971). Recently, in Zimbabwe
(Todd et al. 1995) and Zaire (Bourdoux et al. 1996) iodization of
salt resulted in cases of IIH. While the condition in any one
patient can be serious and requires immediate medical care, IIH
is a temporary and rare phenomenon that can be successfully
treated if detected. Typically, IIH occurs during the first few years
of an iodization program as a result of the many large goiters
caused by the previous iodine deficiency, thus making it an iodine
deficiency disorder. For a population the benefits of iodine
supplementation on balance far outweigh the risks of increased
hyperthyroidism (WHO 1997). While IIH cannot be fully avoided
it can be minimized by regulating the iodine supply (Todd 1998).
Because of this, recommended salt iodization levels were lowered
in 1996 to 20-40 ppm (WHO 1996).

Controlling Iodine Deficiency Disorders

Management structure and strategy
Recognizing that it was technically feasible as well as affordable
to control IDD, yet effective control programs were not in place, a
group of concerned professionals, mostly IDD researchers with a
strong representation of endocrinologists, took the initiative to
create the International Council for Control of Iodine Deficiency
Disorders (ICCIDD). ICCIDD was officially inaugurated in
Kathmandu, Nepal in 1986. It has an international executive
committee, six regional IDD coordinators and working groups as
well as a multidisciplinary network of members. Based on the
1986 World Health Assembly resolution to prevent and control
IDD, the ICCIDD has worked closely with multilateral, bilateral,
national, and non-governmental organizations to lay out and
implement a global strategy for IDD control. This strategy has
been endorsed by the UN System’s coordination subcommittee on
nutrition (Hetzel 1987a; Hetzel 1988; Hetzel 1996; WHO 1986).
Success in controlling IDD is a complex process involving

Controlling IDD – Studies for Program Management

political, social, and economic dimensions. This process has been

studied in different countries and a model summarizing required
program steps and objectives was proposed and updated to its
present form based on experiences in South East Asia (Hetzel
1987a). It was adopted immediately as the guiding model by the
International Council for Control of Iodine Deficiency Disorders
(ICCIDD), and it has remained fairly constant with only minor
changes being made. It comprises the following steps which are
to be completed in a cyclical fashion by a multi-sectoral National
IDD Control Commission (Hetzel 1988):
1. Assessment (collect data, assess situation)
2. Communication (disseminate findings)
3. Planning (develop or update plan of action)
4. Political decision (achieve political support)
5. Implementation
6. Monitoring and evaluation
The latest published version is presented in Figure 2.

Prevalence of IDD
Urinary iodine
Blood sport TSH Population at risk
Prevalence of IDD
Evaluate Situation

Allocation Health
Implementation Professionals
Program Findings
of Programme and Public

Achieve Develop or
Political Update
Will Plan of Action

Community Intersectoral
Groundswell Commission

Figure 2. The ”Social Wheel” of IDD control (Hetzel 1996)

Stefan Peterson

In the following sections selected technical aspects of the

IDD “Social Wheel” control cycle as recommended by ICCIDD,
WHO and UNICEF are reviewed (WHO 1994a).

Assessing IDD
Many indicators exist for individual iodine deficiency disorders.
The selection of indicators for assessment depends on
acceptability, technical feasibility, cost, and the indicator’s
measurement performance. It is generally recommended to
combine at least two indicators, such as thyroid size and urinary
iodine concentration. Commonly used indicators are presented in
Table 9 together with criteria for judging the severity of the
public health problem (WHO 1994a).

Table 9. Summary of IDD prevalence indicators and criteria for classification of

severity of public health problem (WHO 1994a).

Severity of public health problem

Indicator Target (prevalence)
Population Mild Moderate Severe

Goiter grade > 0 SACa 5.0-19.9% 20.0-29.9% >30.0%

> 30.0%
Thyroid volume >97th SAC 5.0-19.9% 20.0-29.9%
Centile by ultrasound b < 20

Median urinary iodine SAC 50-99 20-49

level (µg/L) > 40.0%

> 40
TSH >5mU/L whole Neonates 3.0-19.9% 20.0-39.9%

Median Tg C/Ad 10.0-19.9 20.0-39.9

(ng/ml serum)c
a SAC = school-age children
b Reference values in (WHO & ICCIDD 1997)
c Different assays may have different normal ranges
d C/A = children and adults

Controlling IDD – Studies for Program Management

The target groups for assessment also merit discussion.

While many of the effects of iodine deficiency manifest
themselves in neonates and during early childhood, this age-
group is not very accessible. Adult women, especially pregnant
women, need adequate iodine status to prevent IDD from
developing in their foetus. However, this group may not be very
appropriate for some indicators since goiters in women do not
necessarily reflect recent iodine status. The most commonly
studied group is children in schools. This highly accessible and
co-operative group is also vulnerable and their thyroid size will
reflect the iodine status in the last few years (Aghini-Lombardi et
al. 1997). Inference is then made from this group to the others,
such as neonates and pregnant women. However, in situations
where school attendance is low the children at school may
provide a biased estimate as their peers at home may be worse
off (WHO 1994a). The construction of sampling frames for school
selection must take into account student geographic origin and
stratify for presumed different degrees of iodine deficiency.

The effects of IDD should be communicated to the public and
politicians. Topics to communicate include the effects on growth
and development, the effects on the brain, the effects on school
performance, the effects on productivity, and the effects on
quality of life. Widespread understanding of the scope and cause
of IDD should result in support and demand for interventions.
Press coverage is assured during launching ceremonies of
legislation, iodation plants and major meetings (Daily News
1994; Kwena 2000). In addition, flyers and posters communicate
the impact of IDD and encourage the consumption of iodized salt.
“Question and answer” booklets have also been produced in some
countries. A few countries have made use of “IDD days” to raise
awareness. ICCIDD now stresses the need to improve
communication at all levels and has produced a communication
guide which will be translated into several languages (ICCIDD
1999a; Ling & Reader-Wilstein ).

Stefan Peterson

Plan of action & Achievement of political will

The ICCIDD regional groups serve to promote national programs

by coordinating events, providing technical advice, fundraising,
and progress monitoring. With their support ICCIDD
recommended managerial structures have been re-created in
many countries. Thus multisectoral National Control
Commissions for IDD implement the strategy and its components
in individual countries.
Using an intersectoral commission a plan of action is drawn.
This involves a situation analysis of the salt sector and salt
consumption as a basis for introducing salt iodization.
Recommended groupings such as a salt producers’ association are
then formed. In addition, salt producers’ co-operatives are formed
to organize small salt producers.
In situations where iodized salt is deemed to be long in
coming, other control methods, primarily iodized oil, can be used
as temporary control efforts while striving to achieve universal
salt iodization.
The international effort to raise awareness and achieve
political will has been very successful, as confirmed by the
enactment of salt iodization legislation in a growing number of
countries. Much operational and financial support for this process
has been rendered by UNICEF in light of the goal to eliminate
IDD by the year 2000 (Grant 1994). Great strides have been
made towards this target but the challenge of achieving full
elimination remains.

With UNICEF and other donor assistance, salt iodization
machinery has been imported and installed in salt producing
countries, potassium iodate has been procured, and personnel
has been trained. Where control methods other than salt have
been used, countries have generally been supported by outside

Monitoring and Evaluation

The initial strategy spelled out monitoring and evaluation as the
final step of the “social wheel” (Hetzel 1988). The subsequent

Controlling IDD – Studies for Program Management

publication of the model talks of evaluation alone and lists

indicators such as prevalence of IDD, urinary iodine, and
bloodspot TSH, see Figure 2 (Hetzel 1996). Subsequent technical
documents also suggest methods and indicators for process
monitoring of the intervention, i.e. salt iodization (Sullivan et al.
1995; WHO 1994a).

IDD in Sub-Saharan Africa

Iodine deficiency and IDD are widespread in Africa. A third of the
total population is considered at risk of IDD and 86 million, 16%
of the population, is thought to have goiters. Severe IDD affects
large parts of Central, Eastern, and Southern Africa
(WHO/UNICEF/ICCIDD 1993). Further details are given below
for two severely endemic countries, the Central African Republic
and mainland Tanzania.

Central African Republic

The Central African Republic is a landlocked country across the
Ubangi river from the Ubangi area in the Democratic Republic of
Congo (former Zaire). IDD in the Ubangi area was extensively
studied by Ermans et al who found severe iodine deficiency and a
high prevalence of IDD. Based on findings of increased urinary
thiocyanate levels they concluded that cassava was an
aggravating factor for IDD (Ermans et al. 1980b; Thilly 1992). In
the Central African Republic IDD was found to be highly
prevalent both in smaller area surveys (Biassoni et al. 1990b;
Mongonou 1982) and in a national survey (Lantum 1991). All
parts of the country are highly endemic with total goiter rates
between 40 and 90%. Approximately two thirds of the population
is affected by IDD (WHO/UNICEF/ICCIDD 1993).
Most of the salt in the country is imported. Salt from
Cameroon penetrates the western parts of the country. It is
increasingly iodised following the monopolistic producer
SELCAM’s successful iodation efforts (Lantum 1992). Other salt
sources are located in South Africa, Namibia, and the Sudan in
the East. Traditional salts, such as “natron” from Lake Chad and
the production of salt by burning grass into ashes, are also

Stefan Peterson

available (ICCIDD 1997a; Thilly 1992). Successful trials have

been conducted with iodized water in the more arid parts of the
country (Yazipo et al. 1995a; Yazipo et al. 1995b). The national
primary health care commission oversees IDD control efforts,
which includes monitoring salt imports and monitoring water
iodization (Lantum 1992).
Cassava is the main staple crop (Fresco 1986). The
thiocyanate load from cassava consumption has been implicated
as an aggravator of goiter in CAR (Biassoni et al. 1990a), and
reduced consumption suggested as a means to alleviate the
problem (Biassoni et al. 1991). Another condition attributed to
dietary cyanide-exposure from cassava, is the paralytic disease
konzo, which has also been reported in CAR (Peterson 1994;
Tylleskär et al. 1994).

Mainland Tanzania
Mainland Tanzania has a large landmass with coastline,
highlands, mountains, and floodplains. Studies in the
southwestern part of the Tanzanian highlands have
demonstrated high goiter rates, cretinism, and highly prevalent
hypothyroidism in schoolchildren as well as neonates, which
were amenable to correction using injections of iodized oil
(Wächter et al. 1986; Wächter et al. 1985). Nation-wide systematic
surveys of iodine deficiency disorders were gradually conducted
in numerous small-scale school goiter-surveys during 1980
through 1990. Total goiter rates between 1 and 88% were found
with an estimated 40% of the population affected
(WHO/UNICEF/ICCIDD 1993). The summary map of goiter rates
by district shows severe IDD in most highland and floodplain
regions in Tanzania (NCCIDD-TFNC 1990). Based on the goiter
rates established in the school surveys, van der Haar, Kavishe
and Gebre-Medhin estimate that 8 million people have goiter,
160,000 cretinism, 450,000 more mildly affected with
cretinoidism, and that 30% of total perinatal mortality is caused
by iodine deficiency (van der Haar et al. 1988).
The Tanzania Food and Nutrition Center (TFNC) developed
a strategy for National IDD control in 1985. The plan called for
the use of iodized oil as an interim control method in severely
affected areas to prevent further brain damage there, while

Controlling IDD – Studies for Program Management

working towards a universal salt iodization program. Injection of

iodized oil was initially tried in two districts but abandoned due
to cost and shortage of trained staff. Instead, iodized oil capsules
(IOC) were distributed beginning in 1986. Using an initial
criterion of district Visible Goiter Rate >10%, a total of 20
districts were targeted for IOC distribution. Later the criterion
was revised to Visible Goiter and grade 1B >10% and 5 more
districts were added. Supported by the Swedish International
Development Authority, the objective was to distribute two IOC
(400mg) every two years to the general public 2-45 years of age,
excluding persons with multi-nodular goiter, and one IOC
(200mg of iodine) to children aged 12-23 months of age. (ICCIDD
1997b; Sanga 1994).
The efforts to iodize salt in Tanzania began in 1973.
However, technical, administrative, and coordination problems
among the actors stifled the effort (Sanga 1994). The salt
iodization effort should have been re-launched with Dutch
support together with the start of the IOC distribution. However,
due to several sources of delays it was only in 1991 that UNICEF
imported three iodization machines from India and installed
them in three major salt works. These first machines were
followed by 42 more machines until 1995 (UNICEF 1996). Many
of these were smaller rotary drum type mixers provided to 200 or
so small-scale salt producers located mainly on the southern
coast. These producers were encouraged to form co-operatives
and channel all their salt through a centrally located iodization
machine. The installed iodization capacity at 150,000 tons per
year was theoretically sufficient to satisfy the national need
assuming constant salt production throughout the year.
Constraints on the program include seasonal salt production and
only about 140 salt producers being officially registered.
Government has been unwilling to centralize salt production to a
few larger producers, presumably for reasons of maintaining job
opportunities (Sanga 1994; UNICEF 1996).
In the central parts of the country production of “foothill salt”
from salty soils is widespread. Some 7,500 women are believed to
be involved. By certain estimates this production is quite
substantial amounting up to 40,000 metric tons, although it has
yet to be substantiated how widespread exclusive use of the salt
is (ICCIDD 1997b; Momburi 1993).

Stefan Peterson

A 1995 law requires all salt for human consumption to be

iodized whether produced in the country or imported (some
10,000 metric tons per year) (ICCIDD 1997b). Domestic salt
production takes place mainly on the coast and in the extreme
west of the country with complex distribution patterns requiring
an estimated 4-6 weeks to reach the consumer (TISCO 1992).
Consuming an average of 8 grams of salt per day, Tanzanian
consumers appear to have strong preferences because of the
“taste” and the price for coarse, crudely refined, large-crystal salt
distributed in 50 kg reused sacks of different material. The
coarse salt is often moist and does not hold iodine very well.
Furthermore, it is retailed loose in markets, freely exposed to the
sun, providing opportunities for rapid iodine loss. Refined table
salt in small 0.5 kg plastic bags is widely available, however, at
approximately double the price of the coarse salt. It is mainly
used to sprinkle on special foods such as roasted meat. Taking
these production, distribution, and retail patterns into account,
salt iodization levels were set at 75-100 ppm at the factory so as
to achieve household levels of 37 ppm or more (Sanga 1994).
A National Control Commission for IDD (NCCIDD) was
organized for all stakeholders. TFNC, under the Ministry of
Health, is the secretariat. An association for salt producers has
been formed to encourage dialogue, distribution of the UNICEF
supplied iodization machines, distribution of potassium iodate,
and initially providing appropriate packing materials. Support
was also provided for monitoring. The monitoring responsibility
is divided among the different actors. These are the Ministry of
Water, Energy and Minerals (MOWEM) for the production sites,
the National Food Control Commission (NFCC) with their district
health assistants for the distribution chain, and TFNC for the
household consumption level (ICCIDD 1997b; Sanga 1994).
Larger production sites were given a salt titration laboratory. To
facilitate production site monitoring, a number of vehicles were
supplied to the zonal mining officers in MOWEM. For more
accurate quality control and enforcement during the distribution
chain, NFCC set up network of regional salt titration
laboratories. Peripheral Health Assistants’ and TFNC’s
monitoring has largely used rapid field test kits. Altogether the
external support for the iodization program amounted to 80% of
the budget during the first few years (Sanga 1994).

