EUROMEDLAB 2009 Innsbruck

ISW # 20

In Vitro Hemolysis: Causes, Prevalence, 
Effects, Measurement & Solutions

• Causes, Prevalence & Effects Prof. G. Lippi

• Measurement & Solutions Prof. N. Blanckaert
• Questions & Discussion
• Co‐chairs Dr. S. Green & Prof. V. Palicka

Innsbruck, June 2009

 European Preanalytical
Scientific Committee (EPSC)

Innsbruck, June 2009

 In Vitro Hemolysis: Causes, Prevalence, 
Effects, Measurement & Solutions

Causes, Prevalence & Effects
Prof. G. Lippi
University Hospital
Verona, Italy

Innsbruck, June 2009

 4

In vitro hemolysis: DEFINITION

Hemolysis (or haemolysis), from the Latin hemo
(blood) and lysis (to break open), is the release of
hemoglobin and other intracellular components
from erythrocytes to the surrounding plasma,
following damage or disruption of the cell membrane.

In vitro hemolysis has always been a major concern for clinical

laboratories worldwide, due since it seriously impacts patient care
and a laboratory's reputation through its affect on test results.

Innsbruck, June 2009

 5

In vitro hemolysis: DEFINITION

- The upper reference limit for free hemoglobin
is 2 mg/dL for plasma and 5 mg/dL for serum.
- Visually, hemolysis is defined as free
hemoglobin concentration > 30 - 50 mg/dL,
conferring detectable pink/red hue to serum or plasma.
- It becomes clearly visible in specimens containing as low as
0.5% lysed erythrocytes.

Innsbruck, June 2009

 6

Innsbruck, June 2009

7

In vivo hemolysis: CAUSES

• This clinical condition may have > 50 causes,

including hereditary, acquired and iatrogenic conditions:
• Autoimmune hemolytic anaemia • DIC
• Hemoglobinopathies • Transfusion reactions
• Drugs • Heart valves
• Severe infections • HELLP syndrome.

• Typically, in vivo hemolysis does not depend on the technique

of the healthcare provider and it is thus virtually unavoidable and
potentially insurmountable.
Innsbruck, June 2009
 8

In vitro hemolysis: CAUSES

In vitro hemolysis can occur beginning at the patient’s bedside

and continue through sample handling, processing and storage.
Factors vary depending upon the patient’s condition (fragile
veins), the skill of the person collecting the sample (training),
and the local environment (temperature, length of transport).

Innsbruck, June 2009

 9

Innsbruck, June 2009

 10

In vitro hemolysis:
ROOT CAUSES

Innsbruck, June 2009

 11

Hemolytic specimens are a frequent

occurrence in laboratory practice.
• The prevalence is described as being as
high as 3.3% of all routine samples afferent
to a clinical laboratory.
• They account for 40%–70% of all
unsuitable specimens.
• They are the primary cause of unsuitable
specimens, nearly five times the rate of the
second.
• In vitro hemolysis remains the leading
cause of unsuitable specimens for both
outpatient and inpatient samples, for routine
and stat specimens.

Innsbruck, June 2009

 12

In vitro hemolysis:
PREVALENCE (STAT)
z 505 hemolysed specimens identified (3.3%)
- 64% small degree (50 mg/dL free Hb)
- 31% intermediate degree (50 to 300 mg/dL free Hb)
- 5% by high degree (>300 mg/dL free Hb)
z The % of hemolysed specimens was similar:

- 3.5% intensive care units

- 3.4% dept. of organ transplantation
- 3.3% emergency dept
- 3.1% internal medicine and surgery dept

Innsbruck, June 2009

 13
In vitro hemolysis:
PREVALENCE (Routine)
z The total number of hemolysed specimens
recorded was 8440 (5.6%)

Innsbruck, June 2009

 14
In vitro hemolysis:
INTERFERENCES

Innsbruck, June 2009

 15
In vitro hemolysis:
INTERFERENCES
• Hemolysis interference is approximately linearly dependent on
the final concentration of free Hb in the specimen.
• It generates a consistent trend towards overestimation of:
– ALT & AST
– Creatinine
– Sreatine kinase (CK)
– Iron
– LDH
– Lipase
– Magnesium
– Phosphorus
– Potassium
– Urea
• It generates a consistent trend towards underestimation of:
– Albumin
– ALP
– Chloride & Sodium
– GGT
– Glucose
• Clinically meaningful variations of AST, chloride,
chloride, LDH,
potassium and sodium were observed in specimens with mild
hemolysis (free Hb ~60 mg/dL
mg/dL). ).

Innsbruck, June 2009

 In vitro hemolysis: 16

INTERFERENCES
Although slightly hemolysed specimens might
still be analyzable, gross hemolysis (0.9%,
~150 mg/dL of free Hb), influences the
reliability of routine coagulation testing.

