Scand J Clin Lab Invest 2004; 64: 175 – 184

The reference change value: a proposal to

interpret laboratory reports in serial testing
based on biological variation
C. RIC Ó S, * * F . CA V A , * { J . V. GARCÍ A- LA RI O , * { A . HE RN Á NDEZ, * §
N. I G LE SIA S, * C. V . J IM É NEZ, * } J . M IN C H I NE L A , * } C. P ERIC H , * ,
M. SIMÓ N , * * * M . V . D OM EN EC H §§ & V . Á L V A R E Z * { {
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

*Laboratoris Clı́nics Hospital Vall d’Hebron, Barcelona; {Area de Laboratorio, Fundación

Hospital Alcorcón, Madrid; {Laboratorio de Análisis Clı́nicos, Hospital de Motril, Granada;
§Laboratori Clı́nic de L’Hospitalet, Barcelona; }Laboratori Clı́nic del Barcelonès Nord,
Barcelona; ,Laboratori Clı́nic Bon Pastor, Barcelona; **Laboratori Clı́nic Intercomarcal
Vilafranca, Barcelona; {{Laboratori Clı́nic de Cornellá de Llobregat, Barcelona, Spain;
§§Laboratori Clı́nic Manso, Barcelona

Ricós C, Cava F, Garcı́a-Lario JV, Hernández A, Iglesias N, Jiménez CV,

Minchinela J, Perich C, Simón M, Domenech MV, Álvarez V. The reference
change value: a proposal to interpret laboratory reports in serial testing based on
biological variation. Scand J Clin Lab Invest 2004; 64: 175–184.
For personal use only.

Background: A proposal to calculate and use the reference change value (RCV)
as an objective guide for interpreting the numerical results obtained in clinical
laboratory serial testing is introduced in this study. Methods: A database
showing the results of a compilation of 191 publications on biological variation
and including information on a number of analytes provided the standardized
criterion based on biology for calculating the RCVs. Results: For each of the 261
analytes included in the study, the RCV was determined using Harris’s formula,
replacing analytical imprecision with the desirable specification of analytical
quality based on half the within-subject biological variation at 95% probability
levels. The result is a guide for a common criterion to identify clinically
significant changes in serial results. Conclusions: The RCV concept is an
approach that can be offered by laboratories to assess changes in serial results.
The RCV data in this study are presented as a point of departure for a widely
applicable objective guide to interpret changes in serial results.

Key words: Biological variation; quality assurance; reference change value

Carmen Ricós, Unitat de la Qualitat, Laboratoris Clı́ncis Hospital Universitari

Vall d’Hebron, Passeig Vall d’Hebron 119, ES-08036, Barcelona, Spain. E-mail.
[email protected]

IN T R O D U C T I O N such as defining quality specifications, designing

quality control rules, checking for errors (delta
Data on biological variation (BV) are used in checks) using consecutive results [1, 2], and to
the medical laboratory for various purposes provide information regarding patient health
status. This study deals with the last application.
*Members of the Committee for Quality Assurance Laboratory tests provide essential informa-
and Laboratory Accreditation of the Spanish Society
for Clinical Biochemistry and Molecular Pathology tion for the medical decision process, but they
(SEQC). are subject to intrinsic variability, which affects
DOI 10.1080/00365510410004885 175
 176 C. Ricós et al.

the final results generated. In general terms, be minimized by providing standard instruc-
proper interpretation of laboratory results tions to patients before taking samples and by
requires the following: using written protocols for sample collection,
transport, handling, centrifugation, etc. Analy-
. Knowledge of the type and magnitude of tical variation can be reduced by carefully
variation that is included in the result. maintaining equipment, by properly training
. Use of an approach that allows comparison personnel and by using written, standardized
of the result with criteria that define a state of procedures.
health. There are several general approaches to
Several components of variation affect labora- interpretation of laboratory results, e.g. com-
tory results. Some relate to laboratory activity parison with cut-off values, comparison with
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

(preanalytical and analytical variation) and population reference values, or comparison

others relate to the normal human biological with previous results from the same indivi-
variation in the levels of analytic constituents. dual. The conventional reference interval
Biological variation has two components: within- approach is the most commonly used aid to
subject and between-subject variation. Within- interpret laboratory results, but it has limita-
tions. Marked ‘‘individuality’’ of the analyte
subject or intra-individual BV is the random
explains some of the difficulties in laboratory
fluctuation of a human body constituent around
test interpretation with this method. The
the homeostatic setting point. Additional
variations in the levels of an analyte in
sources of intra-individual variation include
most individuals over time are smaller than
the influences of the natural ageing process,
the dispersion of the reference interval, and
gender, weight, diet, exercise, daily seasonal
intra-individual BV is generally less than inter-
For personal use only.