Controlling IDD – Studies for Program Management

I. Peterson S, Légué F, Tylleskär T, Kpizingui E, Rosling H. Improved cassava-

processing can help reduce iodine deficiency disorders in the Central African
Republic. Nutrition Research 1995; 15(6):803-812.
II. Peterson S, Rosling H, Tylleskär T, Gebre-Medhin M, Taube A. Endemic Goitre in
Guinea [Limitations of urinary iodine / thiocyanate ratios as indicator of
antithyroid effect in population studies]. Lancet 1995; 345(February 25):513-514.
III. Peterson S, Assey V, Forsberg B, Greiner T, Kavishe FP, Mduma B, Rosling H,
Sanga B, Gebre-Medhin M. Coverage and cost of iodized oil capsule distribution in
Tanzania. Health Policy and Planning 1998; 14(4):390-399.
IV. Sundqvist J, Wijetunga M, Assey V, Gebre-Medhin M, Peterson S. Salt iodation and
risk of neonatal brain damage. Lancet 1998; 352(July 4):34-35.
V. Peterson S, Sanga A, Eklöf H, Bunga B, Taube A, Gebre-Medhin M, Rosling H.
Classification of Thyroid Size by Palpation and Ultrasonography in Field Surveys.
Lancet 2000; 355 (January 8):106-110. Addendum in Lancet 2000; 355, (June 3):1995-

Figure 3. Location and titles of studies in Africa

Stefan Peterson


The overall aim of the studies presented in this thesis is to
improve the methodologies to assess the severity of iodine
deficiency, to monitor the implementation of iodine
supplementation programs, and to evaluate program impact. The
specific objectives of the five studies are as follows:

- to elucidate whether the thiocyanate load from cassava

significantly aggravates the effects of iodine deficiency in three
cassava consuming communities of the Central African
Republic and draw conclusions for IDD control policy (paper I);

- to analyze the mathematical and physiological properties of

the iodine/thiocyanate ratio as an epidemiological indicator for
the postulated goitrogenic effects at urinary I/SCN ratios <3
µg/mg using data from Tanzania and the Central African
Republic (paper II);

- to estimate the population time supplemented by Iodized Oil

Capsule (IOC) distribution over a nine year period in a
Tanzanian population of 7 million and to estimate the cost and
effectiveness of the three distribution strategies used (paper

- to study the relationship among the three main iodine

deficiency indicators in three vulnerable groups and to relate
this to the iodine content in table salt in a Tanzanian
community (paper IV);

- to determine the effects on goiter rate of the WHO 1994

change in palpation criteria for goiter and to compare the
relative precision between the old and new palpation systems
and ultrasound determined goiter in a school survey in rural
Tanzania (paper V).

Controlling IDD – Studies for Program Management

This thesis consists of several studies, Figure 3. A range of
quantitative and qualitative research methods have been used,
Table 10. The sections below present the five individual studies
grouped into three areas related to their relevance in different
stages of the IDD control program cycle. Study I and II relate to
the assessment stage, study III to the implementation stage, and
study IV and V to the evaluation stage. For each of the three
areas material and methods are presented, results summarised
and briefly commented on.

Table 10. Summary of research methods used in this thesis.

Method Paper I Paper II Paper III Paper IV Paper V

Epidemiology √ √ √ √
& Biostatistics
Biochemical √ √ √
Qualitative √ √
Cost analysis √

Modeling of √
secondary data

Assessing Iodine Deficiency and Goitrogenic

Effects (paper I,II)
Having arrived in the Central African Republic to study health
effects of cassava (Tylleskär et al. 1994), I was struck by finding a
goiter in every other child I examined. Since the country is
extremely dependent on cassava the potential goitrogenic effects
could be considerable as an explanatory factor. The
epidemiological indicator urinary I/SCN ratio had been proposed
to assess the presence of the potential goitrogenic effects
(Delange et al. 1980). A considerable scientific debate has

Stefan Peterson

addressed the issue (Bourdoux et al. 1978; Bourdoux et al. 1980;

Cliff et al. 1986a; Delange & Ahluwalia 1983; Delange & Ermans
1971; Delange et al. 1982; Gaitan 1990). However, the question of
the goitrogenic effect’s practical importance for iodine deficiency
contrlp efforts at population level and the utility of the I/SCN
ratio for assessment of potential goitrogenic effects was unclear.

Materials and Methods

In study one, cassava processing practices were elucidated in two

rural and one urban setting using focus group discussions and
participant observation (Khan & Manderson 1992; Scrimshaw &
Gleason 1992). A 24 hour dietary recall was made in 20 randomly
selected households to establish the importance of cassava in the
diet (Cameron & Van Staveren 1988). Casual urine samples were
collected for analysis of iodine and thiocyanate in adults in the
interviewed households as well as among school children in the
local school in each community. Samples were immediately
frozen until they were analyzed in Sweden for iodine by a
modified Sandell−Kolthoff method (Sandell & Kolthoff 1937) and
for thiocyanate (Lundquist et al. 1979), respectively. All goiter
gradings were performed independently in triplicate according to
Perez et al. (1960). In case of differences, consensus on grade was
reached after discussion. The iodine/thiocyanate-ratio (I/SCN)
(Delange et al. 1980) was calculated as the mean, median, and
mode of the individual samples' ratios obtained through the
division of each iodine-concentration (Φg/dl) by the same
specimen's thiocyanate-concentration (mg/dl).
In study two, a literature review was carried out for use of
the I/SCN ratio. The Central African Republic study and
additional data from Tanzania were re-analyzed according to the
various practices identified in the literature.


All three Central African communities studied were iodine

deficient, as judged by urinary iodine concentrations. All three
study sites depended on bitter cassava as the main staple. Bitter
cassava was the dominant staple consumed in all households.
However, processing practices differed between the communities,

Controlling IDD – Studies for Program Management

with a recent introduction of a shortcut processing method in a

western village, see Figure 4. In semi-urban Bangui cassava was
bought already processed from the village.
1st night 2nd 3rd 4th 5th 6th

(Central village)

(Western village)

Legend Soaking Mashing Basket fermentation

Sun-drying Pounding Crushing with machete

Figure 4. Cassava processing practices in the two rural areas of CAR.

This change in processing practices resulted in tripled urinary

thiocyanate levels compared to the low levels where traditional
processing or traded cassava products were consumed. Associated
with this was a near doubling of the goiter rate indicating a
possible goitrogenic effect, see Table 11.
Table 11 Cassava processing, urinary iodine and thiocyanate and goiter rates in the
studied populations
Western Village Central Village Bangui

Cassava processing shortcut traditional traded


School Population School Population School

I µg / dl
mean ± SEM 6.98±0.73 10.80±2.21 6.17±0.81 5.76±0.45 6.30±4.21
median 5.78 4.57 4.19 4.45 1.65
n 48 84 47 103 61

SCN µmol / l
mean ± SEM 239±19 231±17 48±8 81±9 65 ±6
median 232 186 31 57 57
n 48 84 47 103 61

Crude TGR(%) 51 50 28 28 22
95% Conf.Int. 43-59 38-62 20-36 18-38 16-28
n 152 72 101 95 184

Stefan Peterson


20 Ce nt ral Sch oo l

mode 1 5 - 1 7. 9
10 median 2 2 .7
mean 3 1. 0

0 6 12 18 24 30 36 42 48 54 6 0 >60

Ur inary I/ SCN rat io ( µg/ mg)


20 We st e rn Sch oo l
mode 0 -2 .9
median 5. 6
mean 7 .4

0 6 12 18 24 30 36 42 48 54 60
Urinar y I/ SCN rat io ( µg/ m g)

Figure 5. Distributions of urinary Iodine/Thiocyanate ratios in two groups of

Central African schoolchildren.

The mean I/SCN ratio failed to indicate the possible

goitrogenic effect in the western area using the postulated cut-off
< 3 µg /mg, see Figure 5.

Controlling IDD – Studies for Program Management

The literature review in paper II revealed that the ratio is poorly

defined. It has been calculated in several different ways yielding
widely different results, namely as averages of individual ratios:

mean I/SCN (Hennart et al. 1982) and median I/SCN

Mathematically expressed as:

I1 In
( _________ + ... + ________ )
mean _________________________ OR median ( ________ ... _________ )

and as ratios of group averages:

mean I/mean SCN (Delange 1980) and median I/median SCN

(Konde et al. 1994)

Mathematically expressed as:

∑ ( I 1-n )
( _____________ )
n median ( I 1-n )
_____________________ OR ___________________________
∑ ( SCN1- n ) median ( SCN1-n )
( ______________ )

Using material from the western part of the Central African

Republic and from Tanzania the four different options were
calculated, Table 12. These different approaches yield different
results varying by a factor of 2 to 4. This is partly explained by
the effects of skewed distributions. Furthermore, it can be
mathematically shown that the ratio obtained from the group
means gives larger weight to the individual ratios with high SCN
values, causing this mean to be lower than the mean obtained
using individual ratios.

Stefan Peterson

Table 12. Comparisons of urinary I/SCN ratios calculated in different ways for
data from the Central African Republic (CAR) and Tanzania.
CAR Tanzania

n=132 n=217
Averages of individual I/SCN ratios
Mean µg/mg (±SEM) 10.1 (±1.7) 35.0 (±10.0)
Median µg/mg 5.2 9.4

Ratios of group I/SCN averages

Mean µg/mg 6.9 7.5
Median µg/mg 4.2 8.7

Crude Goiter Rate 47% 59%

I (µg/dl)
mean (±SEM) 9.4 (±1.4) 5.6 (±0.4)
median 4.8 3.6

SCN (mg/dl)
mean (±SEM) 1.4 (±0.08) 0.74 (±0.08)
median 1.1 0.41

There are also physiological limitations in the use of the

ratio. Above the kidney threshold for thiocyanate re-absorption
around 1.5 mg/dl, the urinary thiocyanate concentrations are not
linearly related to the serum concentrations that could exert the
possible antithyroid effect (Bourdoux 1995; Rosling 1994). The
ratio will also be affected by the 10-12 fold seasonal variation of
thiocyanate concentrations (Casadei et al. 1990).
We found that parts of the Central African Republic are severely
to moderately affected by IDD and that goitrogens from
insufficient cassava-processing covaries with increased
prevalence of goitre (Peterson 1994). This covariation has
subsequently been confirmed by others (Biassoni et al. 1998).
However, the main cause of the observed goiters is iodine
deficiency. Distributing iodine should be the main priority as it
can overcome the possible thiocyanate effects (Delange et al.
1994; Gaitan & Dunn 1992). Promotion of effective cassava
processing can complement iodine distribution, but it should not

Controlling IDD – Studies for Program Management

delay the initiation of the iodine supplementation. Effective

processing of cassava is associated with much lower urinary
thiocyanate levels. This has subsequently also been
demonstrated in West Africa (Bilabina et al. 1995). There is no
justification to discourage the consumption of cassava from an
IDD point of view and continued studies of goitrogens during the
assessment stage of an IDD program can detract from the
number one priority − to distribute iodine.
The iodine/thiocyanate ratio has mathematical and
physiological shortcomings. Its further use should be
discouraged. It is recommended to judge absolute iodine and
thiocyanate levels independently instead of combining them into
a ratio, which obscures the absolute levels (Figure 1).

Distributing Iodine (paper III)

Iodine can be distributed in many ways. Many countries have
included IOC in their iodine supplementation programs but
Tanzania has to our knowledge had the most extensive
experience. While many studies have examined IOC efficacy
using different doses and time intervals, the literature does not
contain any information on large-scale supplementation
programs’ coverage and cost. We set out to determine the
proportion of targeted person-time effectively covered according to
program objectives, as well as the cost and effectiveness of
different strategies to distribute iodised oil capsules to targeted

Materials and Methods

In study three, we examined 57 distribution rounds of iodized oil

capsules from 1986 to 1994 in 27 Tanzanian districts with a total
population of 7 million. We reviewed IOC distribution reports
and administrative records, interviewed past and present
program managers, and conducted supervision visits to selected
districts. The primary and secondary data was analyzed using a
computer spreadsheet model for the proportion of targeted
population-time covered as outlined in Figure 6.

Stefan Peterson

Number of


first distribution


covered covered

2 years delay of Time

Figure 6. Person-time covered in the distribution of Iodized Oil Capsules -
schematic representation

From the start of district IOC distribution, the objective was

to cover everyone between 1 and 45 years of age every two years.
Census data for each district was updated to the year of the
analysis, adjusted for the proportion of the population in the
target age-group, and used as denominator to calculate the
coverage in any one distribution round (Mbalilaki 1988). The
mean of individual districts’ distribution coverages was
calculated for first and repeat distribution rounds as well as for
all combined. Based on the objective to repeat distribution after
two years the total population-time covered was calculated
(hatched areas in Figure 6). This covered population-time was
divided by the total population-time targeted (total area under
curve in Figure 6) to yield the population time covered according
to program objectives.
For the 20 distribution rounds in 1992-93 a detailed review
was done of the distribution strategies planned and finally
employed. Costs were determined from a health care provider
point of view including opportunity costs of labor and vehicles but
excluding overhead cost to the donor as well as direct and indirect
costs to the beneficiaries. Cost per dose was determined by
dividing the total drug cost and the distribution cost by the
number of doses reported ingested. Comparisons were made
between the three distribution strategies used.

Controlling IDD – Studies for Program Management


Reported district distribution coverage ranged from 20% to 96% of

the targeted population, with an overall mean coverage of 64%.
Coverage declined over subsequent distribution rounds. Multiple
linear regression analysis suggested that coverage was
significantly and inversely related to distribution round (p <
0.05), and positively related (not significantly) to initial visible
goiter rate severity (p =0.2).
Repeat distribution was not timely, occurring with a mean
delay beyond target of 1.25 years. Taking also this into account,
43% of targeted person-time was covered according to program
objectives of distributing at two-year intervals. Of the person-
years not covered, 42% were due to less than total coverage and
58% due to delayed repeat distribution beyond the planned two-
year interval.
In a sub-study of the 20 distribution rounds between 1991
and 1993, we found that three different distribution strategies
had been used to organize mass distribution campaigns in one
particular day in each village. Some districts received central
funding for fuel and health worker per-diems in addition to
receiving the IOC. They set up a “district team” which toured the
district using district vehicles to organize and distribute iodine
supplements in all villages. Other districts received only the IOC
and were told to integrate the distribution into Primary Health
Care (PHC). Eight out of the nine districts attempting this
distribution strategy did not accomplish the task before the
capsules were about to expire. To ensure rapid distribution
before expiration date, some of these districts set up a “national
district team” to facilitate iodine distribution. Staff from the
national program initiated and supported the distribution. These
teams, using government cars, were supported with money for
fuel and per-diem pay.

Stefan Peterson

Table 13. Mean coverage and cost per ingested dose (2 IOC) in 1992 USD.

Distribution strategy District team PHC National+ Total

District team

No. of distribution rounds

planned 11 9 4 20
carried out and analyzed 10 1 4 15

Mean coverage 61% 56% 68% 62%

Cost of 2 IOC 0.30 0.30 0.30 0.30

Mean cost of unaccounted for IOC /

dose reported ingested* 0.16 0.22 0.20 0.18

Mean management & distribution cost /

dose reported ingested 0.05 0.03 0.07 0.06

Mean total cost /

dose reported ingested 0.51 0.55 0.56 0.53
(for 2 year duration)

The “district team” approach was the most cost effective at 51

cents per dose, corresponding to an annual cost of 26 cents, see
Table 13. The analysis was driven by the lower proportion of IOC
unaccounted for under this approach. This finding was fairly
insensitive to different assumptions on the proportion of
unaccounted-for IOC actually wasted. The cost analysis explains
this by revealing that the cost of the oil capsules constitute more
than 90% of program cost at current coverage levels, Table 14.

Controlling IDD – Studies for Program Management

Table 14. Cost profiles for iodized oil capsule distribution in a district using
"district team" distribution and for the Expanded Program on Immunization.
Programs Iodized oil capsule Expanded Program on
distribution Immunization

IOC/vaccine 92% 10%

Transport 3% 13%
Social Mobilization 5%
Labour 2% 42%
Equipment NA 13%
Building and other costs NA 5%
Mgmt, training and supervision 3% 12%
TOTAL 100% 100%
Coverage 63% 80%

Source: Mbozi district distribution report and Bateson (1992).

NA: Not applicable and therefore not costed.


