Anda di halaman 1dari 13





A  basic  quality  function  is  that  of  deciding  whether  the  product  conforms  to  specifications.  This  function  is  generally  
called   acceptance.   To   arrive   at   a   decision,   the   primary   step   is   inspection.   Inspection   is   the   comparison   of   certain  
attributes  and  dimensions  of  a  product  against  specifications  to  find  out  if  the  product  is  within  the  prescribed  limits.  
Inspection by attributes is  the  inspection  whereby  either  the  item  is  classified  simply  as  conforming  or  nonconforming  
with  respect  to  a  specified  requirement  or  set  of  specified  requirements,  or  the  number  of  nonconformities  in  the  item  
is  counted.  Inspection  by  attributes  includes  inspection  for  conformity  of  items  as  well  as  inspection  for  number  of  
nonconformities  per  hundred  items.

Inspection  by  variables  is  the  inspection  by  measuring  the  magnitude  of  a  characteristic  of  an  item.  
Acceptance   inspection   is   a   necessary   part   of   manufacturing   and   may   be   applied   to   incoming   materials,   to   partially  
finished   product   at   various   intermediate   stages   of   manufacturing   process   (in-­‐process   inspection),   and   to   the   final  
product.   Acceptance   may   also   be   carried   out   by   the   purchaser   of   the   manufactured   product.   Inspection   consists   of  
several  steps:  
1. interpretation  of  the  specification;  
2. measurement  of  the  product;  
3. comparison  of  the  product  with  specification;  
4. judgment  as  to  conformance;  
5. disposition  of  the  product;  and  
6. recording  of  the  data  obtained.  
1. 100%   inspection   can   be   a   formidable   task   unless   the   100%   inspection   is   performed   with   automatic   test  
equipment.   In   addition,   it   is   not   always   successful,   particularly   when   a   large   number   of   items   have   one   or  
more   characteristics   that   are   marginal   dimensionally,   in   appearance   or   in   performance   (close   to   or  
concentrated   about   a   tolerance   or   limit   of   appearance   or   performance).   Under   these   conditions,   sorting   by  
manual  or  automatic  methods  is  likely  to  classify  some  conforming  items  as  nonconforming,  and  vice  versa.  
In   addition,   100%   testing   by   manual,   visual   or   automatic   methods   can   be   unsatisfactory.   It   can   sometimes  
degenerate  into  superficial  100%  inspection  when,  in  fact,  sufficient  money,  time  and  staff  are  not  available.  
100%  inspection  is  not  viable  if  the  inspection  method  necessitates  destructive  testing.  
2. Sampling  inspection    
Sampling  may   be   defined   as   the   process   of   removing   an   appropriate   number   of   items   from   a   population   in  
order   to   make   inferences   to   the   entire   population.   In   sampling,   one   must   consider   the   laws   of   probability.  
There   are   certain   risks   involved;   namely,   the   risk   of   error.   A   producer’s   risk   (α)   is   the   probability   of   rejecting  
a  good  batch,  whereas  a  consumer’s  risks  (β)  is  the  probability  of  accepting  a  bad  batch.  
Population  (N)  is  the  totaling  of  all  actual  or  conceivable  items  of  a  certain  class  under  consideration.  
Sample  (n)  is   a   portion   of   a   material   collected   according   to   a   defined   sampling   procedure.   The   size   of   any  
sample   should   be   sufficient   to   allow  all   anticipated   test   procedures   to   be   carried   out,   including   all   repetitions  
and  retention  samples.  If  the  quantity  of  material  available  is  not  sufficient  for  the  intended  analyses  and  for  
the  retention  samples,  the  inspector  should  record  that  the  sampled  material  is  the  available  sample  and  the  
evaluation  of  the  results  should  take  account  of  the  limitations  that  arise  from  the  insufficient  sample  size.   A  
random  sample  is  a  sample  chosen  in  such  a  manner  that  one  object  has  a  good  chance  of  being  selected  as  
another.   It   is   a   sample   in   which   the   different   fractions   of   the   material   have   an   equal   probability   of   being  
Sampling   inspection   is   used   in   lieu   of   100%   inspection   because   of   the   disadvantages   stated   above.   It   is  
common  knowledge  that  on  may  types  of  inspection,  even  several  100%  inspections  will  not  eliminate  all  of  
the   defective   product   from   a   stream   of   product,   apportion   of   which   is   defective.   The   best   protection   is,   of  
course  having  the  product  made  right  in  the  first  place.  
Sampling   inspection   methods   can   be   single,   double,   or   multiple   sampling.   In   single   sampling,   a   decision   is  
reached   after   only   one   sampling.   Figure   1   shoes   the   schematic   operation   of   single   sampling.   In   double  
sampling,   a   decision   is   obtained   after   the   result   of   the   second   sampling   is   known.   Figure   2   outlines   the  
schematic   operation   of   double   sampling.   Multiple   sampling   methods   are   merely   extensions   of   the   above.  
Figure  3  is  the  schematic  operation  of  multiple  sampling.  

