CAP Laboratory Improvement Programs
The Origin of Reference Intervals
A College of American Pathologists Q-Probes Study of “Normal Ranges”
Used in 163 Clinical Laboratories
Richard C. Friedberg, MD, PhD; Rhona Souers, MS; Elizabeth A. Wagar, MD; Ana K. Stankovic, MD, PhD, MPH;
Paul N. Valenstein, MD
Context.—Standards have been developed for establish-
ing reference intervals, but little is known about how
tervals are determined in practice, interlaboratory vari
tion in intervals, or errors that occur while setting refer-
tence intervals.
Objectives—To determine (1) methods used by clinical
laboratories to establish reference intervals for 7 common
analytes, 2) variation in intervals, and (3) factors that con-
tribute to establishment of “out ervals.
Design—One hundred sixty-three clinical laboratories
provided information about their reference intervals for
Potassium, calcium, magnesium, thyroid-stimulating hor-
‘mone, hemoglobin, platelet count, and activated partial
thromboplastin time.
Results—Approximately half the laboratories reported
‘conducting an internal study of healthy individuals to va
idate reference intervals for adults. Most laboratories relied
‘on external sources to establish reference intervals for pe-
[tert sts te commonly compare to
erence interval before caregivers make physiological
assessments, medical diagnoses, or management det
The importance of reference intervals is underscored by
Us "regulation: the ‘Clinical Laboratory. Improvement
Amendments of 1988 require that Iaboratories that into-
duce an unmodified, US Food and Drug Administaton
Cleared or approved nonwaived test system “verify that
the manufacturers reference intervals (normal values) are
Appropriate forthe laboratory's patent population" Labs
sratorkes that modify US Food and Drug Administration -
Spproved tests or develop thir wn assays are required
WO establish ther owen relerence intervals for thir assays,
‘Accepted for publication Sepiember 8, 2008
From the Department of Pathology, BayState Medical Cente, Sping
Fld, Mas (Dr Fiedbeugy, the Depart wsis, College of
Anetiran Paholgise, Norivikl ll (Me Souersy the Departs of
Pathology, The David Geflen School of Medicine a UCLA, Lox Angeles,
Clif (DF Wagar) Preanalytical ‘Systems, BD Diagnoses, Fanklin
Likes, NJ (Dr Stankovi), and Department of Pathology, t Joseph Mer
(6) Hospital, Ann Arbor Mich (DF Valente)
tutors have nw telean ruarrsalameses she pwd
‘companies described in this article
Reprints: Paul N Valenstein, MD, Department of Pathology St Joseph
Mercy Hospital, Ann Arbor, MA 48106-0995 (e-mail paul valentin,
om
248 Arch Patol Lab Med Vol 121, March 2007
diatric patients. There was slight variation in intervals used
by the central 80% of study laboratories, but some labo-
ratories outside the central 80% had surprisingly low and
tervals. In some cases the
hhad no overlap. For ex-
ample, one laboratory considered a hemoglobin of 13.8
g/dL ina woman to be “low” while another considered the
same value to be “high.” Three percent of reference inter-
vals contained a limit that qualified as an “outlier” using
standard statistical tests; we could not identify any practice
Conclusions-—Many laboratories adopt reference inter-
vals from manufacturers without on-site testing of healthy
validated with on-site studies.
(Arch Pathol Lab Med. 2007;131:348-357)
Regulations also specity that reference intervals be includ-
tin laboratory ports or made available upon request
to individuals who order test, pos
Reference intervals are of 2 types? The most common
type has been termed feassothted and is derived fon
2 Feference sample of persons who ate in good health
For example, the reference interval normally reported for
Serum potasslum is health-associated. The cental 93% of
the healthy adult population tested by many laboratories
has serum potassium levels between 33 and 5.1 mEq/L,
and these Hts define the serum potasium reference in:
terval The other type of reference interval has been
termed decsion-hnef and defines specific medical decision
limits that clinicians use to diagnose or manage patients
For example, a serum total cholesterol of more than 200
mg/L. defines the level at which diet and exerse are ist
recommended by most authorities fo lower cholesterol in
otherwise healthy adults Smilary the International Nor
malized Ratio range of 2.0 40 30 defines what some au-
thors consider appropriate long-term anticoagulation in
gate thot a itn risa who have
Elude Medical bileaet prosthetic aortic valves” Refer:
ence intervals that incorporate medical decision limits are
Sten defined with cinal trials and adopted by labora
tories from the medical iterstur. In ths study Wwe inves
tigated the process clinical laboratories use to establish
helti-asoctated reference intervals. These intervals are
Reference Intervals—Fredberg etalalso popularly known as reference ranges, normal values, no
ina res, blogic! reeence intersected Tees,
Health-associated reference intervals vary from labora-
tory to laboratory. For example, ina survey of 525 clinical
laboratories the lower limit of the laboratory's total serum
calcium reference interval ranged from 8.3 0 88 mg/dL
(th and 90th percentile of laboratories) and the upper
limit from 10.2 to 10.7 mg/dL. Some of this variation may
have been due to differences among laboratorics in clinical
service needs, analytic platforms, populations of healthy
individuals, or analytic imprecision that was present when
reference intervals were determined. Yet some of the var-
iation may have resulted from different approaches labo-
ratories used to establish their reference intervals, as the
[process for establishing health-associated reference inter
‘als has historically been poorly defined.