The tendency for distribution coverages to decline over

subsequent distribution rounds has been observed previously
with Vitamin A distribution campaigns (West & Sommer 1987). It
is attributed to “program fatigue” among health workers (Greiner
1991) and the emergence of rumors that the IOC was a secretly
promoting “family planning” or other schemes to control the
population (Magambo et al. 1990). Although the difference was
not significant, the tendency towards higher population coverage
in places with more visible goiter suggests that population
motivation is higher in such places and that IOC, and possibly
other direct-contact supplementation methods, will work best
where the population is highly motivated. The need for
information and communication activities to increase the
population’s awareness and motivation in this regard is crucial.
Control of iodine deficiency with IOC requires direct
interaction with each recipient at regular but long intervals as
there is no “herd immunity”, such as in immunization where an
80% coverage may stop transmission. In contrast, 100% of the
population needs to be covered regularly and timely to eliminate
IDD. This was not achieved. Distribution coverage was
incomplete and delays to repeat distribution common and
substantial. In fact, there are bigger potential gains from

Stefan Peterson

reducing distribution delays than from increasing distribution

coverage. This highlights the need for proper and careful
planning of iodine distribution in order to avoid administrative
delays and to ensure that iodine distribution occurs during an
appropriate agricultural season. This has implications for the
distribution chain: the donor, the national program, the district,
and the local leadership.
More intensive supervision of IOC distribution in more
expensive distribution-campaigns is cost-effective compared to
unsupervised primary health care distribution, see Table 13. The
“district team” approach adds outside funds for distribution
expenses to the district and brings coverage rates up, cuts losses
of IOC, and lowers the annual cost of supplementation to USD
0.26. The cost analysis explains this finding since more than 90%
of the total cost is for the drug. This points out the high cost of the
IOC itself at 15 US cents per capsule (1992). By comparison, a
Vitamin A capsule costs only 2.8 US cents and the capsules
constituted 40% of the total average cost for a large scale
distribution program in the Philippines (Loevinsohn et al. 1997).
Providing expensive IOC to an underfunded and over-stretched
distribution system by “integrating into Primary Health Care”
may therefore have been based more on ideology than evidence.
At present drug prices there is scope for increased cost-
effectiveness of IOC delivery through investments in distribution,
supervision and community mobilisation rather than just
delivering an expensive drug to expire on a pharmacy shelf. This
lesson should apply also to other high item-cost health
interventions with intermittent distribution.
Onchocerciasis control programs that annually distribute
ivermectin have proven to be successful when communities direct
the treatments themselves. Community directed treatment is
organized in endemic areas by explaining program objectives and
strategies to the affected population. The population decides on
how to organize the distribution and, most importantly, selects its
own community-based distributors who are trained on basic
principles of the treatment and then organize with their
communities to treat everyone during a convenient time of the
year. Using village population lists, coverages of 82% rising over
the years to 92% have been achieved in Ecuador villages of about
5,000 persons (Guderian et al. 1997). In Uganda an analysis of

Controlling IDD – Studies for Program Management

the sustainability of the distribution indicated that a significant

predictor of sustainable distribution over several years was
whether the community had chosen the distributor itself
(Katabarwa & Mutabazi 1998). Community ownership, direct
and visible effects of the treatment, regular availability of drugs,
and community confidence in the initiative are all listed as
reasons for the strategy’s success (Tarimo 2000). Lessons to learn
for iodized oil capsule distribution include the need to inform
communities and let them direct the activity in order to achieve
sustainability. However, this success comes at a cost, the per
capita distribution cost alone was estimated at USD 0.29 to
achieve an 85% coverage in a Ugandan population of 46,000
(Kipp et al. 1998).

Evaluating Program Impact (paper IV,V)

Iodizing salt is the long-term strategy chosen by most countries.
WHO and ICCIDD recommended iodization at 100 parts per
million for bulk production in warm, moist climates (WHO
1994a). Tanzania adopted this recommendation. Following the
reports of iodine induced hyperthyroidism among older subjects
with large goiters WHO lowered the recommendation to 20-40
ppm and suggested confirmatory surveys of urinary iodine to
ensure a median of 100-200 µg/l (WHO 1996). Such surveys are
customarily done in school-children, yet even more vulnerable
target groups are the pregnant woman, the foetus and the
newborn. However, no biological indicators are usually assessed
in these groups of prime beneficiaries due to logistical difficulties
(WHO 1997). Yet knowing their status and relation to school-
children’s indicators is critical before taking a decision to change
salt iodine content. We studied the relation of iodine deficiency
indicators in different vulnerable groups in a Tanzanian
community with full coverage of salt presumably iodized at 100
ppm at its site of production.
A convenient marker of the last years’ iodine status in a
community is the distribution of thyroid sizes in schoolchildren
(Aghini-Lombardi et al. 1997). This indicator is then used as a
marker for the likely presence of other, more serious disorders

Stefan Peterson

such as cretinism and infant deaths as well as a summary

indicator of the adequacy of the iodine supply (Dulberg 1985). A
number of different classification systems exist for grading
thyroid size using palpation or ultrasound. In addition to the
reduction in number of goitre grades, the 1994 WHO palpation
system implies a change in definition of what constitutes a
“goiter”. Furthermore, ultrasound estimation of thyroid volume
has been recommended as more precise and objective than
palpation (WHO 1994a). We studied the implications of the
change in palpation goitre definition and compared the
obtainable precision using palpation with that of ultrasonic
estimation of thyroid volume in a goiter survey in rural Tanzania.

Material and Methods

Relation of outcome indicators and target groups (paper IV)

We studied the iodized salt distribution coverage and iodine
status in different vulnerable groups in Ilembula village in the
severely iodine deficient Njombe District (Wächter et al. 1986;
Wächter et al. 1985), where iodized salt had been available for 1-
2 years (Sundqvist & Wijetunga 1996). We surveyed village salt
iodine content twice during one month using rapid test kits for
household and school-children’s samples and titration for retail
samples (Sullivan et al. 1995). Thyroid size was graded according
to Perez et al. (1960) in school-children and women in households
and delivering at Ilembula Hospital. Two examiners graded
independently and settled differences in grade through re-
examination and discussion. Urinary iodine concentration was
measured in schoolchildren and women at delivery. Cord blood
was collected on filter paper from consecutive deliveries and
frozen at minus 20 until analysed by the Neonatal Delfia
technique at the National Swedish Hypothyroidism laboratory.
The iodine deficiency indicators collected in the different groups
were evaluated against WHO severity “benchmarks”, see Table 9
(WHO 1994a).

Controlling IDD – Studies for Program Management

Goiter definition (paper V)

In this study three independent, blinded examiners did double
estimation of thyroid size with the old and new palpation
classification systems in 225 primary school children in a
highland Tanzanian village. Palpation results were compared to
double ultrasound estimation of thyroid volume according to
Brunn et al. (1981 and Delange et al. (1997). Kappa values were
calculated as measures of agreement. Logistic regression was
used to study the influence of ultrasonographically determined
thyroid size parameters on whether a palpator considered a child
goitrous or not. Using ultrasound “goiter” by WHO cutoff values
as gold standard the sensitivity and specificity of the two
palpation systems was calculated (WHO & ICCIDD 1997).

Table 15. Iodine deficiency (ID) indicators and WHO severity grading in different
population groups (WHO 1994a)

School- Household Women at Neonates

children Women Delivery
n 60 25 111 123

Total Goiter Rate (%) 62 52 55

(95% conf int) (49-74) (31-73) (46-64)
ID severity severe severe severe

Urinary iodine (µg/l)

Median 140 68

<20 µg/l 0% 10 %
<50 µg/l 7% 32 %
<100 µg/l 28 % 62 %
ID severity none mild

TSH (mU/l whole blood)

Median 1.5

>5 mU/l 25 %
>7.5 mU/l 15 %
>10 mU/l 10 %
ID severity moderate

Stefan Peterson


In Ilembula village semiquantitative test-kits indicated that 98%

of household salt-samples contained iodine at 25 ppm.
Twentynine retail samples had a median titration iodine content
of 27 ppm (range 11-127), but only 17 % complied with the
national recommendation (>37 ppm). Thus almost all salt from
households was commercially produced and iodized during the
study month, although at relatively low levels. Iodine deficiency
indicators and their severity grading are given in Table 15.
Goiter rates indicated “severe“ iodine deficiency in all groups
examined suggesting delayed goiter regression since median
urinary iodine in the school-children was normal. However,
irrespective of whether the delivering women came from the local
ward, local or neighbouring district cord-blood TSH and urinary
iodine indicated “moderate“ and “mild“ ID, suggesting that the
neonates were still at risk of brain damage.

Our comparative study of the two palpation systems and the

"gold standard" ultrasound demonstrated that the recent change
in goiter definition entailed an increase in estimated goiter rates
by up to 25% in the studied population. This is explained by the
effective lowering of the cut-off between a normal thyroid and a
“goiter”, Table 3. This has major implications for follow-up
studies using the new classification system when baseline
studies were done with the old system, such as in Tanzania. At
global level the number of IDD affected will increase considerably
due to the combined effects of lowering the cut-off for goiter and
decreasing the acceptable prevalence to 5%. Furthermore, the
lowering of the cut-off for goiter decreases specificity to a level of
around 30% where it becomes impossible to demonstrate
elimination of IDD with the new classification system due to false
positives, Figure 7.

Controlling IDD – Studies for Program Management

WHO 1960 System WHO 1994 System

Measured Goiter Rate Measured Goiter Rate
100% 100%

90% 90%

80% 80%

70% 70%

60% 60%
True Neg
50% 50%
False Neg
40% 40%

30% 30% False Pos

20% 20%
True Pos
10% 10%

0% 0%
0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

True Goiter Rate True Goiter Rate

Figure 7. Graphical illustration of goitre criterion performance using body-

surface normated thyroid volume by ultrasound determined goiter as gold
standard in 225 primary school children in highland Tanzania. Palpation
using WHO 1960 criterion: sensitivity 56%, specificity 76%. WHO 1994
criterion: sensitivity 80%, specificity 29%. After Taube 1986.

There is considerable random misclassification of goiter status

upon repeated palpation which, however, results in minimal
changes in estimated goiter rates, see Table 16. The level of
reclassification in palpation with a kappa value of 0.57 is very
close to the result for ultrasound, 0.63 indicating similar precision.
Palpation is more influenced by thyroid width than by thyroid
length or depth, while ultrasound estimates all three parameters
equally. Ultrasound estimation of thyroid volume is demonstrated
to be imprecise due to measurement error and lobe shape
deviation from the ellipsoid form. This results in a wide confidence
interval where the 95% confidence interval for an estimated
volume of 10 ml is ±4 ml. Converted to a goiter rate this results in
considerable random misclassification making ultrasound, like
palpation, unreliable at the individual level but reliable at
population level as long as the misclassification is random.

Stefan Peterson

Table 16. Intra-rater variation. Cross-tabulations of repeated palpation

(examiner A) and ultrasound determinations of total goitre rate in 225
primary school children in highland Tanzania. Number of children and (Total
Goiter Rate).



Examination Results
First round Goiter

Second round Goiter No Yes TOTAL

Palpator A No 90 23 113
(WHO'60) Yes 25 87 112 (50%)
TOTAL 115 110 225 21% 1% 0.57
Palpator A No 33 21 54
(WHO'94) Yes 16 155 171 (76%)
TOTAL 49 176 225 16% -2% 0.53
Ultrasound No 41 17 58
Body surface Yes 14 153 167 (74%)
Normated TOTAL 55 170 225 14% -2% 0.63
* Labelling of agreement according to (Altman 1991): kappa <0.2: poor; 0.21-
0.4: fair; 0.41-0.6: moderate; 0.61-0.8: good; 0.81-1: very good

The study of outcome indicators revealed that while all the salt
was iodized the iodine concentrations varied greatly at retail
level indicating either poor quality of iodation or that large losses
of iodine had occurred during transport and storage. We found
that, while an assessment of the easily accessible and
customarily evaluated school-children showed adequate or even
slightly high urinary iodine, other vulnerable groups lagged
behind. Women at delivery showed signs of inadequate iodine
status and goiter, although the relation between iodine intake
and thyroid hormone status in pregnancy is not clearcut and
pregnancy-induced changes cannot be ruled out (Elnagar et al.
1998; Eltom et al. 2000b). However, newborns showed signs of
thyroid stress indicating a high turnover rate of the little
available iodine (Delange 1998). Hypothyroidism, even if

Controlling IDD – Studies for Program Management

transient, has been linked with negative effects on school

achievement and cognition (Huda et al. 1999) and our findings
suggest that the important target group of neonates may still be
at risk of brain damage at the iodine levels found in the study.
The findings could be explained by insufficient time since the
introduction of iodized salt, or that current salt iodine levels at
household level, while sufficient for school children, may have
been inadequate for other vulnerable groups.
In addition our study demonstrated that goiter regression
may be delayed in all groups compared to thyroid hormone and
urinary iodine levels necessitating longer periods before goiter
reassessments to allow time for new "iodine replete" cohorts of
school-children to replace the formerly iodine-deficient children
(Aghini-Lombardi et al. 1997; Zhao et al. 1999).
It remains to be seen whether universal salt iodization is
feasible in countries with widespread small-scale salt production
or whether integration with other control methods, such as IOC,
will be necessary to achieve elimination of IDD. However, once
iodized salt is available in sufficient quantities appropriate
monitoring is necessary to adopt international recommendations
on salt iodization levels to local circumstances in order to avoid
side effects of excess iodine intake. In evaluating the prevalence
of iodine induced hyperthyroidism, it should also be noted that
already while the iodine deficiency continues the prevalence of
IIH is twice that of iodine sufficient communities (Aghini-
Lombardi et al. 1999). It is important to monitor salt iodine
concentrations during production and distribution down to
household level, as well as to assess iodine status with improved
indicators in all vulnerable groups in the local setting, before
adapting international recommendations on new iodization
levels at production. Measures also need to be put in place to
reduce variations in salt iodine content during production.
Varying salt iodine content may not only be a risk factor for IIH
but also to reduce the effectiveness of iodized salt in reducing
thyroid volume (Zhao et al. 1999).

For goiter surveys in settings where ultrasound surveys are

not feasible, we find that appropriately trained palpators using
the old WHO classification of goiter remain "good enough" to
monitor elimination of iodine deficiency by the goitre criterion.

Stefan Peterson

However, an allowance for a 10% acceptable prevalence must be

made. This 10% cut-off should be applicable for palpation surveys
only, to allow for the 10% false positives that result from a
specificity of 90% which seems to be the typically attainable
performance using the Perez criterion reaches (Smyth et al.
1999). It is, however, not an argument against the lowering of the
cutoff for true prevalence of goiter to 5%. However, it should be
noted that applying the new palpation criterion and the lower 5%
cutoff doubled the estimated number of goitrous individuals in
the world from 2-300 million to 686 million. The population
considered at risk of iodine deficiency simultaneously rose from
1,000 to 1,600 million (Clugston & Bailey 1996; Dunn & van der
Haar 1990; Hetzel 1988; WHO/UNICEF/ICCIDD 1993).
Subsequent additions have increased the population considered
at risk to 2.2 billion −more than one third of the world’s
population− and the number of goitrous to 740 million (WHO
The differences between 225 repeat ultrasound
measurements in our study were larger than errors reported from
small validation exercises elsewhere (Foo et al. 1999;
Zimmerman et al. 2000). However, the attainable precision in
large-scale routine surveys in remote low-income country settings
has not been determined before. Our findings in support of the
continued utility of palpation are important since
ultrasonography is technically difficult and costly to use in the
remote highland regions of low-income countries where IDD is
likely to linger.
The findings also demonstrate the importance of clarity of
definitions in epidemiology and the need to avoid inadvertent
oversights, such as the effective lowering of the goiter definition
in the 1994 ICCIDD/WHO publication which effectively rendered
palpation useless with its low specificity (WHO 1994a). It is
suggested that using the research publication process with peer-
review to scrutinize major changes may help in this regard.