Inspect a
sample of n

If the number of
defects found in
the samples  

Does not Exceeds

exceed standard   standard  

Do not accept
Accept the lot   the lot  

Figure  1.  Schematic  operation  of  single  sampling  method.  
Exercises  on  sampling  inspection:  
1. Using   the   double   sampling   method,   3   defectives   were   found   during   the   first   sampling.   Give   the   disposition   of  
the  lot.  
Table   showing   the   criteria   with   normal   inspection  
for  lots  consisting  501-­‐1,200  pieces  of  items.  
Sampling   n   Cum.  n   Ac   Re  
First   50   50   1   4  
Second   50   100   4   5  
2. After   the   fourth   sampling,   four   defectives   were   found.   The   yield   of   the   fifth   sampling   was   three   defectives  
more.  Is  the  lot  approved  or  rejected?  
Table   showing   the   criteria   with   normal   inspection  
for  lots  501-­‐1,200  pieces  of  items.  
Sampling   n   Cum.  n   Ac   Re  
First   20   20   *   3  
Second   20   40   0   3  
Third   20   60   1   4  
Fourth   20   80   2   5  
Fifth   20   100   3   6  
Sixth   20   120   4   6  
Seventh   20   140   6   7  

Inspect a first
sample of n1

Number of
defects found  

Does not
exceed C1   Exceeds C1 but Exceeds C2  
not C2  

Inspect second
sample of n2

Total number of
defects found  

Does not
accept   exceed C2   Exceeds C2   reject  

Figure  2.  Schematic  operation  of  double  sampling  method.  