rhythms and, of course, the alterations caused

individual BV. The ratio of intra-individual
by pathological processes. The individual
to inter-individual BV is called the index of
homeostatic setting points also vary, and this individuality (II). The lower the II, the greater
variation is known as between-subject or inter- is the inherent individuality of the analyte
individual variation [3, 4]. Biological varia- tion tested. The test results of a person can be
is generally expressed in terms of coefficients of outside of his/her healthy range but still lie
variation (CVI and CVG for intra-individual within the population-based reference interval.
and inter-individual variation, respectively). A In the case of two consecutive results, both
great deal of effort has been dedicated to esti- may be within the reference interval but the
mating the biological variation of the constitu- difference between them may be clinically
ents analysed in clinical laboratories and a significant for that particular person. In
recent compilation of this data is now available general terms, the usefulness of population
for a large number of quantities [5 – 8]. reference intervals for monitoring patients is
Analytical and biological variations are limited when the II is less than 0.6 and
random, can be considered Gaussian and are acceptable when the II is greater than 1.4
cumulative. The total variance associated with a [3, 9 – 11].
laboratory result (s2T) is expressed by the When several consecutive results for one
following formula: person are available (serial testing), there is
S2T ~S2P zS2A zS2I another interesting approach for interpretation
of laboratory results. In serial testing, results
with s2P being the preanalytical, s2A the ana- for a specific analyte are evaluated over time in
lytical and s2I the intra-individual variation. a single individual. To confidently report a
Expressed in terms of coefficient of variation, significant change in serial results, the difference
the formula is as follows: in these results must exceed the difference

1=2 explained by the inherent variation. Thus,
CVT ~ CV2P zCV2A zCV2I
assuming that some CVP is essentially included
The lower the total variation in a test result, the in the CVI and assuming that it can be
easier it is to distinguish small but clinically minimized through adherence to good written
important changes that guide medical decision- standard operating procedures, good training
making. Preanalytical sources of variation can and good laboratory practices, the total
 The RCV proposal 177

variation for each result in a single sample components of variation are given for 267
analysed once is: analytes in the healthy state and several patho-

1=2 logic situations. Additionally, other reported
CVT ~ CV2A zCV2I information, e.g. number of subjects, certain
Because the variation involved is random and conditions (age, gender, fasting, state of health),
symmetrically distributed, we can with a certain study period, sampling time, number of samples
probability derive the interval within which the per subject, and analysis results (mean values,
values of variation will lie by using standard analytical coefficient of variation) are provided.
normal deviates (Z-scores) [4, 11]. To conclude The present work only deals with the data from
that the difference between two serial results is healthy subjects.
significant and could be biologically relevant For each analyte studied, the CVI values
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

(change in health status), the difference must be compiled were listed in ascending order to
greater than the sum of the variation of each clearly identify extreme values. These values
result. This difference is called the critical were segregated and carefully studied to deter-
difference or the reference change value. This mine whether there was a reasonable basis for
latter term emphasizes that the interpretation is eliminating them from the calculations. Exam-
based on the difference in a result from a single ples of excluded data were values corresponding
individual with respect to his/her previous to non-fasting subjects for calculating CVI and
results, instead of on population-based values. CVG for glucose and triglycerides. The median
The reference change value (RCV) is expressed of the remaining values was then calculated.
by the following formula [3, 4, 11, 12]: For each analyte included, the reference

1=2 change value was calculated using the above-
RCV~ZP
CV2A zCV2I z CV2A zCV2 I


mentioned formula [1, 3], replacing the analy-
For personal use only.