Controlling IDD – Studies for Program Management

Elimination, not Eradication
Joint international and national efforts focusing on salt
iodization has made remarkable progress towards reducing
iodine deficiency in the last decade. Presently 63 out of 99
reporting countries claim that more than 50% of their households
consume iodized salt and in 26 countries the reported level
reaches 90% of the households (UNICEF 2000). However, the
battle is not won and the issue of sustainability is critical. A
nutritional deficiency like IDD will never be eradicated in the
same way as smallpox was. Supplementation needs to be
sustained forever or else iodine deficiency will return (Dunn
1998). To control iodine deficiency, each and every member of the
deficient community needs to be reached regularly with the
intervention as, unlike immunizable diseases, no “herd
immunity” is provided when the large majority of the target
population is covered. Yet another difference between the major
infectious and nutritional disorders among poor populations is
that nutritional deficiencies constitute no threat to the affluent
populations of the world, whereas infectious diseases do.
During the last decade iodine deficiency control programs
have been initiated in most low-income countries with
considerable external support. In the case of Tanzania, this has
amounted to 80% of the program budget (Sanga 1994).
Sustaining such high levels of support seems unlikely. The
interdependence between countries is not as important for IDD
as with infectious diseases such as smallpox, polio, or measles
(Dayrit 1998). Over time the manifestations of IDD may also
disappear making the perceived need for iodine supplementation
less obvious (Alnwick 1998).
The need to build an iodine deficiency control program that is
sustainable is crucial no matter what iodine distribution system
is set up. This also applies to the communication, monitoring, and
enforcement systems. The experience from Tanzania is
illuminating in this regard. The responsibility to monitor the
program is divided between several ministries. Monitoring the
production sites is the responsibility of the Ministry of Water,

Stefan Peterson

Energy and Minerals (MOWEM) and their regional mining

officers. However, a 1998 assessment found that they have not
been supervised since 1995 (Carlsson et al. 1999). Although both
the central Ministry and the zonal officers were supplied with
vehicles for site monitoring, these visits were not taking place,
except when donor-funded and then at a high cost. A subsequent
visit by TFNC to the southern salt producing region revealed that
no monitoring was taking place at the production sites and that
the supervision vehicle had been sold (Assey et al. 1998).
Monitoring of the distribution process is the responsibility of
the National Food Control Commission, Ministry of Health. While
this monitoring initially worked relatively well it is now largely
not performing, or at least not reporting, since outside funding
stopped (Assey et al. 1998). Now only household monitoring
continues under the direction of the Tanzania Food and Nutrition
Centre, also largely dependent on external funding.
Sustainability issues apply also to the internal monitoring
process at the salt factories. While the larger ones were equipped
with titration equipment, site visits have not given evidence of
regular use of this equipment or regular use of a quality control
logbook. Use of rapid test kits seem more widespread but is also
less accurate, providing essentially an indication of whether
iodine is present or not (Carlsson et al. 1999). In summary,
internal and external salt production quality control systems
broke down, the distribution monitoring started well but has
declined; although household monitoring has partially continued,
the monitoring process does not fulfill the adequacy criteria, see
Table 8.
As for the actual production of iodized salt, the donation and
installation of the iodization machines was the beginning of new
procedures. Previous to iodization, salt could be packed directly
into sacks on the salt pan and then taken straight for sale or
storage. Iodization increases the number of production steps
thereby increasing the salt price or decreasing the profit margin.
In small salt producing co-operatives, the relative costs for
transporting the salt to the iodation plant and producing iodized
salt is likely to be even greater than in large operations, perhaps
even exceeding the profit. Careful analysis of the economic
consequences of salt iodization is crucial for sustainability.
The technical challenges include both installation and

Controlling IDD – Studies for Program Management

maintenance of the iodization equipment. In many settings

machines have never been installed due to expense or technical
unsuitability of the equipment. Where installed, upkeep of
machines has proven difficult to organize. Following machine
breakdowns, new more sustainable ways of iodizing were
pioneered using a hand spray pump to put iodine on salt heaps
with subsequent manual mixing (Assey et al. 1998). This method
seems to be more sustainable for iodization in low cost settings,
but the quality of the iodization has yet to be evaluated.
While the Tanzania salt producers’ association was charged
with distributing iodization machines, potassium iodate and
special lined bags, this body has not functioned well. There is
dissatisfaction among producers with the association’s
functioning as a discussion forum and in giving access to
iodization machines and potassium iodate. In fact, the lack of a
sustainable source and mechanism for re-supply of potassium
iodate is one of the major constraints cited by salt producers. The
originally donated stock of labelled, lined polyethylene bags was
soon used up and then largely replaced by second-hand
unmarked bags (Carlsson et al. 1999). These constraints are
shared by salt traders from other countries. A Mombasa workshop
for salt producers from 13 Eastern and Southern African
countries identified the main barriers to effective salt iodization:
the lack of monitoring systems; ineffective enforcement; the
existence of taxes and duties on iodine; legal trade barriers
between countries; and the large number of small scale salt
producers who lack technical skills and resources (Kwena 2000).
Yet Tanzania has made great advances in controlling iodine
deficiency. The preliminary findings from schools surveys in 16
endemic districts indicate sufficient to high urinary iodine
concentrations in all but 2 districts (Kimboka 2000). While the
challenges to uphold and finetune this success are many, this
study concurs with a recent evaluation of the IDD program in
concluding that the main priority in the current Tanzanian
resource-constrained situation is to put a salt iodization
production and monitoring system in place, starting in the
factories. It was suggested that this will require commitment,
accountability, and organizational changes in the concerned
government agencies and NCCIDD as well as a more proactive
secretariat (Carlsson et al. 1999).

Stefan Peterson

Monitoring does not equal Evaluation

Monitoring and evaluation are presented together as the last step
of the global IDD control strategy (Hetzel 1987b; Hetzel 1988). In
the latest published version of the IDD “social wheel,” monitoring
does not appear, presumably this was inadvertently left out,
Figure 2 (Hetzel 1996). Be that as it may, the tendency to see
“monitoring and evaluation” as the same is widespread in
academic medicine and public health and to the disadvantage of
both concepts. Monitoring and evaluation are different activities
(Barutwanayo et al. 1993; Rubin 1995; UNICEF 1990), see Box 1.

Box 1. Definition of terms. Adapted after Barutwanayo et al. 1993; Rubin 1995;
UNICEF 1990.
Monitoring a continuous activity to measure the process of
carrying out activities for program implementation
(also called implementation evaluation or process

Evaluation a periodic assessment of a program’s outcomes and

impact (effectiveness evaluation). It may also
provide information on effectiveness, efficiency,
and sustainability

Effectiveness a measure of the extent to which original

objectives have been achieved

Efficiency an economic term measuring the cost at which the

objective was achieved

Efficacy an intervention’s effect under ideal circumstances,

typically determined in controlled clinical trials

It has been suggested that monitoring findings suggesting

possibilities of impact should precede any impact evaluation
(Greiner 1997). This seems logical. If you have no monitoring data
suggesting that the intervention has reached its targeted
population, why should you spend resources to evaluate impact?
An evaluation, when carried out, should first validate the
monitoring findings and second reassess the situation both in
terms of outcomes, such as changes in behaviors and laboratory

Controlling IDD – Studies for Program Management

parameters, and in terms of impact, changes in morbidity and

mortality (Barutwanayo et al. 1993). Similarly, for obstetric
services it is suggested to monitor availability, utilization, and
quality of the service as process indicators before measuring
maternal mortality (Koblinsky et al. 1994; Maine et al. 1997).
Moreover, the monitoring information will help the interpretation
of the impact findings, thus opening up the “black box” research of
evaluating impact alone in the absence of any explanatory
information on the implementation process (Rossi & Freeman

Defining the implementation process

The widespread practice of confusing monitoring with evaluation
and to concentrate on measuring impact indicators may reflect
the tendency for nutrition programs to concentrate more on “what
to do” than on “how to do it” (Field 1985). There is a need to focus
attention on the implementation process since any intervention is
dependent on a favorable programmatic context to translate its
objectives into community effectiveness. While any one
intervention has a certain efficacy under controlled conditions,
the process of implementing it in a program has a number of
steps that are contingent upon each other to produce the desired
outputs, outcomes, and impacts. For curative care these steps
have been outlined as a formula (Tugwell et al. 1985).
Community effectiveness is the multiplicative combination of
conditional probabilities of efficacy, diagnostic accuracy, health
provider compliance, patient compliance, and coverage. Assuming
independence, Tugwell suggests the following calculation:

Community effectiveness = Efficacy x Diagnostic accuracy x

Health provider compliance x Patient compliance x Coverage

For a supplementation or fortification program applied to an

entire population, diagnostic accuracy does not apply. Coverage
translates into access to the intervention and health worker
compliance translates into quality of the intervention while
patient compliance remains. Thus the components:

Community effectiveness = Efficacy x Access x Quality x


Stefan Peterson

This formula demonstrates that even if the efficacy is 100%

given 50% access, 50% acceptable quality and 50% compliance,
the resulting effectiveness is only 12.5% in spite of the impressive
efficacy. An intervention with lower efficacy may even result in
better community effectiveness if it is easier to implement and
sustain. While the assumption of independence may not fully
apply, and the reduction in effect may be overstated, the
calculation gives the rationale for focusing on “the how” of an
intervention and not just “the what” (Field 1985).
This realization is manifested in the evaluation literature’s
recommended practice of describing a program’s “logical model” as
a basis for planning and subsequently evaluating an intervention
(Goodman 1998; MMWR 1999; Moyer et al. 1997). A logic model
describes the complete sequence of events for bringing about
change, often in the form of a flowchart. This chart can also spell
out infrastructure requirements and interrelationships. The logic
model promotes discussion and clarification of the program’s
strategies, thus focusing and improving the program design
(MMWR 1999).

Box 2. Definition of terms for monitoring and evaluation. Adapted after

Bouchet; DiPrete-Brown et al. 1998; Levinson et al. 1999.

Input Funds, materials, goods, training and actions

necessary for project activities

Process The way activities must be carried out to achieve the

project outputs

Output Provision of project goods and services to the target

population; the primary project deliverables

Outcome Intermediate change in the condition, behavior or

practice of the target population

Impact Changes in the occurrence of targeted conditions in

the target population

Benefits Effects resulting from the impact, usually in

combination with other factors

Controlling IDD – Studies for Program Management

The use of flowcharts or tabulations of steps is also part of the

Quality Assurance practice. Quality assurance sets standards and
guidelines, communicates them, and then monitors and tries to
improve performance in relation to these standards (DiPrete-
Brown et al. 1998). Taking a “systems view” of the program
standards, indicators and thresholds are set for inputs, processes,
outputs, outcomes, and impact, see Box 2 and Figure 8.

Monitoring focuses on inputs, processes, and outputs. Using the
model in Figure 8, a system for collecting and compiling data
needs to be designed as the program’s Management Information
System (MIS). By using the MIS to compile required monitoring
data, the information can then be analyzed and used to guide
program decision-making for improved performance (Bouchet;
DiPrete-Brown et al. 1998; Rossi & Freeman 1993). During the
analysis it is important to prioritize and focus interventions on
the main “bottlenecks” identified in the flowchart or systems
model (DiPrete-Brown et al. 1998; Maine et al. 1997).


Most interventions aim to influence morbidity and mortality by

changing proximate determinants, e.g. to reduce cretinism and
goitre by getting all households to regularly consume adequately
iodized salt. The program’s outputs to achieve this is iodized salt
production and distribution all over the country as well as health
education to enlighten consumers on the importance of using
iodized salt. It is then assumed that these outputs will translate
into outcomes in the form of consumers buying and using iodised
salt and getting adequate iodine intake, see Figure 8. Based on
the scientific literature, it is also reasonable to make the
assumption that adequate iodine intake will then translate to the
desired impact in terms of actual morbidity/mortality reductions.
Therefore, once monitoring has indicated an adequate output, e.g.
good coverage with adequately iodized salt, then the evaluation
can be limited to validate the monitoring findings (output) and to
evaluate the outcome and impact (Levinson et al. 1999).

MONITORING (extends to outcomes) EVALUATION (starts from outputs)
____ ___ ___ ___ ___ ___ ___ ___ ___ ___ _____ _________________________________________________________________________________________________ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___

_____________________________________________________________________________________________________________________________________________________________________________________________ ___ ___ ___ ___ ___ ___ ___ ___ ___ ___ __


Constant supply of Potassium iodate Iodized salt produced Adequate iodine Reduction in Human and
potassium iodate added to all salt using in sufficient quality, intake impairmed brain Economic
iodization machine for quantity and with (assessed as development Development
mixing adequate packaging urinary iodine)
Functional iodization Thyroid size
machine reduction
Quality control Titration samples Iodized salt Hormonal levels Perinatal and
laboratory taken daily distributed to all infant mortality
parts of the country reduction
Training Rapid test kits used
Packaging material All salt packaged in Iodized salt retailed
lined bags appropriately

Iodized oil capsules: IOC distribution with Iodized oil capsules Targeted
• Arriving timely adequate community distributed in right population
• In sufficient quantity involvement and dose timely at swallows IOC
• In good quality support of distribution indicated intervals
Communication Messages delivered Knowledgeable: Targeted
material for: adequately • Salt workers communities
• Interpersonal delivery • Wholesalers buy and use
• Mass media delivery • Retailers iodized salt
• Consumers
Figure 8. A systems view of an iodine fortification / supplementation program with examples of issues for monitoring and
evaluation. Outcomes are both biological and behavioral. Adapted from DiPrete-Brown et al. 1998; Levinson et al. 1999 and
Controlling IDD – Studies for Program Management

However, measuring impacts such as mortality and incidence of

cretinism is often more difficult and costly, while evaluation of the
more proximate determinants at the outcome stage is generally
easier and cheaper to perform. Evaluation of such proximate
determinants of impact has therefore been suggested as
alternative and complementary evaluation measures (Blecher
1996; Lorenz et al. 1995; Mosley & Chen 1984; van Norren et al.
1989). In the case of iodine deficiency, urinary iodine
concentrations is a convenient outcome indicator. For evaluation
surveys it can be combined with the most easy to measure impact
indicator, thyroid size. However, such measurements have
precision problems whether done by palpation or ultrasound
(paper V) and goiter regression is delayed after correction of iodine
deficiency (paper IV). Other impact indicators, while occasionally
necessary to measure for validation studies, require extensive
surveys in groups that may be hard to reach (paper IV).

Modifying the “Social Wheel”

There is a good case for separating process monitoring from

evaluations of outcome and impact. Measurements of the process,
outcome, and impact may all be necessary but their periodicity
should be different and measurement of process and outcome need
to precede, if not replace, the more popular impact evaluations
(Schrettenbrunner & Harpham 1993). Impact evaluations are also
time-consuming, logistically challenging, expensive and yield
results that can only be interpreted together with process
indicators (paper IV). Methodological measurement errors may
also give spurious results (paper II, V).
The need to first monitor the intervention, i.e. salt
iodization, is well reflected in ICCIDD’s current Global Action plan
1999-2001 (ICCIDD 1999a) and in forthcoming WHO monitoring
documents (ICCIDD 1999b). It is therefore suggested to stop using
“Monitoring and Evaluation” as one concept in IDD strategies in
order to recognize that process monitoring is a crucial activity in its
own right. Effective monitoring requires appropriate
administrative arrangements and practical procedures. The
strategy and “social wheel” for IDD control could reflect this by
adding a smaller monitoring wheel on to the bigger one at the
implementation stage. As this monitoring wheel is a small wheel

Stefan Peterson

turning a big one it has to turn around many times while the big
wheel makes one turn, i.e. monitoring has to be done throughout
the implementation years while program evaluation could be done
at 3-5 year intervals, see Figure 9. The same addition is needed in
UNICEF’s “triple A” program cycle (UNICEF 1998).

Program functioning ASSESSMENT

Communication strategy
Outcome & Impact
Program Situation


Inputs Implementation Disseminate

Process Input of Programme Findings
Political UPDATE
Will Plan of Action

arrangements and
Partner responsibilities

Figure 9. Modified ”Social Wheel” of IDD control for the second and subsequent
turns. Superimposed is UNICEF’s “Triple A process” and the smaller process
monitoring wheel, which turns faster than the larger program wheel. Additional
evaluation and updating tasks added. Compare Figure 2.

Adding such a “faster-turning” wheel points out the

importance of monitoring as distinct from evaluation. This
distinction helps clarify the steps and requirements for successful
implementation. This relates particularly to the necessary
distinction between the importance and sequencing of measuring

Controlling IDD – Studies for Program Management

salt iodine content (output) and measuring urinary iodine

(outcome). Using rapid test kits or titration on salt is an everyday
event at production and distribution level the outcome of which
should be monitored at least monthly by the IDD program office.
It is part of the monitoring and enforcement process. Salt
titration laboratories are required down to factory and regional
levels within countries, which is technically feasible since they
are relatively simple to manage. Instructions for sampling are
available and should be adapted to the specific situation
(Sullivan et al. 1995).
Assessing iodine intake by measuring urinary iodine,
however, is an intermittent, much more complex activity, which
serves to confirm findings of adequate salt iodine content and
provide impetus for adjusting salt iodization levels. It therefore
has more a characteristic of a periodic evaluation indicator and
requires careful elaboration of sampling procedures.
Furthermore, the laboratory technology for urinary iodine is
complex. Capacity for determining urinary iodine is thus only
required at national level for larger countries, or can even be
shared among groups of countries, since very few countries will
need to run continuos analysis of urinary iodine levels.