Inspect a first
sample of n1

Number of
defects found  

Does not exceed Exceeds C1a but

c1a   not C1b   Exceeds C1b  

Inspect second
sample of n2

accept   Total defects reject  

found in the first
and second

Does not exceed Exceeds C2a but

c2a   not C2b   Exceeds C2b  

  Inspect third
sample of n3
Total defects
  found in the
  three samples  
  Does not Exceeds C3a
exceed c3a   but not C3b   Exceeds C3b  
Figure  3.  Schematic  operation  of  multiple  sampling  method.  
A   sampling   plan   is   a   definite   working   rule   regarding   size   and   frequency   of   sample   and   the   basis   for   acceptance   or  
rejection.  It  is  therefore  a  specification  of  sampling.  It  requires  three  numbers  specified.  One  is  the  number  of  items  
(N)  in  the  lot  or  batch  from  which  the  sample  is  drawn.  The  second  is  the  number  of  items  (n)  in  the  random  sample  
drawn   from   the   lot.   The   third   is   the   acceptance   number   (c),   which   can   be   specified   by   the   acceptable   quality   level  
AQL  is  defined  in  the  ABC  standard  as  follows:  “The  AQL  is  the  maximum  percent  defective  (or  maximum  number  of  
defects   per   hundred   units)   that   for   purpose   of   sampling   inspection,   can   be   considered   satisfactory   as   a   process  
average.”   The   acceptance   criteria   for   serious   defects   should   be   more   severe   than   for   trivial   defects.   Thus,   relatively  
low   AQL   values   are   used   for   those   types   of   defects   that   would   have   serious   consequences   and   relatively   high   AQL  
values  are  given  to  those  defects  that  are  of  little  importance.    
Different  sampling  plans:  
1. The   “n   plan”   should   be   used   with   great   caution   and   only   when   the   material   to   be   sampled   is   considered  
uniform   and   is   supplied   from   a   recognized   source.   Samples   can   be   withdrawn   from   any   part   of   the   container  
(usually   from   the   top   layer).   The   n   plan   is   based   on   the   formula   n   =   1   +   𝑁,   where   N   is   the   number   of  
sampling   units   in   the   consignment.   The   value   of   n  is   obtained   by   simple   rounding.   A   minimum   number   of  
containers  needs  to  be  sampled,  e.g.  if  N  is  less  than  or  equal  to  4,  then  every  container  is  sampled.  According  
to   this   plan,   original   samples   are   taken   from   n  sampling   units   selected   at   random   and   these   are   subsequently  
placed  in  separate  sample  containers  (see  Appendix  1).  The  control  laboratory  inspects  the  appearance  of  the  
material   and   tests   the   identity   of   each   original   sample   according   to   the   relevant   specification.   If   the   results  
are   concordant,   the   original   samples   are   combined   into   a   final,   composite   sample   from   which   an   analytical  
sample  is  prepared,  the  remainder  being  kept  as  a  retention  sample.  It  is  important  to  recognize  that  the  “n  
plan”  is  not  statistically  based  and  should  be  used  only  as  a  guiding  principle.  This  is  also  referred  to  as  square  
root  system.  
2. Government  sampling  plan  uses  the  master  tables  which  give  sample  sizes,  acceptance  and  rejection  numbers.  
It  is  necessary  to  know  whether  normal,  tightened  or  reduced  inspection  schemes  are  to  be  followed.    
A  table  for  sample  size  code  letters  is  given  in  Table  1.  This  gives  the  relationship  between  the  lot  or  batch  
size  and  the  code  letter  than  determines  the  sample  size.  The  general  inspection  levels  on  the  right-­‐hand  side  
of  the  tables  are  the  ones  to  be  used  in  most  cases.  Unless  otherwise  specified,  Inspection  Level  II  should  be  
used.  However,  Inspection  Level  I  may  be  specified  when  less  discrimination  is  needed,  or  Level  III  may  be  
specified  for  greater  discrimination.  
Whenever   the   quality   history   is   unsatisfactory   or   when   there   are   other   good   reasons   for   suspicious   about  
quality,   a   severe   acceptance   criteria   is   used.   Thus,   tightened   inspection   is   conducted,   instead   of   normal  