No single authoritative Source specifies the process that
clinical laboratories should use to establish health-associ-
ated reference intervals. The document that most closely
approaches an authoritative source is the Clinical and Lab-
‘oratory Standards Institute (CLSI; formerly NCCLS) doc-
uument C28-A2. The CLSI document is based on the sem-
inal work of Solberg and colleagues;~ who served in the
19805 on the Expert Panel on Theory of Reference Values
of the International Federation of Clinical Chemistry and
the Standing Committee on Reference Values of the Inter-
national Cotineil for Standardization in Haematology.
This CLSI standard recommends one approach for es
tablishing reference intervals of a newly developed or
modified analytical test system andl a second, abbreviated,
20-specimen approach for validating the transfer of refer
ance intervals among, comparable analytical plaorms,
The more involved approach is applicable to instrument
manufacturers first determining reference intervals for
their test systems or to laboratories that develop their own
assays of modify commercially available assays. The less
involved appro allows linia Iaboatores using
modified US Food and Drug Administration-approved as-
says to validate reference intervals supplied By the man-
ufacturers of their analytic instruments.
The current version of the Centers for Medicare and
Medicaid Services (CMS) Suraey Procedures and Interpretive
Guidelines for Laboratories and Laboratory Seroices indicates a
“Maboratory must evaluate an appropriate number of spec-
imens to verity the manufacturers claims for normal val
tes of, a8 applicable, the published reference ranges.”
The CMS provides no guidance about what constitutes an
“appropriate” number of specimens, when age-specil
and secspecificinterals need to be establishes, of how
data from tested healthy individuals are to be used to
verify a manufacturer’ claims. The CMS agents make
judgmental determinations about each facility’ efforts to
Yalidate reference intervals on a case-by-case basis during
the course of laboratory inspections.
Lite is known about how laboratories establish health-
associated reference intervals in practice. During 2004,
1.3% of laboratories enrolled in the College of American
Pathologists (CAP) Laboratory Accreditation Program
‘were cited by inspectors for failing to validate reference
intervals properly, but the specific types of omissions were
rot documented and it snot clear whether inspectors ape
proached this problem in a consistent manner A survey
{f 500 laboratdries conducted by the CAP in 2001 found
that 390 (78% of laboratories) adopted manufacturers’
published values for their reference’ intervals Appro
‘Ach Pathol Lab Med—Vol 131, March 2007
mately one fifth of laboratories reported receiving, hel
from manufacturers in validating 2 manufacturers refer
cence interval either inthe form of stalstcal consultation,
test materials, or procedures. The survey did not collect
information about the proportion of laboratories. that
adopted manufacturers’ felerence intervals without test-
ing any healthy individuals.
The aims of the present study were to (1) describe meth-
‘ods actually used ‘by clinical laboratories to establish ref-
ference inteFvals for common analytes, (2) document inter
laboratory variation inthe reference intervals used in prac-
tice, and (3) identify institutional factors and practices that
influenced the reference intervals a laboratory adopted.
We considered the aims of this study to be important
for two reasons: Fist, the validation of reference intervals
can consume considerable laboratory time and resources.
Methods for validating reference intervals that require less
effort but that still produce reliable intervals may be at-
tractive to clinical laboratories operating with tightly con-
strained resources. Second, we were concerned that some
laboratories may have adopted unusual reference intervals
that posed patient safety risks, because of some sort of
‘oversight or conceptual ezzoe. We wished to examine how
often “outlier” reference intervals (defined in the “Mate-
als and Methods”) were found in laboratories, and
‘whether any’ particular laboratory practices predisposed
to the adoption of outlier interval
MATERIALS AND METHODS
Study Design
The study was conducted according tothe Q-Probes study for
‘mat previously described, which reli on a convenience sample
‘of clinical laboratories that subscribe to the CAP Q.Prabes bench-
‘marking program: After refinement of a standardized data cok
Tection instrament, CAP Q-Prabes subscribers were male data
collection instructions in Tate 2005,
Participants wore asked to provide thei laboratories’ lw and
high vals for referonce intervals fr 7 analytes: potassium, total
calcium, magnesium, thyrotd-stimulating hormone (TSH), be-
moglobin,phtelet count, and activated. partial Uuromboplastin
time (aPTH}. Adult values (assuming a 32-year-old male inpae
tient) and pediatric values suming an S-yearold male inpa-
tient) were collected. In addition, hemoglobin reference intervals
were collected for female adult 3nd pediaeic inpatients,
Fr each analyte, the fllowing additonal information was col
lected unit of measure, primary specimen type, ana¥tc instruc
rent manufacturer year the reference interval was orginally es
tablished, year of the most recent revalidation of the relerence
interval, and year the primary instrument was placed nto ser
vice: The methods the laboratory used to determine reference
Intervals for each analyte were also ascertained. Participants were
ssked whether they festod "healthy individuals” as part of the
process they used to determine reference intervals, ut no wxplict
{Sefiition of» healthy individual was provided.