The role of the community in iodine control must also be

reassessed. Enrolling community based non-governmental
organizations in monitoring and advocacy for iodized salt is one
promising avenue (Pandav et al. 1995). While public education to
create awareness and demand for iodized salt is stressed as a
determinant of a successful program (Dunn 1996a), the current
placement of communication in the “social wheel” as a step
preceding the intervention may need to be revised. The reason is
that communication issues also need to become part of the
program implementation concurrent with salt iodization.
Communication also needs to become more relevant to different
contexts. Attention and research needs to be devoted to defining:
the different target audiences and the benefits, costs and other
barriers that interfere with desired behaviors. These audiences
include actors such as the salt workers whom we want to spend
the extra energy and money to iodize all salt appropriately.
Messages need to be developed that will promote these behaviors
most effectively, and the most effective channels of

Stefan Peterson

communication need to be identified for each targeted actor

(Cabanero-Verzosa 1999). Use of such strategic communication
was highly effective in increasing consumer knowledge of IDD
and iodized salt in a campaign in Ecuador (Cabanero-Verzosa
1999). Other target audiences, requiring different messages and
channels will include politicians, salt producers, salt factory
workers, and salt retailers. This can be achieved by more
countries using the available ICCIDD communication strategy
(Ling & Reader-Wilstein). An example is the recent press
campaign in India to maintain the ban on sale of non-iodised salt
(CFAR 2000). Including communication as part of the
intervention also implies that its delivery needs to be monitored
and its effect evaluated in the last stage of the program cycle.

Reassessing the Situation

The strategy

A combined focus on process monitoring and impact evaluation

will assist programs to reassess the feasibility of their control
strategies during the evaluation stage. The blueprinted focus on
universal salt iodization has been largely successful in most
settings (Kavishe 1996). Focusing attention on iodizing the salt
has also been wise. Goitrogens do not merit attention until iodine
has reached the population (paper I,II). The logistics of
distributing iodized oil capsules have also precluded their
widespread use for effective mass supplementation (paper III).
Most iodine deficiency control programs have now reached
maturity and need to reassess the experience of applying this
initial strategy. As they set out on the next turn of the control
cycle, they may need to be less dogmatic; e.g., they may need to
recognize that small scale salt iodation using salt co-operatives
and shared iodization machines may not be feasible in all
settings. Especially as structural adjustment towards fewer,
large-scale salt producers may not be politically acceptable as it
may create soceio-economic inequalities. Programs need the
flexibility to try out new economically and practically viable
alternatives for special contexts. One such example is the
development in Tanzania of iodization of salt in heaps directly on

Controlling IDD – Studies for Program Management

the salt pan, using portable pesticide sprayers. Studies on the

performance of such methods and other practices in use will be
necessary, and form the basis for adapting new international
recommendations for iodization levels (paper IV; PAMM 1998;
WHO 1996).
Ensuring sustainable and functional monitoring of salt
iodization, distribution, and use is crucial to strengthen and
sustain the salt iodization program (Pandav et al. 1995; Wang et
al. 1997). Adequately iodized salt will then supply the large
majority of the population with iodine. The second step is then to
identify geographic or socio-economic pockets where iodized salt
does not work. For example, iodization may not be a viable
alternative when people can access local salt supplies without the
use of a producer, when the market is dominated by many small-
scale salt producers, or when there are price differences between
iodated and non-iodated salt. In such pockets the severity of
iodine deficiency needs to be assessed. If severe enough to
warrant control efforts, distribution of iodized oil capsules should
be considered as an alternative control strategy and monitoring
should be developed. Paper III indicated that IOC distribution
was most effective where IDD was most severe and especially
when local health workers were facilitated to interact with the
local communities. The final step is to devote special attention to
fine-tuning the supplementation to optimally cover vulnerable
groups such as neonates, pregnant women as well as groups with
special dietary practices.

The impact on IDD

Ultimately programs need to demonstrate their impact in

reducing IDD. Some of the impacts may be difficult and costly to
measure, such as reduced incidence of cretinism or decreased
perinatal death rates. Other impacts are comparatively easy to
demonstrate such as reductions in thyroid size and
hypothyroidism. However, it is essential to select optimal
methods for remote and resource poor environments. The
sampling procedures and methods used ultimately depend on the
human and financial resources available to conduct the
evaluation. In countries with per-capita health expenditures
around USD 5 per year these resources are very limited. Research

Stefan Peterson

is needed to validate convenient indicators of outcome and impact

(e.g., urinary iodine and thyroid size in school children) against
impact in more vulnerable target groups (e.g., pregnant women
and newborns) that are harder to reach.
Further developments in this regard could include assessing
program impact through studying the distribution of thyroid
volumes in a community. While ultrasound measures thyroid
volume, the present interpretation of ultrasound data introduces
the same dichotomization as “goiter” by making reference to the
97th percentile. This has led to considerable debate on what
should be the cutoff (Delange 1999; Foo et al. 1999; Xu et al.
1999; paper V). However, dichotomizing thyroids into “normal”
and “goiter” allows no quantification of thyroid size, nor any
statement on how far from “normal” a volume is. It is also not
possible to merge data across age, body surface area, and sex
groups. Similar to anthropometry it is thus suggested to develop
thyroid volume z-scores. (Foo & Mafauzy 1999; WHO 1995).
Further study will be required of the practically attainable
precision of ultrasound surveys under field-conditions and
available budget frames. Standardization of the statistical
methodology for assessing ultrasound measurement error will
help in that regard. The present findings suggest that due to the
combined ultrasound measurement error and error resulting
from the assumption of an ellipsoid thyroid shape, the estimated
volume will differ considerably from the true volume in any one
individual. The apparent precision may thus be largely spurious
as long as the ellipsoid approximation method is used (Brunn et
al. 1981; paper V) while field adaptations of other, more precise
methods could improve the precision (Hegedüs et al. 1983).

The management structure

Most studies in this thesis have focused on biomedical aspects of

methods for monitoring and evaluation. However, as programs
reach their evaluation stage, it will also be necessary to evaluate
the functionality of the institutional arrangements with regards
to monitoring and enforcement. This includes the achievements
and constraints of partners on the National Control Commission
for IDD in fulfilling their envisioned roles. While the NCCIDD
was a great step forward in putting all stakeholders around a

Controlling IDD – Studies for Program Management

table in Tanzania, it does not follow that they have the same
objectives. The difference between the health sector and the salt
industry is not only one of different educational backgrounds, as
often pointed out in the IDD literature, but each has different
incentives and goals. The health sector needs to recognize that
the salt industry’s primary motive and justification is to make a
profit. Effective monitoring, enforcement, and social marketing of
iodized salt will assist the serious partners in that regard, while
poor enforcement will undercut the collaborating salt industries
as non-iodizing competitors will sell 5-10% cheaper (or with a
correspondingly larger profit) in the “large volume, small profit
margin” salt business. Experience from Tanzania suggests that
the institutional arrangements for monitoring need to be
rearranged to give a clear and unified responsibility for
monitoring of salt iodization to one body. Such unified
responsibility for monitoring and management has been
identified as an essential requirement in scaling up a project to a
program (Greene & Kevany 1994).
Donors need to see the totality of their support, so as not to
waste resources overall by attempting to make marginal savings,
e.g. on distribution expenses for expensive medications. Similarly
when partners agree to co-finance an intervention, provisions
must be made to ensure that all necessary ingredients are
actually supplied in adequate quality, such as in an arrangement
where the donor buys the drug and the government foots the
distribution bill. Competitive bidding needs to be used for
purchase of inputs and researchers can facilitate this process by
evaluating comparative efficacies between interventions, e.g.
Chinese and French IOC, rather than merely continuing study
the effect of giving different doses of one manufacturer’s product.
The other, equally important requirement is to set up
procedures that are within the capacity of national agencies’
financial resources, especially with regards to recurrent costs
(Greene & Kevany 1994). Small scale salt producers in Tanzania
unable to sustain the cost of iodization machines, thus resorting
to spraying and hand-mixing, is one example of inappropriate
technology. Another is the inability to sustain factory monitoring
of salt iodization in Tanzania in light of the excessive costs to
transport officials from several ministries with considerable cost
for per-diems for the necessary visits.

Stefan Peterson

Continued modest donor funding also of recurrent costs for

IDD control could indeed be justified as very cost-effective
utilization of donor funds (World Bank 1993) and in line with
what was done for immunization programs. However, if
international organizations and donors are careful to support
establishment of sustainable procedures in each setting and if
appropriate advocacy is done, it could become a national priority
to cover the recurrent costs of an IDD control program. At costs
around 5-10 US cents per person per year salt iodization is
arguably within reach of national resources, even on a 5 dollar
per capita health budget. And provide "... the highest cost-
effectiveness of any health intervention available in the world
today" (World Bank 1993).

Controlling IDD – Studies for Program Management

1. We found that studied parts of the Central African Republic
are severely to moderately affected by IDD. Goitrogens from
insufficient cassava-processing may have aggravated IDD in one
area. Although promotion of effective cassava processing may be
beneficial, such action should by no means delay the
establishment of effective national iodine supplementation.

2. The urinary I/SCN ratio failed to identify the possible

goitrogenic effect observed in our Central African study.
Furthermore, the ratio was found to be mathematically ill-
defined, it has physiological shortcomings, and suffers from
seasonal variation. It is recommended to judge absolute urinary
iodine and thiocyanate levels independently rather than using a

3. In IDD endemic areas not covered by iodized salt, iodized oil

capsule distribution remains an important tool to achieve
elimination of IDD, provided that high coverage is achieved.
However, the overall coverage achieved in Tanzania was only
43% of the targeted population time. The attempt to integrate the
IOC distribution into primary health care resulted in lower
coverage than other strategies in Tanzania. The cost of the IOC
constitutes more than 90% of program costs and significant
savings can be made from reducing capsule waste. Available
funds will be better used if distribution is focused on severely
affected communities and a larger proportion of resources is
allocated to social mobilization and organisation, labor, training,
and supervision. These findings are of general relevance also for
campaign distribution of other medications with long distribution
intervals in low-income countries.

4. All the salt was iodized in a highland study community.

However, the iodine concentrations varied greatly at retail level
and indicated that factory quality control was insufficient or that
losses of iodine during distribution had been larger than expected.
The easily accessible and customarily evaluated school children
showed adequate or even slightly high urinary iodine. However,

Stefan Peterson

other vulnerable groups such as women at delivery and

newborns showed signs of inadequate iodine status. Goiter
regression may be delayed in all groups compared to
normalization of thyroid hormone and urinary iodine levels.
Local studies are necessary to adapt international
recommendations on salt iodization levels to local circumstances.
This will help avoid side effects of excess iodine intake such as
hyperthyroidism, yet ensure adequate iodine intake in the
population. It is recommended to monitor salt iodine
concentrations at both production and distribution down to the
household level, as well as to assess iodine status in all
vulnerable groups as a basis for adapting international
recommendations on new iodization levels at production.

5. The WHO 1994 goiter classification system with the 5%

prevalence cut-off will due to its low specificity make
demonstration of elimination of iodine deficiency by palpation
impossible. This will require programs to invest in costly
ultrasound equipment and training. However, ultrasound
estimation also has considerable errors. Appropriately trained
palpators using the 1960 Perez definition of goiter remains
optimal for monitoring elimination of iodine deficiency in many
resource-poor settings. We recommend a return to palpation with
the Perez goiter criterion with a 10% acceptable prevalence for
future goiter surveys in resource-poor countries.

6. Under the stewardship of ICCIDD, WHO and UNICEF iodine

deficiency control programs across the globe have made
remarkable progress over the last decade. They have benefited
from large international support and used internationally
recommended strategies. As these programs now enter their
second cycle of implementation adjustments need to be made to
each national context. Methodologies for assessing severity of
iodine deficiency and setting up standardized control strategies
have been successful. However, monitoring methodologies now
need to be strengthened, especially with regards to sustaining
and acting on ongoing process monitoring of the implementation.
Starting from the salt production stage, monitoring needs to be
institutionalized step by step. Appropriate sampling procedures
need to be worked out. Once one step is successful monitoring of
the next step needs to follow until there is indication of sufficient

Controlling IDD – Studies for Program Management

implementation to warrant a more refined evaluation. Instituting

a quality assurance process with a committed organization
responsible for maintaining a management information system
and acting on an ongoing monitoring is a top priority. Where
iodized salt fails, iodized oil capsule distribution will need
considerable strengthening to successfully control IDD through
investments in the distribution and monitoring.
For impact evaluation there is need to refine international
guidelines for aspects such as evidence based methodology for
thyroid size estimation, and the relation of iodine status
indicators in different vulnerable groups. In addition to
production of these much needed practical guidelines and
information materials from international organisations, key
methodological developments also need to undergo the peer
review process and be published in regular scientific journals.

Stefan Peterson

Many people have helped out during the course of this work. Thank you all. In
particular I would like to acknowledge:

The many local health workers, village leaders, and community guides that
helped my way around Tanzanian and Central African villages and all the
people that so readily took part in the studies. Thanks are also due to the past
and present IDD researchers and practitioners across the world, in particular
the people in ICCIDD who have pushed the drive against IDD.

Hans Rosling for inspiration, never-ending enthusiasm, mentoring and your

readiness to let me run away and do other things beyond research. Agneta and
Magnus Rosling for hospitality and generosity in lending Hans to me.

Mehari Gebre-Medhin for good scientific discussions and preparing me so well

for work in Africa. Mehari and Susanne for the good injera-wat.

Karin Wennqvist for your life-long dedication to building an international

public health center in Uppsala and your ever competent support.

Ted Greiner for sharing your knowledge, experience, material and contacts so
liberally and for the many manuscript revisions.

Finn Diderichsen for your generosity, support and many useful discussions.

Thorkild Tylleskär for introducing me to field research in Africa, subsequent

collaboration and friendship.

Adam Taube for not only enlightening my student parties at the student
nation but also subsequently my scientific endeavours.

Wilbald Lorri, Festo Kavishe, Benedicta Mduma and “tovarish” Kimboka for
taking me into TFNC so readily. Vincent Assey, Alfred Sanga and Bernard
Bunga for our many days and evenings together in Dar es Salaam and up-
country, the good collaboration and for sharing Tanzania so generously with
me. Nicholas Mlingi and everybody else at TFNC for your hospitality, help and

Francois Légue and Eugène Kpizingui for good collaboration and UNICEF and
the Ministry of Health for support in the Central African Republic. Staff of
ÖrebroMissionen in Sweden and the Central African Republic for collaboration,
support and hospitality., CAR.

Kristine Eklund and Inga Andersson for assistance with computers,

paperwork, bookings and a whole lot more. Birgitta Jennische for sending the
many IDD articles to me and the Medical Library for so competently

Controlling IDD – Studies for Program Management

and enthusiastically digging them out.

Everybody past and present at IMCH and the Dept of Nutrition for help,
companionship and discussions, scientific and other.

Colleagues at Socialmedicin and the Dept of Public Health Sciences in

Stockholm for scientific interaction and the many coffee breaks we spent
discussing statistics, public health and development.

Gösta Tibblin and other former colleagues at the Dept of Family Medicine
where I worked and was inspired during the preparation of this work.

Hampus Eklöf for leaving your expecting family to suffer the hardships of
measuring thyroids in Tanzania. Birger Forsberg for good collaboration and
encouragement. Johanna Sundqvist and Mevan Wijetunga for long days and
nights in Ilembula and the even longer analysis and writing phase. Vincent
Assey, Lena Carlsson, Claes Guthenberg, Gunilla Sparell, and Ann-Sofie
Eriksson for laboratory analyses.

Kurt Svärdsudd and Bengt Höjer for the stimulating discussions and
constructive advice during my “half-time” seminar.

John Dunn, Francois Delange and Pierre Bordoux for inspiration during the
start of my work. Michael Zimmerman, Pavel Langer, and Mária Tajtáková for
stimulating scientific debate.

Linley Chiwona-Karltun for progressing from being a fellow IDD researcher to

becoming part of the family.