inspection.   When   the   quality   history   is   good   enough,   a   reduced   inspection   is   permitted.   This   allowed   the  
concentration  of  attention  of  inspection  on  those  products  where  attention  seems  to  be  needed  most.  
Normal  inspection  is  designed  to  protect  the  producer  against  having  a  high  proportion  of  lots  not  accepted  
when  quality  is  better  than  the  AQL.  In  fact,  the  producer  is  being  given  the  benefit  of  any  doubt  that  arises  
due  to  sampling  variability.  But  the  consumer  needs  protection  too,  and  this  is  achieved  by  arranging  that  the  
produces  is  not  given  the  benefit  of  the  doubt  blindly  and  invariably,  but  only  for  as  long  as  he  proves  worthy  
of   it.   If   at   any   time   the   sampling   results   show   that   his   process   average   is   probably   worse   than   the   AQL,   he  
forfeits  his  right  to  the  benefit  of  the  doubt  (that  is,  his  right  to  normal  inspection),  and  tightened  inspection  
is  instituted  to  protect  the  consumer.    
The   rule   is   that   tightened   inspection   has   to   be   used   for   the   following   lots   as   soon   as   two   out   of   any   five   or  
fewer   successive   lots   on   original   inspection   have   not   been   accepted.   Once   tightened   inspection   has   been  
instituted,   it   remains   in   force   for   every   lot   until   five   successive   lots   have   been   accepted   on   tightened  
inspection,   then   normal   inspection   is   restored.   This   requirement   is   quite   a   severe   one,   as   acceptance   on  
tightened  inspection  is  more  difficult  than  on  normal  inspection,  but  once  there  is  evidence  that  quality  worse  
than  the  AQL  has  been  produced,  the  producer’s  right  to  benefit  of  the  doubt  cannot  be  restored  until  it  is  safe  
to  do  so.  
Table  1.  Sample  size  code  letters  
Reduced   inspection   is   used   when   there   is   evidence   that   the   product   quality   is   consistently   better   than   the  
AQL.   Where   this   happens   and   there   is   reason   to   believe   that   good   production   will   continue,   sampling  
inspection   no   longer   serves   the   purpose   of   segregating   the   good   lots   from   the   bad   ones.   However,   inspection  
cannot  be  dispensed  with  altogether,  as  a  warning  is  needed  if  the  production  quality  worsens.  
The   four   special   levels,   S-­‐1   to   S-­‐4   at   the   left   hand   side   of   the   table,   are   included   for   the   special   case   where  
relatively  small  sizes  are  necessary  and  large  sampling  risks  can  or  must  be  tolerated.    
The   AQL   and   the   sample   size   code   letter   shall   be   used   to   obtain   the   sampling   plan   from   Tables   2,   3,   4   (for  
single  sampling  plans).   For   a   specified   AQL   and   a   given   lot   size,   the   same   combination   of   AQL   and   sample   size  
code   letter   shall   be   used   to   obtain   the   sampling   plan   from   the   table   for   normal,   tightened   and   reduced  
1. Crude  drugs  
a. Sampling  of  material  in  bulk  
If   initial   inspection   indicates   that   the   batch   is   uniform,   take   samples   as   follows.   When   a   batch  
consists   of   five   containers   or   packaging   units,   take   a   sample   from   each   one.   From   a   batch   of   6–50  
units,  take  a  sample  from  five.  In  the  case  of  batches  of  over  50  units,  sample  10%,  rounding  up  the  
number  of  units  to  the  nearest  multiple  of  10.  For  example,  a  batch  of  51  units  would  be  sampled  as  
for  60  —  i.e.  take  samples  from  six  packages.  
From   each   container   or   package   selected,   take   three   original   samples,   taking   care   to   avoid  
fragmentation.  Samples  should  be  taken  from  the  top,  middle  and  bottom  of  the  container.  In  the  case  
of  sacks  and  packages,  the  three  samples  should  be  taken  by  hand,  the  first  from  a  depth  of  not  less  