1=2 tical imprecision with the desirable specification

RCV~ð2Þ1=2
ZP
CV2A zCV2I
of analytical quality based on biological varia-
in which RCV is the reference change value, ZP tion (0.5 CVI) [10 – 13].
is the standard deviation for the appropriate  1=2
probability of error, and CVA and CVI are the RCV~ZP ð2Þ1=2 CV2I zð0:5
CVi Þ2
analytical and intra-individual biological coeffi-
When the range of probability of change is
cients of variation.
bidirectional and fixed at 95%, ZP is equal to
In this paper we introduce a proposal to
1.96.
calculate and use the RCV as an objective,
flexible guide for identification of significant
differences between serial results that could RESULTS
have clinical relevance. Included in the formula
are the following: Data for the 261 analytes comprising the
. Intra-individual variation. Values are taken database are presented in Table I. This table
from an extensive compilation of data on includes CVI data, desirable specifications for
components of biological variation. CVA, the calculated RCV values for these
. Analytical variation. Desirable specifications specifications and 95% probability levels, using
(goals) proposed for analytical variation are bidirectional Z-scores.
based on biological variation.
. Z-scores at two levels of probability (95% and
DISCUSSION
99%) for bidirectional situations.
Laboratory test results are often used for
MATERIALS AND METHODS monitoring patients, which involves reviewing
these results over time. When analytes are strongly
To perform this study we used a database [5 – 8] regulated, the intra-individual variation is
showing the results of a compilation of 191 publi- narrower than the inter-individual variation
cations on biological variation and including (the ratio CVI/CVG is small) and comparison
information for 316 body fluid constituents. of serial results with the population-based refer-
Estimates of intra-individual and inter-individual ence range does not provide useful information
 178 C. Ricós et al.

TABLE I. Proposal for reference change values (RCVs) according to biological variation, analytical quality
specifications and probability level using bidirectional Z-scores.

RCV Desirable
Biological specification
variation (CVA~0.5 CVI)
Analyte CVI CVA desirable RCV95%

S- 11-Deoxycortisol 21.3 10.7 66.0

S- 17-Hydroxiprogesterone 19.6 9.8 60.7
S- 5’Nucleotidase 23.2 11.6 71.9
U- 5-HIAA concentration, 24 h 20.3 10.2 62.9
S- a1-Acid glycoprotein 11.3 5.7 35.0
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

S- a1-Antichymotripsine 13.5 6.8 41.8

S- a1-Antitrypsin 5.9 3.0 18.3
S- a1-Globulins 11.4 5.7 35.3
U- a1-Microglobulin concentration, overnight 33.0 16.5 102.3
P- a2-Antiplasmin 6.2 3.1 19.2
S- a2-Globulins 10.3 5.2 31.9
S- a2-Macroglobulin 3.4 1.7 10.5
U- a2-Microglobulin output, overnight 29.0 14.5 89.9
S- a-Amylase 9.5 4.8 29.4
U- a-Amylase concentration, random 94.0 47.0 291.3
S- a-Amylase, pancreatic 11.7 5.9 36.3
S- a-Carotene 35.8 17.9 110.9
S- Acid phosp, tartrate-resistant (TR-ACP) 8.0 4.0 24.8
S- Acid phosphatase (ACP) 8.9 4.5 27.6
S- Acid phosphatase activity, prostatic (PAP) 33.8 16.9 104.7
For personal use only.

P- Activate partial thromboplastine, time 2.7 1.4 8.4

S- Adenosine deaminase (ADA) 11.7 5.9 36.3
S- Alanine aminopeptidase 4.1 2.1 12.7
S- Alanine aminotransferase 24.3 12.2 75.3
S- Albumin 3.1 1.6 9.6
U- Albumin concentration, first morning 36.0 18.0 111.6
S- Aldosterone 29.4 14.7 91.1
U- Aldosterone concentration, 24 h 32.6 16.3 101.0
S- Alkaline phosphatase 6.4 3.2 19.8
S- Alkaline phosphatase, bone isoform 6.2 3.1 19.2
S- Alkaline phosphatase, placental 19.1 9.6 59.2
U- Ammonia output, 24 h 24.7 12.4 76.5
S- Androstendione 11.5 5.8 35.6
S- Angiotensin converting enzyme (ACE) 12.5 6.3 38.7
P- Antithrombin III 5.2 2.6 16.1
S- Apolipoprotein A1 6.5 3.3 20.1
S- Apolipoprotein B 6.9 3.5 21.4
S- Ascorbic acid 25.8 12.9 80.0
S- Aspartate aminotransferase 11.9 6.0 36.9
S- a-Tocopherol 13.8 6.9 42.8
S- b2-Microglobulin 5.9 3.0 18.3
B- Basophils, count 28.0 14.0 86.8
S- b-Carotene 36.0 18.0 111.6
S- b-Cryptoxanthin 36.7 18.4 113.7
S- b-Globulins 10.1 5.1 31.3
S- Bilirubin, conjugated 36.8 18.4 114.0
S- Bilirubin, total 25.6 12.8 79.3
S- C Peptide 9.3 4.7 28.8
S- C Telopeptide type I collagen 8.0 4.0 24.8
U- C Telopeptide type I collagen/Creat 35.1 17.6 108.8
S- C3 Complement 5.2 2.6 16.1
S- C4 Complement 8.9 4.5 27.6
S- CA 125 antigen 29.2 14.6 90.5
S- CA 15.3 antigen 5.7 2.9 17.7
S- CA 19.9 antigen 24.5 12.3 75.9
 The RCV proposal 179

TABLE I. (Continued ).