My colleagues and friends in Uganda and WHO/IMCI for good collaboration

and many new insights.

Friends in MSF Stockholm for being my social life throughout the process.

My parents Gunvor and Assar Peterson for giving me the support and
encouragement to become what I am. And for paying all those taxes that paid
the way.

My mother-in-law Anne Marie Swartling for support and practical help to allow
me to pursue and finish up this work.

My immediate family: Malin Swartling for critical realism, scientific

discussions, patience and encouragement. Holger for helping out on the
computer and being such a fun distraction.

And then, foremost, the ones I should have listed but forgot to…

This work was supported by Sida / SAREC, InDevelop, Uppsala University, and
the Stockholm County Council.

Stefan Peterson

Assey V, Kusindah B, Bunga B &

REFERENCES Munga J. Progress report on salt
iodation in Kilwa, Lindi and
Mtwara districts. TFNC Report
Aghini-Lombardi F, Antonangeli L,
No 1842. Dar es Salaam: Tanzania
Martino E, Vitti P, Maccherini D,
Food and Nutrition Center, 1998.
Leoli F, Rago T, Grasso L, Valeriano
R, Balestrieri A & Pinchera A. The
Baniahmad A, Tsai S, O'Malley B &
spectrum of thyroid disorders in an
Tsai M. Kindred thyroid hormone
iodine-deficient community: the
receptor is an active and
Pescopagano survey. J Clin
constitutive silencer and a
Endocrinol Metab 1999;84:561-6.
receptor for thyroid hormone and
retinoic acid responses. Proc Natl
Aghini-Lombardi F, Antonangeli L,
Acad Sci 1992;89:10633-7.
Pinchera A, Leoli F, Rago T,
Bartolomei AM & Vitti P. Effect of
Barutwanayo M, Bassalia D,
iodized salt on thyroid volume of
Birabuza A, Celacollette C, Keita
children living in an area previously
M, Madji N, Maiga A, Maoude H,
characterized by moderate iodine
Ndihokubwayo H, Niangue J,
deficiency. J Clin Endocrinol Metab
SidiMohamedO M & Yazipo E.
Malaria control in Africa:
Guidelines for the evaluation of
Alnwick D. Candidate noninfectious
national programs (Africa Child
disease conditions. Global disease
Survival Initiative Combatting
elimination and eradication as public
Childhood Communicable
health strategies. Bull World Health
Diseases (ACSI-CCCD). Atlanta,
Organ 1998;76 Suppl 2:55-60.
Georgia: US Dept of Health and
Human Services, CDC,
Altman D. Practical statistics for
International Health Program
medical research. London: Chapman
Office, 1993.
& Hall, 1991.
Bateson A. Average cost profile
Anonymous. Millet - a possibly
prepared for the 15th EPI Global
goitrogenic cereal. Nutr Rev
Advisory Group meeting.: EPI-
GPV, WHO, Geneva, 1992.
Anonymous. Interrelationships
Benmiloud M, Chaouki ML,
between selenium deficiency, iodine
Gutekunst R, Teichert HM, Wood
deficiency, and thyroid hormones.
WG & Dunn JT. Oral iodized oil
Proceedings of a workshop.
for correcting iodine deficiency:
Aberdeen, United Kingdom, March
optimal dosing and outcome
25-27, 1991. Am J Clin Nutr
indicator selection. J Clin
Endocrinol Metab 1994;79:20-4.
Anonymous. Editorial comment on:
Biassoni P, Ravera G, Bertocchi J,
Interobserver variation in the
Schenone F & Bourdoux P.
assessment of thyroid enlargement: A
Influence of dietary habits on
pitfall in surveys of the prevalence of
thyroid status of a nomadic
endemic goitre. Food and Nutrition
people, the Bororo shepherds,
Bulletin 1994;15:69-70.
roaming a central African region
affected by severe iodine

Controlling IDD – Studies for Program Management

Evidence that cassava ingestion

Eur J Endocrinol 1998;138:681-5. increases thiocyanate formation: a
possible etiological factor in
Biassoni P, Ravera G, Schenone F, endemic goitre. J Clin Endocrinol
Green J & Bertocchi J. Hormonal and Metab 1978;46:613-621.
biochemical patterns in subjects from
a new endemic goiter area in the Bourdoux P, Delange F, Gerard M,
Central African Republic. Mafuta M, Hanson A & Ermans
Thyriodology 1991;3:63-68. AM. Antithyroid action of cassava
in humans. In: Role of cassava in
Biassoni P, Schenone F, Bertocchi J, the etiology of endemic goitre and
Green J, Mela G & Balestra V. cretinism. IDRC Monograph136e.
Ouham-Pendé: a new endemic goiter Ottawa, Canada: International
area in the Centroafrican Republic Development Research Centre,
(C.A.R.). Preliminary observations on 1980.
school children. Thyriodology
1990a;2:35-40. Bourdoux P, Ermans A, Mukalay
wa Mukalay A, Filetti S & Vigneri
Biassoni P, Schenone F, Ravera G, R. Iodine-induced thyrotoxicosis
Balestra V, Green J & Bertocchi J. in Kivu, Zaire. Lancet
Ouham-Pendé: a new endemic goiter 1996;347:552-3.
area in the Centroafrican Republic
(CAR): epidemiological survey on Brunn J, Block U, Ruf G, Bos I,
7261 subjects. Thyriodology Kunze W & Scriba P. Volumetrie
1990b;2:73-80. der Schilddrüsenlappen mittels
real-time sonographie. Dtsch Med
Bilabina I, Fliniaux M, Doh A, Desmet Wochenschr 1981;106:1338-1340.
G, Bour H & Brazier M. Evaluation
de la teneur en iode et en acide Burgi H, Supersaxo Z & Selz B.
cyanhydrique d'aliments consommés Iodine deficiency diseases in
au Togo. Cah Nutr Diét 1995;30:300-5. Switzerland one hundred years
after Theodor Kocher's survey: a
Blecher M. Use of indicators in health historical review with some new
care management. World Health goitre prevalence data. Acta
Forum 1996;17:287. Endocrinol 1990;123:577-90.

Bleichrodt N & Born M. A meta- Burrow G (ed) Neonatal Thyroid

analysis of research on iodine and its Screening. New York: Raven,
relationship to cognitive devlopment. 1980.
In: Iodine deficiency and brain
damage. Report of the Franklin Buttfield I & Hetzel B. Endemic
Symposium 1994. J. Stanbury (ed) goitre in Eastern New Guinea
Philadelphia: Cognizant with special reference to the use
Communications Corporation, 1994. of iodised oil in prophylaxis and
treatment. Bull World Health
Bouchet B. Monitoring the quality of Organ 1967;36:243-262.
primary health care. Bethesda, MD:
Quality Assurance Project. Cabanero-Verzosa C. Strategic
communication for development
Bourdoux P. Measurement of projects. Washington DC: World
thiocyanate in serum or urine yields Bank, 1999.
different information. J Anal Toxicol
1995;19:127. Calaciura F, Mendorla G,
Distefano M, Castorina A, Fazio T,
Bourdoux P, Delange F, Gerard M, Motta R, Sava L, Delange F &
Mafuta M, Hanson A & Ermans AM. Vigneri R. Childhood

Stefan Peterson

IQ measurement in infants with Research and Industry 1992;37:38-

transient congenital hypothyroidism. 41.
Clin Endocrinol 1995;43:473-477.
Choufoer J, van Rhijn M,
Cameron M & Van Staveren W (eds). Kassenaar A & Querido A.
Manual on methodology for food Endemic goiter in western New
consumption studies. Oxford: Oxford Guinea. Iodine metabolism in
University Press, 1988. goitrous and non goitrous
subjects. Clin Endocrinol
Cao X-Y, Jiang X, Dou Z, Rakeman M, 1963;23:1203-17.
Zhang M, O'Donnell K, Ma T, Amette
K, DeLong N & DeLong G. Timing of Clements F, Gibson H & Howeler-
vulnerability of the brain to iodine Coy J. Goiter prophylaxis by
deficiency in endemic cretinism. N addition of potassium iodate to
Engl J Med 1994a;331:1739-44. bread. Experience in Tasmania.
Lancet 1970;:489.
Cao X-Y, Jiang X-M, Kareem A, Dou Z-
H, Rakeman M, Zhang M-L, Ma T, Cliff J, Essers S & Rosling H.
O´Donell K, DeLong N & DeLong G. Ankle clonus correlating with
Iodination of irrigation water as a cyanide intake from cassava in
method of supplying iodine to a rural children from Mozambique.
severely iodine-deficient population J Trop Pediatr 1986a;32:186-189.
in Xinjiang, China. Lancet
1994b;344:107-110. Cliff J, Lundquist P, Rosling H,
Sörbo B & Wide L. Thyroid
Carlsson J, Ismail S, Jitta J & Tekle E. function in a cassava-eating
Working with nutriton. A population affected by epidemic
comparative study of the Tanzania spastic paraparesis. Acta
Food and Nutrition Centre and the Endocrinologica 1986b;113:523-
National Nutrition Unit of Zimbabwe. 528.
Sida Evaluation 99/10. Stockholm:
Sida, 1999. Clugston G & Bailey K. WHO. In:
SOS for a billion. B. Hetzel & C.
Casadei E, Cliff J & Neves J. Pandav (eds). Dehli: Oxford
Surveillance of urinary thiocyanate University Press, 1996.
concentration after epidemic spastic
paraparesis in Mozambique. Journal Clugston G, Beaudry M & Delange
of Tropical Medicine and Hygiene F. Iodine-induced thyrotoxicosis.
1990;93:257-261. Lancet 1996;347:1051.

CFAR. Dissemination on universal Clugston G, Dulberg E & Pandav

salt iodisation. New Dehli: Centre for C. Iodine Deficiency Disorders in
Advocacy and Research. ICCIDD, South East Asia. In: The
2000. Prevention and Control of Iodine
Deficiency Disorders. B. Hetzel, J.
Chanoine JP, Toppet V, Bourdoux P, Dunn & J. Stanbury (eds).
Spehl M & Delange F. Smoking Amsterdam: Elsevier, 1987.
during pregnancy: a significant cause
of neonatal thyroid enlargement. Br J Cobra C, Muhilal, Rusmil K,
Obstet Gynaecol 1991;98:65-8. Rustama D, Djatnika, Suwardi SS,
Permaesih D,
Chauhan S, Bhatt A, Bhatt M & Muherdiyantiningsih, Martuti S &
Semba RD. Infant survival is
improved by oral iodine
Majeethia K. Stability of iodised salt supplementation. J Nutr
with respect to iodine content. 1997;127:574-8.

Controlling IDD – Studies for Program Management

Connolly R. An increase in Scientific publication no 502, 1986.

thyrotoxicosis in southern Tasmania
after an increase indietary iodine. Delange F, Benker G, Caron P,
Med J Aust 1971;1:1268-71. Eber O, Ott W, Peter F, Podoba J,
Simescu M, Szybinsky Z,
Cooke RD & Maduagwu EN. The Vertongen F, Vitti P, Wiersinga W
effect of simple processing on the & Zamrazil V. Thyroid volume and
cyanide content of cassava chips. urinary iodine in European
Journal of Food Technology schoolchildren: standardization of
1978;13:299-306. values for assessment of iodine
deficiency. Eur J Endocrinol
DailyNews. Status of salt industry in 1997;136:180-7.
Tanzania. In Daily News. Dar es
Salaam, 1994. Issue April 29. Delange F, Ekpechi L & Rosling H.
Cassava cyanogenesis and iodine
Dayrit M. Report of the workgroup on deficiency disorders. Paper
noninfectious diseases. Global presented to the International
disease elimination and eradication workshop on cassava safety,
as public health strategies. Bull Ibadan, Nigeria, Acta
World Health Organ 1998;76 Suppl Horticulturae, International
2:80-84. Society for Horticultural Sciences,
Delange F. Administration of iodized Delange F & Ermans A. Role of a
oil during pregnancy: a summary of dietary goitrogen in the etiology
the published evidence. Bull World of endemic goiter on Idjwi island.
Health Organ 1996;74:101-108. Am J Clin Nutr 1971;24:1361-1373.
Delange F. Screening for congenital Delange F, Iteke FB & Ermans
hypothyroidism used as an indicator AM. Nutritional factors involved
of the degree of iodine deficiency and in the goitrogenic action of
of its control. Thyroid 1998;8:1185-92. cassava. IDRC Monograph184e).
Ottawa, Canada: International
Delange F. What do we call a goiter? Development Research Centre,
Eur J Endocrinol 1999;140:486-8. 1982.
Delange F & Ahluwalia R. Cassava Delange F, Vigneri R, Trimarchi F,
toxicity and thyroid: research and Filetti S, Pezzino V, Squatrito S,
public health issues. IDRC Bourdoux P & Ermans A. The
monograph 207e). Ottawa, Canada: iodine/thiocyanate ratio. In: Role
International Development Research of cassava in the etiology of
Centre, 1983. endemic goitre and cretinism.
A.M. Ermans, N.M. Mbulamoko, F.
Delange F, Bastani S, Benmiloud M, Delange & R. Ahluwalia (eds).
DeMayer E, Isayama M, Koutras D, Ottawa, Canada: International
Muzzo S, Niepomniszcze H, Pandav C Development Research Centre,
& Riccabona G. PAHO/WHO 1980.
technical group on endemic goiter,
cretinism, and iodine deficency 5th DeLong G, Leslie P, Wang S, Jiang
meeting 1983 Lima, Peru. Definition X, Zhang M, Rakeman M, Jiang J,
of endemic goiter and cretinism, Ma T & Cao X. Effect on infant
classification of goiter size and mortality of irrigation water in a
severity of endemias, and survey severely iodine deficient area of
techniques. In: Towards the China. Lancet 1997;350:771-3.
eradication of endemic goiter,
cretinism, and iodine deficiency. J. DeLong GR. Dynamics of
Dunn (ed) Washington D.C.: PAHO environmental supplementation of

Stefan Peterson

iodine: four years' experience of Dunn J, Crutchfield H &

iodination of irrigation water in Gutekunst R. Methods for
Hotien, Xinjiang, China. Arch measuring iodine in urine:
Environ Health 1998;53:238-9. ICCIDD/WHO/UNICEF, 1993.
Dunn J & van der Haar F. A
DeMaeyer E, Lowenstein F & Thilly practical guide to the correction of
C. The control of endemic goitre. iodine deficiency: International
Geneva: WHO, 1979. Council for the Control of Iodine
Deficiency Disorders (ICCIDD),
DiPrete-Brown L, Miller-Franco L, WHO, UNICEF, 1990.
Rafeh N & Hatzell T. Quality
assurance of health care in Dustin J & Ecoffey J. A field test
developing countries. Second ed. for detecting iodine-enriched salt.
Bethesda, Maryland: Quality Bull World Health Organ
Assurance Project, 1998. 1978;56:657-658.

Dulberg E. A model for predicting the El Mahdi E, Eltom M, Karlsson F

developmental IDDs from goitre & Mukhtar E. Water goitrogens
data. IDD Newsletter, ICCIDD 1985;1. and endemic goiter in the Sudan.
Trop Geogr Med 1986;38:180-182.
Dunn J. Alternatives to salt and oil for
iodine supplementation. In: The Elnagar B, Eltom A, Wide L,
prevention and control of iodine Gebre-Medhin M & Karlsson FA.
deficiency disorders. B. Hetzel, J. Iodine status, thyroid function
Dunn & J. Stanbury (eds). and pregnancy: study of Swedish
Amsterdam: Elsevier Science and Sudanese women. Eur J Clin
Publishers, 1987a. Nutr 1998;52:351-5.