Innsbruck, June 2009

 17

The European Preanalytical Scientific Committee (EPSC), in 

collaboration with the International Federation of Clinical Chemistry
(IFCC) Working Group on Patient Safety, have designed a questionnaire
to collect data on prevalence and management of hemolytic specimens
referred to the clinical laboratories for clinical chemistry testing.
Surveys completed: 391
Innsbruck, June 2009
 18

Countries

Innsbruck, June 2009

 19

Type of facility

Innsbruck, June 2009

 20

Size of the Laboratory

Specimens/year

Innsbruck, June 2009

 21

Are you aware of the problem that some

lab tests might be affected by hemolysis?

Innsbruck, June 2009

 22

Do you systematically monitor the number and

the origin (wards, facilities, etc.) of hemolysed
specimens that you receive in your lab?

Innsbruck, June 2009

 23

What is the percentage (approximately) of

hemolysed specimen you receive?

Innsbruck, June 2009

 24

Where do most hemolysed specimens

come from?

Innsbruck, June 2009

 The IFCC Working Group on laboratory
errors and patient safety
25

Innsbruck, June 2009

 Leading unresolved issues:

• Why there is poor concern and information on this common

problem?
• Why clinical labs are reluctant to share their data on incidence
and sources of hemolysed samples, thus allowing benchmark?
• Why available guidelines aren’t followed by most laboratories
around the world?
• Are guidelines not “evidence-based” or is compliance with
guidelines poor in laboratory practice?
• Do we need mandatory rules?

Innsbruck, June 2009

 In Vitro Hemolysis: Causes, Prevalence, 
Effects, Measurement & Solutions

Mesaurement & Solutions
Prof. N.Blanckaert
University Hospital
Leuven, Belgium

Innsbruck, June 2009

 HEMOLYSIS
• A SIGNIFICANT PROBLEM:
– Frequent occurrence
– Non‐transparent to the requesting physician
– Interferes with:
• analyte level
• analyte measurement
– Considerable impact, as possible source of 
misleading clinical information

• A PROBLEM THAT DESERVES TO BE 
MANAGED

Innsbruck, June 2009

 HEMOLYSIS
• A PROBLEM THAT DESERVES TO 
BE MANAGED
– Risk assessment‐based investigation:
hence, measure hemolysis occurrence and 
impact
– Root Cause Analysis (5 why’s, fishbone 
diagram, …)
– CAPA :
• Corrective actions: systematic & ad hoc
• Preventive actions

Innsbruck, June 2009

 ISO 15189 standards

Innsbruck, June 2009

 Management of the Hemolysis Issue (1)

• Do you have a systematic approach to 
risk assessment re. hemolysis?
– Assess interference in your assays, as part 
of method selection, validation?
– Measure frequency, degree of hemolysis 
for your current assays?
– System approach to deal with 
interference, in your lab operations and 
for reporting?
– Integral part of your quality system?
Innsbruck, June 2009
 Management of the Hemolysis Issue (2)

• Do you have a systematic approach to 
preventive actions?

• Do you have a systematic approach to 
corrective actions? 

Innsbruck, June 2009

 ISO 15189 standards

Innsbruck, June 2009

 SOLUTIONS (1)
• Create awareness: provide 
information, education

– Laboratory management ‐> leadership

– Phlebotomists

– Requesting physicians

Innsbruck, June 2009

 SOLUTIONS (2)
• Prevention
– Establish proper phlebotomy techniques
– Phlebotomists: education & training, feed‐
back on performance
– Assessment of transport procedures, systems
– Select and validate proper in‐lab 
preanalytical procedures
– Select analytical methods with minimal 
interference

Innsbruck, June 2009

 SOLUTIONS (3)
• Corrective actions

– Continuous assessment, focused actions on 
basis of risk estimation

– Procedures and mechanisms for dealing with 
individual specimens

Innsbruck, June 2009

 Continuous Monitoring (1)
• Monitoring system must be outcome
oriented
• Components of monitoring system:
– Document the hemolysis interference 
profiles of your assays ‐> estimate of impact
– Establish a sound hemolysis measurement 
system
– Monitor your sampling device and 
technique, transportation, preanalytical 
handling systems
Innsbruck, June 2009
 Continuous Monitoring (2)
– Establish procedures to monitor and 
communicate impact of hemolysis on 
clinical information value of your laboratory 
results:

• eg. how often K results are compromised, non‐
reportable (delayed)?

• eg. how often LDH results are compromised, 
non‐reportable (delayed)?