RCV Desirable
Biological specification
variation (CVA~0.5 CVI)
Analyte CVI CVA desirable RCV95%

S- CA 549 antigen 9.1 4.6 28.2

S- Calcium 1.9 1.0 5.9
U- Calcium concentration, 24 h 27.5 13.8 85.2
S- Calcium ionized 1.7 0.9 5.3
U- Calcium output, 24 h 26.2 13.1 81.2
S- Carbohydrate deficient transferrin 7.1 3.6 22.0
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

S- Carcinoembryonic antigen (CEA) 12.7 6.4 39.4

S- Ceruloplasmin 5.7 2.9 17.7
S- Chloride 1.2 0.6 3.7
S- Cholesterol 6.0 3.0 18.6
S- Cholinesterase 7.0 3.5 21.7
S- Cholinesterase, immunoreactive 6.4 3.2 19.8
S- Cholinesterase, activity 5.4 2.7 16.7
S- CK MB% 6.9 3.5 21.4
S- CK MB, activity 19.7 9.9 61.1
S- CK MB, mass 18.4 9.2 57.0
S- Copper 4.9 2.5 15.2
P- Copper 8.0 4.0 24.8
S- Cortisol 20.9 10.5 64.8
S- C-Propeptide type 1 procollagen 8.2 4.1 25.4
S- C-Reactive protein 52.6 26.3 163.0
For personal use only.

S- Creatine kinase (CK) 22.8 11.4 70.7

S- Creatinine 4.3 2.2 13.3
Patient- Creatinine clearance 13.6 6.8 42.1
U- Creatinine concentration, 24 h 24.0 12.0 74.4
U- Creatinine output, 24 h 11.0 5.5 34.1
U- C-Telopeptide type I collagen/creat., 2nd morning 23.5 11.8 72.8
P- Cysteine 5.9 3.0 18.3
S- Dehydroepiandrosterone sulphate 11.6 5.8 35.9
U- Deoxipyridinoline/Creatinine, 24 h 15.4 7.7 47.7
U- Deoxipyridinoline/Creatinine, morning spot 26.5 13.3 82.1
P- Dipeptidyl-peptidase IV 8.2 4.1 25.4
P- Elastase 13.6 6.8 42.1
B- Eosinophils, count 21.0 10.5 65.1
(B)Plat- Epinephrine 25.3 12.7 78.4
P- Epinephrine 48.3 24.2 149.7
B- Erythrocytes, count 3.2 1.6 9.9
U- Oestradiol 30.4 15.2 94.2
S- Oestradiol 18.1 9.1 56.1
P- Factor V coagulation 3.6 1.8 11.2
P- Factor VII coagulation 6.8 3.4 21.1
P- Factor VIII coagulation 4.8 2.4 14.9
P- Factor X coagulation 5.9 3.0 18.3
S- Ferritin 14.9 7.5 46.2
P- Fibrinogen 10.7 5.4 33.2
(B)Erythr- Folate 12.0 6.0 37.2
S- Folate 24.0 12.0 74.4
S- Follicle stimulating hormone (males) 8.7 4.4 27.0
U- Free oestradiol 38.6 19.3 119.6
S- Free testosterone 9.3 4.7 28.8
U- Free testosterone 51.7 25.9 160.2
S- Free thyroxine (FT4) 7.6 3.8 23.6
S- Free triiodothyronine (FT3) 7.9 4.0 24.5
S- Fructosamine 3.4 1.7 10.5
S- Galactosy hydroxylisine 11.8 5.9 36.6
S- c-Globulins 14.6 7.3 45.2
 180 C. Ricós et al.

TABLE I. (Continued ).