Dunn J. Iodized oil in the treatment Elnagar B, Eltom M, Karlsson FA,

and prophylaxis of IDD. In: The Bourdoux PP & Gebre-Medhin M.
prevention and control of iodine Control of iodine deficiency using
deficiency disorders. B. Hetzel, J. iodination of water in a goitre
Dunn & J. Stanbury (eds). endemic area. Int J Food Sci Nutr
Amsterdam: Elsevier Science 1997;48:119-27.
Publishers, 1987b.
Elnagar B, Eltom M, Karlsson FA,
Dunn J. Seven deadly sins in Ermans AM, Gebre-Medhin M &
confronting endemic iodine Bourdoux PP. The effects of
deficiency, and how to awoid them. J different doses of oral iodized oil
Clin Endocrinol Metab 1996a;81:1332- on goiter size, urinary iodine, and
1335. thyroid-related hormones. J Clin
Endocrinol Metab 1995;80:891-7.
Dunn J. The use of iodized oil and
other alternatives for the elimination Elnour A, Lieden S, Bourdoux P,
of iodine deficiency disorders. In: Eltom M, Khalid SA & Hambraeus
SOS for a billion. B. Hetzel & C. L. Traditional fermentation
Pandav (eds). Dehli: Oxford increases goitrogenic activity in
University Press, 1996b. pearl millet. Ann Nutr Metab
Dunn J. What's happening to our
iodine. . J Clin Endocrinol Metab Eltom A, Elnagar B, Elbagir M &
1998;83:3398-33400. Gebre-Medhin M. Thyroglobulin
in serum as an indicator of iodine
status during pregnancy. Scand J

Controlling IDD – Studies for Program Management

Clin Lab Invest 2000a;60:1-8. Fernandez RL. A simple matter of

salt: an ethnography of nutritional
Eltom A, Eltom M, Elnagar B, Elbagir deficiency in Spain. Berkeley:
M & Gebre-Medhin M. Changes in University of California Press,
iodine metabolism during late 1990.
pregnancy and lactation: a
longitudinal study among Sudanese Field J. Implementing nutrition
women. Eur J Clin Nutr programs: Lessons from an
2000b;54:429-433. unheeded literature. Ann Rev
Nutr 1985;5:143-72.
Eltom M. Endemic goitre in the
Sudan. [Doctoral Thesis]. Uppsala Fisch A, Pichard E, Prazuck T,
University, Sweden, 1984. Sebbag R, Torres G, Gernez G &
Gentilini M. A new approach to
Eltom M, Elnagar B, Sulieman EA, combating iodine deficiency in
Karlsson FA, Van Thi HV, Bourdoux developing countries: the
P & Gebre-Medhin M. The use of controlled release of iodine in
sugar as a vehicle for iodine water by a silicone elastomer. Am
fortification in endemic iodine J Public Health 1993;83:540-545.
deficiency. Int J Food Sci Nutr
1995;46:281-9. Foo LC & Mafauzy M. Does the use
of mean or median Z-score of the
Eltom M, Hofvander Y, Torelm I & thyroid volume indices provide a
Fellstrom B. Endemic goitre in the more precise description of the
Darfur region (Sudan). Epidemiology iodine deficiency disorder status
and aetiology. Acta Med Scand of a population? Eur J Endocrinol
1984;215:467-75. 1999;141:557-60.

Eltom M, Karlsson FA, Kamal AM, Foo LC, Zulfiqar A, Nafikudin M,

Boström H & Dahlberg P. The Fadzil MT & Asmah AS. Local
effectiveness of oral iodized oil in the versus WHO/International
treatment and prophylaxis of Council for Control of Iodine
endemic goiter. J Clin Endocrinol Deficiency Disorders-
Metab 1985;61:1112-7. recommended thyroid volume
reference in the assessment of
Emamian S, Nielsen M & Pedersen J. iodine deficiency disorders. Eur J
Intraobserver and interobserver Endocrinol 1999;140:491-7.
variations in sonographic
measurements of kidney size in adult Fresco L. The agricultural sector in
volunteers. Acta Radiol 1995;36:399- the Central African Republic. An
401. assessment of the technical
constraints and potential. World
Ermans A, Bourdoux P, Lagasse R, Bank, Agricultural Sector Report,
Delange F & Thilly C. Congenital Washington DC, 1986.
hypothyroidism in developing
countries. In: Neonatal Thyroid Furnée C. Prevention and control
Screening. G. Burrow (ed) New York: of iodine deficiency: Studies on
Raven Press, 1980a. the efficacy of oral iodized oil.
[Doctoral thesis]. Wageningen
Ermans AM, Mbulamoko NM, Delange Agricultural University,
F & Ahluwalia R. Role of cassava in Netherlands, 1994.
the etiology of endemic goitre and
cretinism. IDRC Monograph136e). Furnée C, van der Haar F, West C
Ottawa, Canada: International & Hautvast J. A critical appraisal
Development Research Centre, 1980b.

Stefan Peterson

of goiter assessment and the ratio of Guderian R, Anselmi M, Espinel

urinary iodine to creatinine for M, Mancero T, Rivadeneira G,
evaluating iodine status. Am J Clin Proano R, Calvopina H, Vierira J
Nutr 1994;59:1415-1417. & Cooper P. Succesful control of
onchocerciasis with community-
Gaitan E. Goitrogens in food and based ivermectin distibution in
water. Ann Rev Nutr 1990;10:21-39. the Rio Santiago focus in Ecuador.
Trop Med Int Health 1997;2:982-8.
Gaitan E & Dunn J. Epidemiology of
iodine deficiency. Trends End. Metab. Gutekunst R. Iodine Deficiency
1992;13:427. cost Germany over one billion
dollars per year. IDD Newsletter
Gandz AS. [Iodized bread in the 1993: 29-31.
prophylaxis of endemic goiter in
children]. Pediatriia 1974;53:77-8. Gutekunst R, Smolarek H,
Hasenpusch U, Stubbe P,
Gerasimov G, Nazarov A, Mayorova N, Friedrich HJ, Wood WG & Scriba
Schischkina A, Arbusova M, PC. Goitre epidemiology: thyroid
Mischenko B & Dedov I. Bread volume, iodine excretion,
iodization for micronutrient thyroglobulin and thyrotropin in
supplementation in iodine deficient Germany and Sweden. Acta
regions of Russia. Thyroid Endocrinol (Copenh) 1986;112:494-
1995;Suppl 1:S-40. 501.

Goodman RM. Principals and tools for Hahn SK. An overview of African
evaluating community-based traditional cassava processing and
prevention and health promotion utilization. Outlook on Agriculture
programs. J Public Health Manag 1989;18:110-18.
Pract 1998;4:37-47.
Hegedüs L, Karstrup S, Veiergang
Grant J. Iodine Deficiency Disorders D, Jacobsen B, Skovsted L &
on the run. Paper presented to the Feldt-Rasmussen U. High
Regional meeting on universal salt frequency of goitre in cigarette
iodization to eliminate IDD in the smokers. Clin Endocrinol
Americas, Quito, Equador, UNICEF, 1985;22:287-92.
New York, 1994.
Hegedüs L, Perrild H, Poulsen L,
Greene J & Kevany J. Projects to Andersen J, Holm B, Schnohr P,
programmes: An international Jensen G & Hansen J. The
perspective. Food and Nutrition determination of thyroid volume
Bulletin 1994;15:227-232. by ultrasound and its relationship
to body weight, age, and sex in
Greiner T. Lessons learned and issues normal subjects. J Clin Endocrinol
emerging from experience with Metab 1983;56:260-263.
universal distribution of vitamin A
capsules in Bangladesh. A report to Hennart P, Bourdoux P, Lagasse R,
the World Bank. ICH, Uppsala Thilly C, Putzeys P, Courtois P,
University, Sweden, 1991. Vis H, Yunga Y, Seghers P &
Delange F. Epidemiology of goitre
Greiner T. Rapid appraisal before and malnutrition and dietary
impact evaluation studies. World supplies of iodine, thiocyanate,
Health Forum 1997;18:66-67. and proteins in Bas Zaire, Kivu,
and Ubangi. In: Nutritional factors
involved in the goitrogenic action

Controlling IDD – Studies for Program Management

of cassava. F. Delange, F. Iteke & A. societatis medicorum suecanæ. A.

Ermans (eds). Ottawa: International Pettersson (ed) Stockholm: Isaac
Development Research Centre, 1982. Marcus Boktryckeri, 1931.

Hetzel B. Iodine deficiency disorders Huda S, Grantham-McGregor S,

and their eradication. Lancet Rahman K & Tomkins A.
1983;ii:1126-1129. Biochemical hypothyroidism
secondary to iodine deficiency is
Hetzel B. An overview of the associated with poor school
prevention and control of iodine achievement and cognition in
deficiency disorders. In: The Bangladeshi children. J Nutr
prevention and control of iodine 1999;129:980-7.
deficiency disorders. B. Hetzel, J.
Dunn & J. Stanbury (eds). ICCIDD. IDD Prevalence and
Amsterdam: Elsevier Science Control Program Data. Central
Publishers, 1987a. African Republic: ICCIDD, 1997a.
Issue May 1997.
Hetzel B. Toward a global strategy for¨~jtd/icci
the prevention and control of iodine dd/mi/idd_032
deficiency disorders. In: The
prevention and control of iodine ICCIDD. IDD Prevalence and
deficiency disorders. B. Hetzel, J. Control Program Data. Tanzania:
Dunn & J. Stanbury (eds). ICCIDD, 1997b. Issue May 1997.
Amsterdam: Elsevier Science¨~jtd/icci
Publishers, 1987b. dd/mi/idd_166

Hetzel B. The prevention and control ICCIDD. ICCIDD Global Action

of iodine deficiency disorders. Plan 1999-2001: ICCIDD, 1999a.
Geneva: ACC/SCN state of the art Issue
series. Discussion paper no 3., 1988.¨~jtd/icci
Hetzel B. The story of iodine
deficiency. An international ICCIDD. Indicators for assessing
challenge in nutrition. Oxford: IDD status. In IDD Newsletter,
Oxford University Press, 1989. 1999b. Issue August.

Hetzel B. Recent progress in the Inglenbleek Y, Jung L, Férard G,

elimination of iodine deficiency Bordet F, Goncalves A & Dechoux
disorders. In: SOS for a billion. B. L. Iodised rapeseed oil for
Hetzel & C. Pandav (eds). Dehli: eradication of severe endemic
Oxford University Press, 1996. goitre. Lancet 1997;350:1542-45.

Hetzel B & Mano M. A review of Jamison D, Mosley W, Measham A

experimental studies of iodine & Bobadilla J (eds). Disease
deficiency during fetal development. control priorities in developing
J Nutr 1989;119:145-151. countries. New York: Oxford
University Press, 1993.
Hetzel B & Pandav C. SOS for a
billion. Dehli: Oxford University Kapil U, Prakash S, Shah A &
Press, 1996. Nayar D. Some factors influencing
loss of iodine from iodized salt.
Höjer JA. Kropfstudien. In: Svenska IDD Newsletter 1996;12:12.
Läkaresällskapets Handlingar. Acta
Karmarkar M & Pandav C.
Interpretation of indicators of

Stefan Peterson

iodine deficiency disorders: Recent Deficiency Disorders. B. Hetzel, J.

experiences. The National Medical Dunn & J. Stanbury (eds).
Journal of India 1999;12:113-7. Amsterdam: Elsevier, 1987.

Katabarwa M & Mutabazi D. The Kohrle J. The trace element

selection and validation of indicators selenium and the thyroid gland.
for monitoring progress towards self- Biochimie 1999;81:527-33.
sustainment in community-directed,
ivermectin-treatment programmes Konde M, Ingenbleek Y, Daffe M,
for onchocericiasis control in Uganda. Sylla B, Barry O & Diallo S.
Ann Trop Med Parasitol 1998;92:859- Goitrous endemic in Guinea.
868. Lancet 1994;344:1675-1678.

Kavishe F. Can Africa meet the goal of Kwena E. Children who don't get
eliminating iodine deficency enough iodine are at great risk. In
disorders by the year 2000? Food and The East African. Nairobi, 2000.
Nutrition Bulletin 1996;17:262-267. Issue February 14-20.

Khan M & Manderson L. Focus groups Lagasse R, P. B, Courtois P,

in tropical diseases research. Health Hennart P, Putzeys P, Thilly C,
Policy and Planning 1992;7:56-66. Mafuta M, Yunga Y, Ermans AM
& Delange F. Influence of the
Kimboka S. Personal Communication dietary balance of
on Tanzania national IDD evaluation iodine/thiocyanate and protein on
1999: TFNC. March, 2000. thyroid function in adults and
young infants. In: Nutritional
Kipp W, Burnham G, Bamuhiiga J, factors involved in the goitrogenic
Weis P & Buttner D. Ivermectin action of cassava. F. Delange, F.B.
distribution using community Iteke & A.M. Ermans (eds).
volunteers in Kabarole district, Ottawa, Canada: International
Uganda. Health Policy and Planning Development Research Centre,
1998;13:167-73. 1982.

Koblinsky M, A T & Daly P. Lamberg BA. Endemic goitre--

Programming for Safe Motherhood: a iodine deficiency disorders. Ann
guide to actioni. Health Policy and Med 1991;23:367-72.
Planning 1994;9:252-66.
Lantum D. Lutte contre les
Kochupillai N. Neonatal troubles dûs à la carence en iode.
hypothyroidism in India. Mt Sinai J Rapport de l'enquête nationale,
Med 1992;59:111-5. 14-25 mars 1991. Ministère de la
Santé Publique et des Affaires
Kochupillai N, Karmarkar M, Sociales, Bangui, République
Weightman D, Hall R, Deo M, Centrafricaine, 1991.
McKendrick M, Evered D &
Ramalingaswami V. Pitutary-thyroid Lantum D. Francophone Central
axis in Himalayan endemic goitre. Africa. In IDD Newsletter, 1992.
Lancet 1973;:1021-1024.
Levinson FJ, Rogers BL, Hicks
Kochupillai N & Pandav C. Neonatal KM, Schaetzel T, Troy L & Young
chemical hypothyroidism in iodine C. Monitoring and evaluation of
deficient enviroments. In: The nutrition programs in developing
Prevention and Control of Iodine countries. Nutr Rev 1999;57:157-

Li J & Wang X. Jixian: A success

Controlling IDD – Studies for Program Management

story in IDD control. IDD Newsletter Oxford University Press, 1996.

Mannar M & Dunn J. Salt
Ling J & Reader-Wilstein C. Ending iodization for the elimination of
iodine deficiency now and forever: iodine deficiency:
ICCIDD. ICCIDD/MI/UNICEF/WHO, 1995.¨~jtd/iccidd/
plan.htm Mao I, Ko Y & Chen MTsoiihs. The
stability of iodine in human sweat.
Loevinsohn B, Sutter R & Costales M. Jpn J Physiol 1990;40:693-700.
Using cost-effectiveness analysis to
evaluate targeting strategies: the Marine D & Kimball O. The
case of vitamin A supplementation. prevention of simple goitre in
Health Policy and Planning man. American Journal of Medical
1997;12:29-37. Science 1922;:163-634.

Lorenz N, Kilima P, Tanner M & May SL, May WA, Bourdoux PP,
Garner P. The right objectives in Pino S, Sullivan KM & Maberly
health care planning. World Health GF. Validation of a simple,
Forum 1995;16:280-2. manual urinary iodine method for
estimating the prevalence of
Lundquist P, Mårtensson J, Sörbo B & iodine-deficiency disorders, and
Öhman S. Method for determining interlaboratory comparison with
thiocyanate in serum and urine other methods. Am J Clin Nutr
(Addendum in Clin Chem 1997;65:1441-5.
1983;29:403). Clin Chem 1979;25:678-
681. Mbalilaki N. Tanzania Census
1988. Dar Es Salaam: Bureau of
MacLennan R, Gaitán E & Clinton Statistics, Ministry of Finance,
Miller M. Observer variation in Economic Affairs and Planning,
grading and measuring the thyroid in 1988.
epidemiological surveys. In: Endemic
Goiter: PAHO, 1969. McCullagh S. The Huon Peninsula
endemic: I. The effectiveness of an
Magambo F, Mlingi N & Bunga B. An intramuscular depot of iodized oil
investigation into factors leading to in the control of endemic goitre.
some resistance to a new innovation: Med J Aust 1963;1:767-9.
The case of iodinated oil capsule
distribution: Rukwa Region, Mlingi N, Poulter N & Rosling H.
Tanzania. Tanzania Food and An outbreak of acute intoxications
Nutrition Centre, TFNC report No from consumption of insufficiently
1242, 1990. processed cassava in Tanzania.
Nutrition Research 1992;12:677-
Maine D, Wardlaw T, Ward V, 687.
McCarthy J, Birnbaum A, Akalin M &
Brown J. Guidelines for monitoring MMWR. Framework for program
availability and use of obstetric evaluation in public health.
services. New York: UNICEF, WHO, MMWR Morb Mortal Wkly Rep
UNFPA, 1997. 1999;48:1-40.
Momburi B. Survey for salt
Mannar M. The iodization of salt for production "foot hill sources" .
the elimination of iodine deficiency The case for Dodoma district. Dar
disorders. In: SOS for a billion. B. es Salaam: Ministry of Water,
Hetzel & C. Pandav (eds). Dehli: Energy and Minerals, 1993.