Innsbruck, June 2009

 Continuous Monitoring (3)
• Use impact results
as Key Performance Indicator (KPI)

– Communicate KPIs to key stakeholders

– Use focused approach: per facility, per ward, 
in‐patients versus out‐patients

– Use KPIs as tool in improvement projects, as 
part of PDCA cycle

Innsbruck, June 2009

 Continuous Monitoring (4)

Innsbruck, June 2009

 SOLUTIONS
• Corrective actions

– Continuous assessment, focused actions on 
basis of risk estimation

– Procedures and mechanisms for dealing with 
individual specimens

Innsbruck, June 2009

42
How do you deal with
hemolysed specimens?
A. Perform all requested tests but do
not report the test results mostly
affected by hemolysis
B. Reject the specimen, refuse to
perform requested tests that are
affected by hemolysis

C. Perform all requested tests but do not

report the test results mostly affected
by hemolysis and ask for an additional
sample
D. Correct results for the hemolysis index
when feasible and provide an
associated interpretative comment (i.e.
"test performed on a hemolysed
specimen")

Innsbruck, June 2009

 How do you deal with individual
hemolysed specimens?
• Need for systematic hemolysis 
assessment, with differential 
approach:

– depending on requested tests

– depending on hemolysis interference 
profile

Innsbruck, June 2009

 How do you deal with individual
hemolysed specimens?
• Report hemolysis impact, not just 
degree of hemolysis!
– if relatively small interference and large 
Total Error budget, report with comment
– if major interference and/or small Total 
Error budget, do not report, request new 
sample
– mathematical correction for hemolysis 
interference is not recommended

Innsbruck, June 2009

 How do you deal with individual
hemolysed specimens?
• For non‐reportable tests, 
communication mode depends on 
potential criticality of the test 
information:

– rule out in vivo hemolysis

– medical emergency request?

– tests with ‘critical’ value intervals

Innsbruck, June 2009

 Hemolysis Assessment: Requirements

• Objective measurement as continuous 
variable is ideal, semi‐quantitative 
assessment is sufficient if enough categories 
can be distinguished
• Assessment must be made selectively, 
depending on requested tests
• Assessment must be done as soon as 
possible following sample collection and 
centrifugation, to minimize the effect on 
TAT

Innsbruck, June 2009

47

How do you check to evaluate sample

hemolysis in your lab?

Innsbruck, June 2009

 Hemolysis Assessment: Methods

• Visual/ camera assessment
Digital image analysis Visual inspection
e.g. PVT’s QSI camera system Use control chart

• Spectrophotometric measurement
‐ Categorical scale: e.g. Olympus AU5400, Siemens Vista,
Siemens Dimension RxL
‐ Continuous scale: e.g. Roche Modular&Cobas systems,
OCD Vitros 5,1 FS

Innsbruck, June 2009

 Measurement-Roche Modular

• Bichromatic wavelength  • Absorbance readings 

pairs: used:
– 480 nm and 505nm (range 1) – hemolysis index: 
ranges 2 and 3
– 570 nm and 600 nm (range 2)
– icterus index: ranges 
– 660 nm and 700 nm (range 3)
1,2 and 3
– lipemia index: range 3

Innsbruck, June 2009

 Measurement-Roche Modular

The formula to calculate the hemolysis index (H) is:

where:
A = scaling factor for hemoglobin
B = corrects H measurement for lipemia
ΔAbs2,3 = absorbance of the 570 ‐ 600 nm and 660 ‐
700 nm bichromatic readings, 
respectively, in relation to the blank absorbances.

Innsbruck, June 2009

 Measurement-Olympus

• Patient samples are diluted with the LIH 
reagent and the absorbance is measured at 6 
wavelengths. If one or more chromogen in a 
potentially interfering concentration is present 
in a sample, flags will be generated and 
reported along with the results of the analyses 
performed on that sample.
These flags characterise
‐ the kind of chromogen (L, I, or H)
‐ the approximate concentration of the
interferent, as a flag category: e.g. +, ++, ++++

Innsbruck, June 2009

 Measurement-Olympus

Innsbruck, June 2009

 Comparative study for measurement
Modular Integra Advia 1800 & 2400
Drabkin (3 labs, n=8) (1 lab, n=3) (1 lab, n=6)
(mg/dL)
Mean Range Mean Range Mean Range
8.5 10 9 ‐ 11 12 11 ‐ 13 N (13) 12 ‐ 14
48 46 43 ‐ 50 48 45 ‐ 49 + (68) 65 ‐ 72
75 74 70 ‐ 80 74 73 ‐ 75 ++ (109) 103 ‐ 114
119 119 112 ‐ 127 118 116 ‐ 119 ++ (174) 165 ‐ 184
217 217 206 ‐ 231 213 212 ‐ 214 +++ (319) 302 ‐ 335

Olympus Beckman Siemens Dimension 

Drabkin (1 lab, n=3) (1 lab, n=3) RxL (2 labs, n=5)
(mg/dL)
Range  Range  Range 
Mean Mean Mean
defined defined defined
8.5 N <50 0 0 1 0 - 25
48 N <50 1 0‐50 2 26 - 50
75 ʺ+ʺ 50‐99 2 50‐100 3 51 - 200
119 ʺ++ʺ 100‐199 3 100‐150 3 51 - 200
217 ʺ+++ʺ 200‐299 6 250‐300 4 202 - 300

Innsbruck, June 2009

Innsbruck, June 2009
 In Vitro Hemolysis: Causes, Effects, 
Prevalence, Measurement & Solutions

Questions & Discussion
Prof. V. Palicka
Charles’ University, University Hospital,
Hradec Králové, Czech Republic

Innsbruck, June 2009