Reference intervals for pointo-care instruments and capillary
(fngestick oF earlobe) blood collections were exchided. In ad
shtion, intervals that were in the process of being reviewed were
‘aclu, a were intervals for secondary laboratory testing sites
Sand analyzers (Gf more than one testing site or analyzer was in
tse and reference intervals ifered between site or analyzers).
Participants vere also queried about several institutional char-
acteristics occupied bed siz, teaching satu, pathology resident
training status, government afation, institution location, istic
tution type, CAP inspection statu, and inspection status by the
Joint Commission om Accreditation of Healthcare Organizations.
Laboratory Characteristics
Participants from a total of 163 institutions submitted data,
Most of the institutions (27%) were located in the United States
Reference Intervals—Friedberg et al 249‘with the remaining located in Canada (2), Australia (1), Lebanon
(), and South Korea (1). Approximately 31% of participating in-
stitutions were teaching hospitals and 15% had 8 pathology re=-
idency program. Within the past 2 years the CAP inspected 78°
‘of the study laboratories. Hospital or laboratory inspections were
conducted by the Joint Commission on Accreditation of Health-
‘are Organizations at 66% of participating institutions. Table 1
displays characteristics of participating institutions.
For chemistry assays the majority of institutions reported that
they most commonly tested serum, rather than plasma. Ninety-
‘one institutions (56.5%) most commonly tested for potassium us-
ing serum 94 (59.1%) used serum for cakium; 92 (579%) used
serum for magnesium; and 120 (78.9%) used serum to test for
TSH.
Statistical Analysis
Where required, calcium and magnesium reference interval
data were standardized to milligrams per deciliter (if they had
bbeen reported in millimoles pet liter or millequivalents per lite).
Prior to performing statistical tests fr association, values were
screened for outliers, Several participating institutions did not
fnswer all of the questions on the questionnaire about demo-
‘graphic characteristics, institutional practices, or reference inter-
‘als for particular analytes or age groups. These institutions were
excluded only from tabulations and analyses that required the
Imissing data elements All statistical analysis were performed
using SAS v9.1 (SAS Institute Inc, Cary, NC).
The low and high values for the reference intervals were tested
for associations with the institutions” demographic and practice
variable information in Tables 1 and 2, as well as the analytic
platform used by the laboratory. Individual associations were
First tested using the nonparametric Kruskal-Wallis test. Variables
‘with significant associations (P ~- 10) were then inclided ina
Forward selection regression model. All remaining variables were
significantly associated atthe .05 significance level
‘We used multivariate analysis of variance to perform a joint
analysis ofthe dependent variables: low and high adult reference
intervals and low and high pediatric reference intervals. This ap-
‘proach simultaneously tests whether the mean low /high vectors
re statistically different for the analytespecific predictor vari
fables. Since the cell counts for the predictor variables were not
‘equal, the significance level was set at OL
‘We investigated outliers/atypical reference intervals using sev-
ral techniques. First, we screened reference interval limits for
‘outliers using 2 tests, the Tukey procedure (in which an outlier
is defined as any value less than the first quartile — 15 > Inter-
quartile Range or greater than the third quartile 1.5 X Inter-
‘uartle Range) and a 2-pass/3 SD procedure (in which an outlier
iS defined as any value tha fell more than 3 standard deviations
from the mean, either during a first pass with al data included
‘or during a second pass in which outliers identified during the
first pass were excluded). We also used a logit regression model
to examine whether unusual reference interval lmits—thove in
the upper or lower decile—wvere associated with any of the de-
‘mognaphic and practice variables listed in Tables Land 2. For the
logit regression, the significance level was set at 05
RESULTS
When Are Reference Intervals Established?