than   10   cm   from   the   top   and   the   second   and   third   from   the   middle   and   bottom   after   cutting   into   the  
side   of   the   package.   Samples   of   seeds   should   be   withdrawn   with   a   grain   probe.   Material   in   boxes  
should   first   be   sampled   from   the   upper   layer;   then   approximately   half   of   the   contents   should   be  
removed   and   a   second   sample   taken.   Finally   after   further   removal   of   material,   another   sample  
should   be   taken   from   the   bottom.   Samples   should   be   as   uniform   as   possible   in   mass.   The   three  
original  samples  should  then  be  combined  into  a  pooled  sample  which  should  be  mixed  carefully.  
Table  2.  Single  sampling  plans  for  reduced  inspection  (Master  table)  
The   average   sample   is   obtained   by   quartering.   Form   the   pooled   sample,   adequately   mixed,   into   an  
even   and   square-­‐shaped   heap,   and   divide   it   diagonally   into   four   equal   parts.   Take   two   diagonally  
opposite   parts   and   mix   carefully.   Repeat   the   process   as   necessary   until   the   required   quantity,   to  
within  ±  10%,  is  obtained  (100–  200  g  for  flowers  and  up  to  10  kg  for  certain  roots).  Any  remaining  
material  should  be  returned  to  the  batch.  
Using  the  same  quartering  procedure,  divide  the  average  sample  into  four  final  samples,  taking  care  
that  each  portion  is  representative  of  the  bulk  material.  The  final  samples  are  tested  for  the  following  
— degree  of  fragmentation  (sieve  test);    
— identity  and  level  of  impurities;    
— moisture  and  ash  content;    
— level  of  active  ingredients,  where  possible.  
A  portion  of  each  final  sample  should  be  retained  to  serve  as  reference  material,  which  may  also  be  
used  for  re-­‐test  purposes,  if  necessary.  
Table  3.  Single  sampling  plans  for  tightened  inspection  (Master  table)  
b. Sampling  of  material  in  retail  packages  
From   each   wholesale   container   (boxes,   cartons,   etc.)   selected   for   sampling,   take   at   random   two  
consumer   packages.   From   small   batches   (1–5   boxes),   take   10   consumer   packages.   Prepare   the  
pooled  sample  by   mixing   the   contents   of   the   selected   consumer   packages   and   proceed   as   described  
above  to  obtain  the  final  sample.  
2. Purified  raw  material  
a. Gross  sample  is  computed  using  the  n  plan.  
b. Test  sample  is  the  amount  of  sample  for  one  complete  analysis.  It  is  dependent  on  the  number  of  tests  
required  in  the  drug  monograph.  The  quantity  needed  to  describe  the  material  is  no  longer  counted.  
Identification   and  Solubility  tests  are  reported  on  pooled  sample. Quantitative  tests  are  performed  
c. Laboratory  sample  is  equal  to  three  times  the  amount  of  sample  for  one  complete  analysis.  
d. Retention  sample  is  two  times  the  amount  of  sample  for  one  complete  analysis.  
3. In-­‐process  items  
a. Gross  sample  is  withdrawn  on  a  time  basis  or  as  portions  of  batch.  
b. Laboratory  sample  is  based  on  GMP,  in-­‐house  or  pharmacopeial  requirement.  
Table  4.  Single  sampling  plans  for  normal  inspection  (Master  table)  
4. Finished  products  
The  minimum  size  of  the  samples  will  be  determined  by  the  requirements  of  the  analytical  procedure  
that  will  be  used  to  test  the  product.  Tests  of  unit  dosage  forms  for  uniformity  of  weight,  volume  or  
content   can   require   a   considerable   number   of   units,   as   can   tests   for   sterility.   Depending   upon   the  
type  of  material,  the  size  of  the  consignment  and  the  way  in  which  the  material  is  packed,  a  unit  to  be  
sampled   may   be   regarded   as   the   transport   container,   e.g.   20   packs   shrink-­‐wrapped   or   boxed  
together,   rather   than   an   individual   container.   The   required   number   of   unit   dosage   forms   is   then  
withdrawn  from  any  individual  container  in  the  selected  transit  container.  When  sampling  finished  
products,  packaging  materials  may  be  retained  for  testing.  