RCV Desirable
Biological specification
variation (CVA~0.5 CVI)
Analyte CVI CVA desirable RCV95%

S- c-Glutamyltransferase 13.8 6.9 42.8

S- Globulins, total 5.5 2.8 17.0
S- Glucose 4.9 2.5 15.2
(B)Erythr- Glucose 6 phosphate dehydrogenase (G6PDH) 32.8 16.4 101.6
B- Glutathion peroxidase 7.2 3.6 22.3
S- Glycated albumin 5.2 2.6 16.1
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

S- Glycated total protein 0.9 0.5 2.8

P,S- Haptoglobin 20.4 10.2 63.2
S- HDL cholesterol 7.1 3.6 22.0
S- HDL1 cholesterol 5.5 2.8 17.0
S- HDL2 cholesterol 15.7 7.9 48.7
S- HDL3 cholesterol 7.0 3.5 21.7
B- Hematocrit 2.8 1.4 8.7
B- Haemoglobin 2.8 1.4 8.7
S- Haemoglobin A1 C 2.1 1.1 6.5
P- Homocysteine 9.0 4.5 27.9
U- Hydroxiproline/creatinine 25.9 13.0 80.3
U- Hydroxiproline/minute, night urine 36.1 18.1 111.9
S- Hydroxybutyrate dehydrogenase 8.8 4.4 27.3
S- Immunoglobulin A 5.4 2.7 16.7
S- Immunoglobulin G 4.5 2.3 13.9
For personal use only.

S- Immunoglobulin M 5.9 3.0 18.3

S- Immunoglobulins k chains 4.8 2.4 14.9
S- Immunoglobulins l chains 4.8 2.4 14.9
S- Insulin 21.1 10.6 65.4
(B)Leuc- Interferon receptor 14.0 7.0 43.4
S- Interleukin-1b 30.0 15.0 93.0
S- Interleukin-8 24.0 12.0 74.4
S- Iron 26.5 13.3 82.1
B- Lactate 27.2 13.6 84.3
S- Lactate dehydrogenase (LDH) 8.6 4.3 26.7
S- Lactate dehydrogenase 1 isoform (LD1) 6.3 3.2 19.5
S- Lactate dehydrogenase 2 isoform (LD2) 4.9 2.5 15.2
S- Lactate dehydrogenase 3 isoform (LD3) 4.8 2.4 14.9
S- Lactate dehydrogenase 4 isoform (LD4) 9.4 4.7 29.1
S- Lactate dehydrogenase 5 isoform (LD5) 12.4 6.2 38.4
P- Lactoferrin 11.8 5.9 36.6
S- LDL cholesterol 8.3 4.2 25.7
S- LDL cholesterol direct 6.5 3.3 20.1
B- LDL receptor mRNA 21.5 10.8 66.6
B- Leucocytes, count 10.9 5.5 33.8
S- Lipase 23.1 11.6 71.6
S- Lipoprotein (a) 8.5 4.3 26.3
S- Lutein 23.7 11.9 73.4
S- Luteinizing hormone 14.5 7.3 44.9
S- Lycopenen 43.1 21.6 133.6
B- Lymphocytes, count 10.4 5.2 32.2
(B)Erythr- Magnesium 5.6 2.8 17.4
(B)Leuc- Magnesium 18.3 9.2 56.7
S- Magnesium 3.6 1.8 11.2
U- Magnesium concentration, 24 h 45.4 22.7 140.7
S- Magnesium ionized 1.9 1.0 5.9
U- Magnesium output, 24 h 38.3 19.2 118.7
(B)Erythr- Mean corpuscular haemoglobin (HCM) 1.6 0.8 5.0
(B)Erythr Mean corpuscular haemoglobin conc (MCHC) 1.7 0.9 5.3
(B)Erythr- Mean corpuscular volume (MCV) 1.3 0.7 4.0
 The RCV proposal 181

TABLE I. (Continued ).

RCV Desirable
Biological specification
variation (CVA~0.5 CVI)
Analyte CVI CVA desirable RCV95%

(B)Plat- Mean platelet volume (MPV) 4.3 2.2 13.3

B- Monocytes, count 17.8 8.9 55.2
S- Mucinous carcinoma-associated antigen (MCA) 10.1 5.1 31.3
S- Myoglobin 13.9 7.0 43.1
U- N-acetyl glucosaminidase concentration,overnight 48.6 24.3 150.6
B- Neutrophils, count 16.1 8.1 49.9
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