Mongonou F. Goitre éndemique

Stefan Peterson

dans la region de Bouar. [Thése pour iodization programs. Report of a

le Doctorat en Medicine.]. Université training workshop., Blantyre,
de Bangui, RCA, 1982. Malawi, PAMM, Emory
University, Rollins School of
Moreno-Reyes R, Boelaert M, el Public Health, Atlanta, Georgia,
Badawi S, Eltom M & Vanderpas JB. 1998.
Endemic juvenile hypothyroidism in
a severe endemic goitre area of Pandav C, Pandav S, Anand K,
Sudan. Clin Endocrinol (Oxf) Wajih S, Prakash S, Singh J &
1993;38:19-24. Karmarkar M. A role for
nongovernmental organizations in
Mosley W & Chen L. An analytical monitoring the iodine content of
framework for the study of childhood salt in northern India. Bull World
survival in developing countries. Health Organ 1995;73:71-5.
Population and Development Review
1984;10 (Suppl):25-45. Passmore R & Eastwood M.
Davidson and Passmore Human
Moyer A, Verhovsek H & Wilson VL. Nutrition and Dietetics.
Facilitating the shift to population- Edinburgh: Churchill Livingstone,
based public health programs: 1986.
innovation through the use of
framework and logic model tools. Can Perez C, Scrimshaw S & Muñoz A.
J Public Health 1997;88:95-8. Technique of endemic goitre
surveys. In: Endemic
Murdoch D, Harding E & Dunn J. GoitreGeneva: WHO, 1960.
Persistence of iodine deficiency 25
years after initial correction efforts Peterson S. Cassava, iodine
in the Khumbu region of Nepal. N Z deficiency disorders and konzo in
Med J 1999;112:266-8. the Central African Republic.
MFS Report No 69. International
Mutamba J. Endemic goitre in Child Health Unit, Uppsala
Zimbabwe. [Master's thesis]. ICH, University, 1994.
Uppsala University, Sweden, 1993.
Pharoah P, Buttfield I & Hetzel B.
NCCIDD-TFNC. Map of Tanzania Neurological damage to the fetus
District Goitre Survey (1980-1990). resulting from severe iodine
Dar es Salaam: National Council for deficiency during pregnancy.
the Control of IDD, 1990. Lancet 1971;1:308-310.

O'Brien G, Mbome L, Taylor A & Pharoah P & Connolly K. A

Poulter N. Variations in cyanogen controlled trial of iodinated oil for
content of cassava during village the prevention of endemic
processing in Cameroon. Food cretinism: A long-term follow-up.
Chemistry 1992;44:131-136. Int J Epidemiol 1987;16:68-73.

Özgen A, Erol C, Kaya A, Özmen M, Pharoah PO & Connolly KJ. Effects

Akata D & Akhan O. Interobserver of maternal iodine
and intraobserver variations in supplementation during
sonographic measurement of thyroid pregnancy. Arch Dis Child
volume in children. Eur J Endocrinol 1991;66:145-7.
Pharoah PO, Ellis SM, Ekins RP &
PAMM. Quality assurance, Williams ES. Maternal thyroid
monitoring and enforcement of salt function, iodine deficiency and
fetal development. Clin
Endocrinol (Oxf) 1976;5:159-66.

Controlling IDD – Studies for Program Management

Phillips D. Iodine, milk, and the 1937; 9-25.

elimination of endemic goitre in
Britain: the story of an accidental Sanga A. Towards the elimination
public health triumph. J Epidemiol of iodine deficiency disorders by
Community Health 1997;51:391-393. the year 2000: A challenge for
Tanzania. [Master's thesis].
Pichard E, Soula G, Fisch A, Rhaly London School of Hygiene and
AA, Diarra M, Sebbag R & Gentilini Tropical Medicine, 1994.
M. Prevention of iodine deficiency
disorders in children in the rural Schrettenbrunner A & Harpham T.
African zone. Ann Pediatr (Paris) A different approach to evaluating
1992;39:71-8. PHC projects in developing
countries: how acceptable is it to
Querido A, Delange F, Dunn J, Fierro- aid agencies? Health Policy and
Benitez R, Ibbertson H, Koutras D & Planning 1993;8:128-35.
Perinetti H. Definitions of endemic
goiter and cretinism, classification of Scrimshaw N & Gleason G (eds).
goiter size and severity of endemias Rapid assessment procedures.
and survey technique. In: Endemic Qualitative methodologies for
goiter and cretinism: Continuing planning and evaluation of health
threats to world health. J. Dunn & G. related programmes. Boston:
Medeiros-Neto (eds). PAHO Sci Pub, International Nutrition
1974. Foundation for Developing
Countries (INFDC), 1992.
Rendl J, Bier D, Groh T & Reiners C.
Rapid urinary iodide test. Exp Clin Sjöberg K & Sundlöf G.
Endocrinol Diabetes 1998;106 Suppl Strumasjukdomen före och efter
3:S12-6. jodprofylaxen [Endemic goitre
before and after iodine
Rosling H. Cassava toxicity and food prophylaxis]. Läkartidningen
security. A review of health effects of 1971;68:980-984.
cyanide exposure from cassava and of
ways to preent these effects. UNICEF Smyth PP, Darke C, Parkes AB,
& ICH, 1988. Smith DF, John R, Hetherton AM
& Lazarus JH. Assessment of
Rosling H. Measuring effects in goiter in an area of endemic
humans of dietary cyanide exposure iodine deficiency. Thyroid
from cassava. In: International 1999;9:895-901.
workshop on cassava safety. M.
Bokanga, A. Essers, N. Poulter, H. Solomons NW. There needs to be
Rosling & O. Tewe (eds). Ibadan, more than one way to skin the
Nigeria: Acta Horticulturae, iodine deficiency disorders cat:
International Society for novel insights from the field in
Horticultural Sciences, 1994. Zimbabwe [comment]. Am J Clin
Nutr 1998;67:1104-5.
Rossi P & Freeman H. Evaluation: a
systematic approach. Newbury Park, Sorcini M, Oliveri A, Fazzini C,
CA: SAGE Publications, 1993. Medda E, Grandolfo M, Aghini-
Lombardi F, Balestrazzi P,
Rubin F. A basic guide to evaluation Giovanelli G & Carta S. Incidence
for development workers. Oxford: of congential hypothyroidism
Oxfam, 1995. (CH): An indicator of iodine
deficient areas in Italy. 1995;.
Sandell E & Kolthoff I. Micro
determination of iodine by a Stanbury J (ed) The damaged brain
catalythic method. Microchem Acta of iodine deficiency: cognitive,

Stefan Peterson

behavioral, neuromotor, and 1992.

educative aspects. New York:
Cognizant communication, 1994. Todd C. Hyperthyroidism and
other thyroid disorders
Sullivan K, Houston R, Gorstein J & WHO/AFRO/NUT/99.1
Cervinskas J (eds). Monitoring WHO/NHD/99.1): WHO/ICCIDD,
universal salt iodisation programmes. 1998.
Atlanta, Georgia: PAMM, MI,
ICCIDD, 1995. Todd C, Allain T, Gomo Z, Hasler
J, Ndiweni M & Oken E. Increase
Sundqvist J & Wijetunga M. Intake of in thyrotoxicosis associated with
Iodised Salt and Iodine Status in a iodine supplements in Zimbabwe.
Community - a survey carried out in Lancet 1995;346:1563-4.
the village of Ilembula in Iringa
province, southern Tanzania. 89/1996. Todd C & Dunn J. Intermittent
Uppsala: ICH, Dept of Pediatrics, oral administration of potassium
Uppsala University, 1996. iodide solution for the correction
of iodine deficiency. Am J Clin
Tarimo E. Final report on Nutr 1998;67:1279-1283.
sustainability of community-directed
treatment of onchocerciasis with Tonglet R, Bourdoux P, Dramaix
ivermectin. Ougadogou: African M, Hennart P & Ermans A.
Programme for Onchocerciasis, Interobserver variation in the
APOC, 2000. assessment of thyroid
enlargement: A pitfall in surveys
Taube A. Sensitivity, specificity and of the prevalence of endemic
predicitive values: a graphical goitre. Food and Nutrition Bulletin
approach. Stat Med 1986;5:585-591. 1994;15:64-70.

Thilly C. Goitre. In: Médecine et Tonglet R, Bourdoux P, Minga T &

hygiène en Afrique Centrale de 1885 Ermans AM. Efficacy of low oral
à nos jours. P. Janssens, M. Kivits & doses of iodized oil in the control
J. Vuylsteke (eds). Brussels: of iodine deficiency in Zaire. N
Fondation Roi Baudouin, 1992. Engl J Med 1992;326:236-41.

Thilly C, Contempre B & Vanderpas J. Tugwell P, Bennett K, Sackett D &

Surcharge en Thiocyanate et Carence Haynes R. The measurement
en Sélénium: Des Cofacteurs dans iterative loop: a framework for the
lÉtiologie du Goitre et du Crétinisme critical appraisal of need, benefits
Endémiques au Nord-Zaire. Bull and costs of health interventions.
Acad R Med Belg 1991;145:440-450. Journal of Chronic Diseases
Thorpe-Beeston J, Nicolaides K,
Felton C, Butler J & McGregor A. Tylleskär T. The causation of
Maturation of the secretion of thyroid Konzo. [Doctoral Thesis]. Uppsala
hormone and thyroid stimulating University, Sweden, 1994.
hormone in the fetus. N Engl J Med
1991;324:532-535. Tylleskär T, Banea M, Bikangi N,
Cooke R, Poulter N & Rosling H.
TISCO. Salt production, importation, Cassava cyanogens and konzo, an
distribution pattern and marketing in upper motoneuron disease found
Tanzania. Dare es Salaam: NCCIDD, in Africa. Lancet 1992;339:208-

Tylleskär T, Légué FD, Peterson S,

Kpizingui E & Stecker P. Konzo in

Controlling IDD – Studies for Program Management

the Central African Republic. Vitti P, Martino E, Aghini-

Neurology 1994;44:959-961. Lombardi F, Rago T, Antonangeli
L, Maccherini D, Nanni P,
UNICEF. Chapter 6. Monitoring and Loviselli A, Balestrieri A, Araneo
Evaluation. In: UNICEF Policy and G & Pinchera A. Thyroid volume
Procedure manual. New York: measurement by ultrasound in
UNICEF, 1990. children as a tool for the
asessment of mild iodine
UNICEF. Small salt producers and deficiency. J Clin Endocrinol
universal salt iodation. New York: Metab 1994;79:600-3.
UNICEF Nutrition Section, 1994a.
Wächter W, Mvungi M, König A,
UNICEF. State of the world's Pickardt C & Scriba P. Prevalence
children: Oxford University Press, of goitre and hypothyroidism in
1994b. Southern Tanzania: effect of
iodised oil on thyroid hormone
UNICEF. State of the world's deficiency. J Epidemiol
children: Oxford University Press, Community Health 1986;40:86-91.
Wächter W, Mvungi M, Triebel E,
UNICEF. Universal Salt Iodation Vanthiel D, Marschner I, Wood W,
Update. Vol 1. New York: , 1996. Habermann J, Pickardt C & Scriba
P. Iodine deficiency,
UNICEF. State of the world's hypothyroidism and endemic
children: Oxford University Press, goitre in Southern Tanzania. J
1998. Epidemiol Community Health
1985; 263-270.
UNICEF. State of the world's
children: Oxford University Press, Wang J, Harris M, Amos B, Li M,
2000. Wang X, Zhang J & Chen J. A ten
year review of the iodine
Untoro J, Ruz M & Gross R. Low deficiency disorders program of
environmental selenium availability the People's Republic of China. J
as an additional determinant for Public Health Policy 1997;18:219-
goiter in East Java, Indonesia? Biol 41.
Trace Elem Res 1999;70:127-36.
West KJ & Sommer A. Delivery of
van der Haar F, Kavishe P & Gebre- oral doses of vitamin A to prevent
Medhin M. The public health vitamin A deficiency and
importance of IDD in Tanzania. Cent nutritional blindnes. A state-of-
Afr J Med 1988;34:60-5. the-art review. ACC/SCN State-of-
the-art series. Nutrition policy
van Norren B, Boerma J & Sempebwa discussion paper No 2. United
E. Simplifying the evaluation of Nations ACC/SCN, 1987.
primary health care programmes. Soc
Sci Med 1989;28:1091-97. WHO. Resolution 29. The 39th
Word Health Assembly. Geneva:
Vermiglio F, Lo Presti V, Finocchiaro WHO, 1986.
M, Battiato S, Grasso L, Ardita F,
Mancuso A & Trimarchi F. Enhanced WHO. National strategies for
iodine concentrating capacity by the overcoming micronutrient
mammary gland in iodine deficient malnutrition (EB. 89.R11), 1992.
lactating women of an endemic goiter
region in Sicily. J Endocrinol Invest WHO. Indicators for assessing
1992;15:137-42. iodine deficiency disorders and
their control through salt

Stefan Peterson

iodation. WHO/NUT/94.6. Geneva: inhibitory action of excess iodide

WHO/UNICEF/ICCIDD, 1994a. on organic iodide synthesis in the
normal thyroid. Endocrinology
WHO. Iodine and health: eliminating 1949;45:504-13.
iodine deficiency disorders safely
through salt iodization: a statement World Bank. Investing in health.
by the World Health Organization. New York: Oxford University
Geneva: WHO/NUT/94.4, 1994b. Press, 1993.

WHO. Physical status: the use and Xu F, Sullivan K, Houston R, Zhao

interpretation of anthropometry. J, May W & Maberly G. Thyroid
Geneva, 1995. volumes in US and Bangladeshi
schoolchildren: comparison with
WHO. Recommended iodine levels in European schoolchildren. Eur J
salt and guidelines for monitoring Endocrinol 1999;140:498-504.
their adequacy and effectiveness.
WHO/NUT/96.13. Yazipo D, Ngaindiro LF, Barriere-
WHO/UNICEF/ICCIDD, 1996. Constantin L, Namboua L, Pichard
E, Ndoyo J & Bourdoux P.
WHO. Joint WHO/UNICEF/ICCIDD [Effectiveness of a water
consultation. Review of findings from iodination system for preventing
7-country study in Africa on leveles of iodine deficiency diseases in
salt iodization in relation to iodine Central Africa]. Santé 1995a;5:9-
deficiency disorders, including 17.
iodine-induced hyperthyroidism.
WHO/NUT/97.5. Geneva, 1997. Yazipo D, Ngaindiro LF, Namboua
L, Ndoyo J, Pichard E, Barriere-
WHO. Progress towards the Constantin L & Bourdoux P.
elimination of Iodine Deficiency Combatting iodine deficiency: the
Disorders (IDD). WHO/NHD/99.4. iodization of water in the Central
Geneva, 1999. African Republic [letter;
comment]. Am J Public Health
WHO & ICCIDD. Recommended 1995b;85:732.
normative values for thyroid volume
in children aged 6-15 years. Bull Zhao J, Xu F, Zhang Q, Shang L, Xu
World Health Organ 1997;75:95-97. A, Gao Y, Chen Z, Sullivan K &
Maberly G. Randomized clinical
WHO/UNICEF/ICCIDD. Global trial comparing different iodine
Prevalence of Iodine Deficiency interventions in school children.
Disorders, MDIS Working paper #1. Public Health Nutrition
Geneva: WHO, Micronutrient 1999;2:173-8.
Deficiency Information System, 1993.
Zimmerman M, Adou P, Torresani
Wolff J, Chaikoff I, Goldberg R & T, Zeder C & Hurrell R.
Meier J. The temporary nature of the Persistence of goiter despite oral
iodine supplementation in
goitrous children with iron
deficiency anemia in Côte d'Ivoire.
Am J Clin Nutr 2000;71:88-93.