In   some   cases   it   may   be   sufficient   to   limit   examination   of   finished   goods   to   visual   inspection   only.   If  
physical  and  chemical  testing  is  required,  however,  the  sampling  units  should  consist  of  whole  packs.  
Individual  packs  should  not  be  broken  open  for  the  purposes  of  sampling.  

5. Packaging  materials  
a. Gross  sample   is   computed   using   the   n  plan.   It   is   the   number   of   packages   to   be   opened   for   withdrawal  
of  samples.  
b. Laboratory  sample  is  determined  using  the  military  standard  for  attributes.  
1. Scoops  
Small  containers  of  solid  materials  may  be  adequately  sampled  using  a  spatula  or  scoop  (See  Figure  4).  The  
samples  are  then  blended  to  provide  a  representative  sample  of  that  container.  The  size  of  the  scoop  depends  
on   the   amount   of   sample   to   be   taken.   A   scoopful   of   sample   should   be   taken   in   a   single   movement   and  
transferred   to   the   sample   container.   Avoid   tapping   the   scoop   to   remove   pharmaceutical   product,   as   this   is  
likely  to  cause  segregation  of  the  sample.  
Figure  4.  Sampling  scoops  for  solids.  
2. Dip  tubes  
Dip  tubes  (See  Figure  5.)  should  be  used  for  sampling  liquid  and  topical  products  and  should  be  made  of  an  
inert  material,  such  as  polypropylene  or  stainless  steel.  
3. Weighted  containers  
For  taking  samples  from  large  tanks  and  storage  vessels,  a  container  in  a  weighted  carrier  can  be  used.  The  
container  is  designed  such  that  it  can  be  opened  at  the  required  depth.  Marks  on  the  cord  used  for  lowering  
the   container   can   be   used   to   determine   when   the   correct   sampling   depth   has   been   reached.   Figure   6   is   an  
example  of  a  typical  weighted  container.  

Figure  5.  Typical  dip  tube.                                        Figure  6.  Typical  weighted  container.  
4. Thieves  
Sample   thieves   should   be   used   when   taking   samples   from   deep   containers   of   solids.   Typical   thieves   are  
shown  in  Figure  7.  
The   plug   thief   typically   consists   of   a   hollow   tube   with   an   inner   rod   that   has   a   tip   on   the   end   to   allow   the   thief  
to   enter   the   powder   bed   in   the   closed   position   (see   Figure   7.i).   The   geometry   of   this   tip   can   influence   the  
sample  taken;  pointed  tips  distort  the  powder  bed  less  than  blunt-­‐tipped  probes,  thereby  reducing  sampling  
error.   Some   thieves   have   a   locking   device   that   allows   the   sample   volume   to   be   set   to   the   required   sample  
weight,  thereby  reducing  the  weight  variation  in  the  sample  population.  
A  chamber  thief  generally  consists  of  two  concentric  tubes  (see  Figure  7.ii);  the  inner  tube  is  solid  except  for  
the   chambers   in   which   the   sample   is   collected.   The   outer   tube   is   hollow   with   openings   that   can   be   aligned  
with  the  chambers  in  the  inner  tube.  A  well-­‐designed  thief  will  have  a  sharp  end  to  minimize  disruption  to  the  
powder  bed.  

Figure  7.  Typical  sample  thieves.  
5. Simple  bag-­‐sampling  spears  
Simple  bag-­‐sampling  spears  are  the  most  commonly  used  instruments  for  taking  samples  from  bags,  because  
they  are  relatively  cheap,  simple  and  quick.  Sampling  spears  generally  have  a  maximum  external  diameter  of  
about  12  mm,  but  can  be  up  to  25mm  in  diameter.  To  obtain  a  good  cross-­‐sectional  sample,  the  spear  should  
be  40–45  cm  in  length.  The  tapered  type  of  sampling  spear  penetrates  bags  easily.  Typical  spears  are  shown  
in  Figure  8.  

Figure  8.  Typical  sampling  spears.  