U- Nitrogen, output 13.9 7.0 43.1

(B)Plat- Norepinephrine 9.5 4.8 29.4
P- Norepinephrine 19.5 9.8 60.4
U- N-telopeptide type I collagen/creatinine 20.5 10.3 63.5
S- Osmolality 1.3 0.7 4.0
S- Osteocalcin (z1 trab) 7.2 3.6 22.3
U- Oxalate concentration, 24 h 44.0 22.0 136.4
U- Oxalate output, 24 h 42.5 21.3 131.7
B- pCO2 4.8 2.4 14.9
B- pH (pH units) 0.2 0.1 0.6
B- pH [Hz] 3.5 1.8 10.8
S- Phosphate 8.5 4.3 26.3
U- Phosphate concentration, 24 h 26.4 13.2 81.8
U- Phosphate output, 24 h 18.0 9.0 55.8
Patient- Phosphate tubular reabsorption 2.7 1.4 8.4
For personal use only.

S- Phospholipids 6.5 3.3 20.1

P- Plasminogen 7.7 3.9 23.9
B- Platelet distribution wide (PDW) 2.8 1.4 8.7
B- Plateletcrit 11.9 6.0 36.9
B- Platelets 9.1 4.6 28.2
S- Potassium 4.8 2.4 14.9
(B)Leuc- Potassium 13.6 6.8 42.1
U- Potassium concentration, 24 h 27.1 13.6 84.0
U- Potassium output, 24 h 24.4 12.2 75.6
S- Prealbumin 10.9 5.5 33.8
S- Procollagen type I C-terminal 7.8 3.9 24.2
S- Procollagen type I N-terminal 6.8 3.4 21.1
S- Prolactin (men) 6.9 3.5 21.4
P- Prolyl endopeptidase 16.8 8.4 52.1
S- Prostatic specific antigen (PSA) 14.0 7.0 43.4
P- Protein C 5.8 2.9 18.0
U- Protein concentration, 24 h 39.6 19.8 122.7
U- Protein output, 24 h 35.5 17.8 110.0
P- Protein S 5.8 2.9 18.0
S- Protein, total 2.7 1.4 8.4
P- Prothrombin, time 4.0 2.0 12.4
U- Pyridinoline/creatinine, morning spot 8.7 4.4 27.0
B- Pyruvate 15.2 7.6 47.1
B- Red cell distribution wide (RDW) 3.5 1.8 10.8
S- Reticulocyte highly fluorescent, count 10.0 5.0 31.0
S- Reticulocyte low fluorescent, count 1.6 0.8 5.0
S- Reticulocyte medium fluorescent, count 13.0 6.5 40.3
S- Reticulocyte, count 11.0 5.5 34.1
S- Retinol 14.8 7.4 45.9
S- Rheumatoid factor 8.5 4.3 26.3
S- Riboflavin 5.2 2.6 16.1
S- SCC antigen 39.4 19.7 122.1
B- Selenium 12.0 6.0 37.2
P- Selenium 12.0 6.0 37.2
S- Sex hormone binding globulin (SHBG) 12.1 6.1 37.5
 182 C. Ricós et al.

TABLE I. (Continued ).

RCV Desirable
Biological specification
variation (CVA~0.5 CVI)
Analyte CVI CVA desirable RCV95%

S- Sodium 0.7 0.4 2.2

(B)Erythr- Sodium 1.8 0.9 5.6
(B)Leuc- Sodium 51.0 25.5 158.1
U- Sodium concentration, 24 h 24.0 12.0 74.4
U- Sodium output, 24 h 28.7 14.4 88.9
S- Superoxide dismutase 17.1 8.6 53.0
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

(B)Erythr- Superoxide dismutase 12.3 6.2 38.1

S- T3-uptake 4.5 2.3 13.9
Saliva- Testosterone 17.3 8.7 53.6
U- Testosterone 25.0 12.5 77.5
S- Testosterone 9.3 4.7 28.8
S- Thyroglobulin 13.0 6.5 40.3
S- Thyroid stimulating hormone (TSH) 19.7 9.9 61.1
S- Thyroxin binding globulin (TBG) 6.0 3.0 18.6
S- Thyroxine (T4) 6.0 3.0 18.6
S- Tissue polipeptide specific antigen (TPS) 36.1 18.1 111.9
S- Tissue polypeptide antigen (TPA) 28.7 14.4 88.9
U- Total catecholamines, concentration, 24 h 24.0 12.0 74.4
S- Transferrin 3.0 1.5 9.3
S- Triglyceride 20.9 10.5 64.8
S- Triiodothyronine (T3) 8.7 4.4 27.0
For personal use only.

S- Tumor necrosis factor-a 43.0 21.5 133.3

S- Urate 8.6 4.3 26.7
U- Urate concentration, 24 h 24.7 12.4 76.5
U- Urate output, 24 h 18.5 9.3 57.3
S- Urea 12.3 6.2 38.1
U- Urea concentration, 24 h 22.7 11.4 70.3
U- Urea output, 24 h 17.4 8.7 53.9
U- Vanilmandelic acid concentration, 24 h 22.2 11.1 68.8
(B)Erythr Vitamin B12 5.2 2.6 16.1
(B)Erythr Vitamin B6 1.4 0.7 4.3
S- VLDL cholesterol 27.6 13.8 85.5
P- Von Willebrand factor 0.001 0.0 0.0
S- Water 3.1 1.6 9.6
S- Zeaxanthine 34.7 17.4 107.5
S- Zinc 9.3 4.7 28.8
P- Zinc 11.0 5.5 34.1

on the patient’s health status [1, 2]. This is true et al. [15] realized that doctors’ opinions are
for many common analytes [5 – 8]; thus other highly influenced by the analytic state of the art
criteria are needed to assess the patient’s status and that the real variations caused by biological
during serial testing. variation are often ignored.
One approach that can be usesd to interpret It would be of help to the clinician if the
changes in serial results is based on medical laboratory could provide an objective means for
criteria. This information is usually obtained interpreting the information contained in analyti-
through a system of inquiries directed to clini- cal reports. One approach towards achieving
cians to determine at what magnitude a differ- this is through the RCV concept. We introduce
ence in serial results indicates a clinical change in a proposal to calculate and use the RCV according
patient status. This system has shown important to the standardized criterion based on biology.
deficiencies and has been recently revised by The general expression for the RCV is:
Sandberg & Thue [14] with a more restricted
1=2
focus related to specific clinical situations. Stöckl RCV~ð2Þ1=2
ZP
CV2A zCV2I
 The RCV proposal 183

All the components of this formula should be standards are easy to meet, an optimum goal
closely examined to justify the proposal as a (CVA~0.25 CVI) has been established [16, 17].
standardized criterion. When these goals are applied to RCV data,
we obtain a guide for desirable, minimum and
Intra-individual variation optimum RCV values. When a laboratory’s
variation (precision) is lower than the desirable
The CVI should not be interpreted as an or optimum goal, the laboratory is able to
estimate of a true value, but is actually a mean establish that small numerical changes in con-
of obviously different intra-individual values. secutive results are significant, thereby provid-
The values for intra-individual variation in this ing higher clinical sensitivity. However, when a
proposal are also the mean of all published laboratory’s variation does not meet the desir-
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

values in healthy subjects for each analyte, able or even the minimum goal, the numerical
compiled in our database. Nevertheless, there difference in consecutive results will have to be
are some theoretical disadvantages to the use of larger for the laboratory to mark the difference
BV data from healthy subjects for estimating as significant, and it will be difficult to detect
RCVs. Intra-individual variation in acute changes in consecutive results that are biologi-
pathologic states may be larger than that seen cally relevant.
in healthy individuals. For this reason the use of At present, most clinical laboratories make
RCV data derived from healthy individuals use of the Laboratory Information Systems
may lead to false-positive detection of changes (LIS). One of the tasks of the LIS is to help
in patients. Nevertheless, it is commonly consi- in the interpretation of laboratory results through
dered better to register a change and then comparison with reference values, cut-off points,
decide that it is unimportant clinically, than to etc. Since many laboratory results are used for
For personal use only.