Exercises  on  sampling  preparation:  
1. For  a  shipment  of  100  labels,  the  acceptance  limit  is  25  defects;  9  defects  were  found  during  sampling.  
a. What  is  the  batch  size?  
b. Give  the  sample  size  code  letter.  
c. What  is  n  using  the  military  standard  tables?  
d. Give  the  AQL.  
e. What  is  Ac  using  the  military  standard  tables?  
f. What  is  the  disposition?  
2. What  is  the  Ac  if  Re  is  23?  
3. Solve  for  the  batch  size  if  n  is  4  using  the  square  root  system.  
4. For  an  AQL  of  0.15,  what  is  the  maximum  number  of  defects  acceptable  for  10,000  pieces  of  cartons?  
5. Fourteen  drums  of  Muriatic  acid  were  received  in  the  warehouse.  
a. If  the  shipment  has  the  same  batch  number  in  all  drums,  how  many  drums  should  be  sampled?  
b. If   the   shipment   consists   of   2   drums   with   one   batch   number   and   12   drums   with   another   batch  
number,  give  the  sample  size  for  each  batch.  
6. In  acceptance  sampling  under  the  ABC  standard,  single  sampling  is  to  be  used  with  Inspection  Level  II,  for  a  
batch  of  100,000  pieces  of  labels.  Give  the  acceptance  criteria  for  normal  sampling  if  the  AQL  is  1.5.  
7. Explain  how  a  sample  size  of  5  units  with  an  AQL  of  150  can  have  an  acceptance  number  of  14.  
8. Fifty  sacks  of  25  kg  each  of  Psidium  guajava  folium  were  received  for  sampling.  Calculate  the  gross  sample  
and  gross  delivery.  
9. Forty  cans  of  Powdered  Luya  Rhizome  were  received  by  the  warehouse.  Inspection  of  the  containers  shows  
that   37   had   the   lot   number   R12-­‐04   printed   on   the   label   while   3   cans   were   identified   with   a   lot   number   of  
R15-­‐04.  How  much  is  the  gross  sample?  
10. Given  below  is  the  list  of  tests  needed  to  measure  the  quality  of  Piperazine  Citrate  USP:  
Identification  :  Chemical  reaction     0.220      g  
Solubility         4.0                g  
Water                                   1.083      g  
Primary  Amines  and  Ammonia       0.5                g  
Assay                                   0.2                g  
a. How  many  grams  of  the  material  are  needed  for  one  complete  analysis?  
b. How  much  is  the  laboratory  sample?  
c. How  much  is  the  retention  sample?  
11. One   hundred   sixty-­‐four   cases   of   glass   bottles   of   30mL,   each   containing   12   dozens,   were   subjected   to  
acceptance  sampling.  Determine  the  following:  
a. gross  sample  
b. laboratory  sample  size  
c. If  cracks  in  some  bottles  were  noted,  
1) what  AQL  should  be  considered:  0.65  or  65?    
2) what  is  the  acceptance  number?  
3) what  is  the  rejection  number?  
British  Standard  on  sampling  procedures  for  inspection  by  attributes  –  Part  0:  Introduction  to  the  BS  6001  attribute  
sampling  system.  Technical  Committee  SS/5.  British  Standard,  1996.  
Lerma,   Norma   V.   and   Marina   O.   Osi.   Drug   and   Cosmetic   Quality   Control   with   Instrumentation,   2nd   ed.   Manila,  
Philippines:  UST  Publishing  House,  1996.  
ISO   2859-­‐1.   Sampling   procedures   for   inspection   by   attributes.   2nd   Edition.   International   Organization   for  
Standardization.  Switzerland.  1999.  
WHO  guidelines  on  quality  assurance  of  pharmaceuticals  :  a  compendium  of  guidelines  and  related  materials.  Vol.  2,  
Good  manufacturing  practices  and  inspection.  –  2nd  ed.  Geneva,  World  Health  Organization,  2007.  
WHO   guidelines   on   quality   control   methods   for   herbal   materials.   Updated   edition   of   quality   control   methods   for  
medicinal  plant  materials,  1998.  Geneva,  World  Health  Organization,  2011.  
Appendix  1.    
Examples  of  types  of  containers  used  to  store  samples  of  starting  materials  and  bulk  products  


                                           Figure  1.  Bag  for  storage  of  samples.                     Figure  2.  Screw-­‐top  containers