miss a change in a patient’s condition. monitoring, it would be useful for all LIS to have
the capability to introduce RCV data and provide
flags or information in laboratory reports on the
Analytical variation
significance of changes in an individual’s serial
The value assigned to this component of the results. Each laboratory would adjust the RCV
formula may be different for each laboratory values according to its own CVA and the interest
since it is established by internal quality control in specific probability ranges.
protocols. Consequently, among laboratories We present the RCV data in this work as
acting independently, the RCV obtained by a point of departure, an objective guide to
applying the formula will also vary. Never- interpret changes in serial results that is widely
theless, to promote harmony in laboratory applicable. This guide is also intended to make
medicine, which is the aim of our scientific laboratory professionals more aware of the
society, we support the use of unified criteria for repercussions of their performance on medical
decision-making. When laboratories maintain decisions.
their analytical CV within the limits derived
from biological variation, there will be a signi-
ficant advance along this line. Then the A C K NO W L E D G E M E NT S
reference change concept will only depend on
intra-individual CVs and the resultant values We thank Celine Cavallo for English
will be common to all laboratories for each language advice.
analyte tested.
In this proposal, we use the quality specifica-
tions based on biology for the CVA in the R E F E RE N C E S
formula. Desirable performance is defined by
CVA~0.5 CVI. This means that intrinsic test 1 Ricós C, Andreu A, Schwartz S. Métodos de
result variability due to biology has been detección del error extraanalı́tico. Rev Diag Biol
increased by 11.8%. For analytes in which 1988; 37: 269 – 71.
2 Fraser CG, Stevenson HP, Kennedy IMG. Biolo-
desirable standards are difficult to meet, a gical variation data are necessary prerequisites for
minimum goal (CVA~0.75 CVI) has been objective autoverification of clinical laboratory
established, and for analytes in which desirable data. Accreditation Qual Assur 2002; 7: 445 – 60.
 184 C. Ricós et al.

3 Fraser CG, Harris EK. Generation and applica- principles to practice. Washington: AACC Press;
tion of data on biological variation in clinical 2001. p. 91 – 116.
chemistry. Crit Rev Clin Lab Sci 1989; 27: 409 – 11 Harris EK, Boyd JC. Statistical bases of reference
37. values in laboratory medicine. New York: Marcel
4 Fraser CG. Change in serial results. In: Biological Dekker Inc; 1995.
variation: from principles to practice. Washington: 12 Harris EK, Yasaka T. On the calculation of a
AACC Press; 2001. p. 67 – 90. ‘‘reference change’’ for comparing two consecutive
5 Ricós C, Alvarez V, Cava F, Garcı́a-Lario JV, measurements. Clin Chem 1983; 29: 25 – 30.
Hernández A, Jiménez CV, Minchinela J, Perich 13 Hyltoft Petersen P, Fraser CG, Kallner A, Kenny
M, Simón M. Current databases on biological D. Strategies to set global analytical quality
variation: pros, cons and progress. Scand J Clin specifications in laboratory medicine. Scand J
Lab Invest 1999; 59: 491 – 500. Clin Lab Invest 1999; 59: 475 – 585.
6 Sociedad Española de Bioquı́mica Clı́nica y 14 Sandberg S, Thue G. Quality specifications derived
Scand J Clin Lab Invest Downloaded from informahealthcare.com by University of Guelph on 05/16/13

Patologı́a Molecular. Comité de Garantı́a de la from objective analyses based upon clinical needs.
Calidad y Acreditación de Laboratorios . Especi- Scand J Clin Lab Invest 1999; 59: 531 – 4.
ficaciones de la calidad analı́tica en laboratorios 15 Stöckl D, Baadenhuijsen H, Fraser CG, Libeer JC,
con distintos niveles de recursos. Quim Clin 2000; Hyltoft Petersen P, Ricós C. Desirable routine
19: 219 – 36. analytical goals for quantities assayed in serum.
7 Ricós C, Alvarez V, Cava F, Garcı́a-Lario JV, Discussion paper from the members of the
Hernández A, Jiménez CV, Minchinela J, Perich External Quality Assessment (EQA) Working
M, Simón M. Biological variation database. http:// Group on analytical goals in laboratory medicine.
www.westgard.com/guest21/htm Eur J Clin Chem Clin Bichem 1995; 33: 157 – 69.
8 Sociedad Española de Bioquı́mica Clı́nica y 16 Fraser CG, Hyltoft Petersen P, Libeer JC, Ricos
Patologı́a Molecular. Comité de Garantı́a de la C. Proposals for setting generally applicable
Calidad y Acreditación de Laboratorios. Base de quality goals solely based on biology. Ann Clin
datos de variación biológica. http://www.seqc.es/ Biochem 1997; 34: 8 – 12.
bd/vb.htm 17 Fraser CG. Quality specifications. In: Biological
9 Harris EK. Effects of intra and interindividual variation: from principles to practice. Washington:
For personal use only.

variation on the appropriate use of normal ranges. AACC Press; 2001. p. 29 – 66.
Clin Chem 1974; 20: 1535 – 42.
10 Fraser CG. The utility of population based Received: 5 July 2003
reference values. In: Biological variation: from Accepted: 14 January 2004