A number of institutions reported that they did not
know the year that their original reference inierval had
been established or the date of their most recent reference
interval revalidation. Nine participants (5.5%) could not
provide the year of their most recent revalidation of aPTT
reference iniervals, while 30 participants (18.4%) did not
know the most recent year that potassium reference inter-
vals were revalidated. Among laboratories that reported
the year of their most recent revalidation, most had reval-
idated reference intervals within the past 5 years. Exclud-
ing aPTT, approximately two thirds reported that they re-
350 Arch Pathol Lab Med—Vol 191, March 2007,
Table 1. Characteristics of Participating Laboratories”
Tnstiatons
Characteristic Non %
Trattaion pe
vivate, sone 8
State, county or ity hospital 18
University hospital 6
Governmental federal 5
Independent lab 5
Other 8
Occupied beds
0-150 st aa
151-300 a as
301-450 % asa
451-600 10
‘600 3 os
Institution location
ity 63 ara
Suburban ” 28
Rural =u 2
Federal installation lab 2 15
Goverrmentalalliaio
Nongovernmental 104 800
NNonfederal governmental 2 162
Federal governmental 3 va
‘writen policy for establishing, revising, or updating intervals
Yes Nz 73.6
No a 268
Unique intervals fr distinct populations by patient age
Yes 40 870
No Fy) 10
Unique intervals fr distinct populations by patient location
Yes 4 25
No 1s? os
Unique intervals fr distinct provider or practice group
Yes 2 12
No 159 ow
Unique intervals fr distinct populations based on disease type
Yes 6 3a
No 153 962
Unique intervals fr distinct populations by patient ethnicity
Yes 7 44
No 153 956
Reference intervals validated inthe same year as new
instrument acquisition
Potassium 6 63.
Calcium 6 a7
Magnesium 8 ora
TSH 6 0
Hemoglobin (male) 8 63.
Hemoglobin (lemale) fs 596
Platelet count 6 ont
apr 2 295,
TSH indicates thyroid Simulating hormone! APT, activated partial
thromboplastin time.
validated their intervals in the same year that a new an-
alyzer was purchased (Table 1). Nevertheless, in some lab-
oratories and for some analytes the most recent
revalidation of a reference interval had not occurred for
more than 10 years, and one participant indicated the lab-
‘oratory had last revalidated several reference intervals in
Reference Intervals—Fredberg etalTable 2. Source of Reference Intervals"
Tnatons, Nox
Hah Hab
Potassium — Calcium Magnesium TSH (Male) (Female) Platelets PTT
"a
Inwemal study of
heathy individuals 701443) 701438) 701886) 751484) 851535) 836525) 81.613) 130082.
Manufacturer's ecommendations!
inserts 55048) 56050) 570363) 460297) 17007) 1HL4 19020 1207.6)
Published literature
textbooks, 1912.0) 17408 16402) 27) 42070) BA72 Ore 60.)
‘Other laboratories ladopted with
intemal validation) 860 966 861) 1065) 860 861) GN) 502)
Nonlaboratony medical stat
recommendations 20a) 405 203) 426 30% 309 349 203)
‘ther laboratories ladopted with-
out internal validation) 0 = 00) += =~ 203) 108 §=—106 = 10H =~ 203)
Other 425) 425 405) 6G9 203 20a) 24a) 108)
Peciaie
Inwemal study of
heathy individuals 350252) 44241) 36259) 210157) 5238 3420 we28 45.620)
Manufacturer's ecommendations!
inserts 45024) 2098) 45024 S940) 1368 1405) 15105) 11.1)
Published literature
textbooks, 42002) 44012) 2.009 16 75610 756607) sur) 26078
‘Other laboratories ladopted with
intemal validation) 760 867 643) 762) 74H 864) 866 5.0.7)
Nonlaboratony medical stat
recommendations 302) 505 322) 507 407 4en 42H 205)
‘ther laboratories ladopted with-
out internal validation) 107) 204 107) 46.0 1068 1068 96a) 40.0)
Other 64a 64) 586 587 320 320 505) 400)
TSH indicates thyroid stimulating hormone; Hgb, hemoglobin and aPTT, awated pavial twomboplastin time
1983, more than 22 years before the study was conducted
and more than 8 years before the laboratory's current an-
alytic instrument was placed in service.
Methods Used to Establish Reference Intervals
roximately half the participants reported that they
conte a internal stu of healthy individuals to help
establish chemistry and hematology reference intervals fOr
adults, and half relied exclusively upon external sources
(manufacturers’ inserts, published literature, intervals
used at other laboratories, or medical staff recommenda-
tions), For aPTT, 130 facilities (823%) reported conducting
fn internal study of healthy individuals to establish ref
ference intervals. These data are shown in Table 2.
Most laboratories relied on external sources to establish
reference intervals for pediatric patients (Table 2). Ap-
proximately one fourth of participants indicated they per-
formed an internal study on healthy individuals to estab-
lish chemistry reference intervals for pediatric patients,
da spproutately 10's of focitescondocted ntemal
Studies to establish hematology and aPTT pediatric ref
erence intervals.
‘One hundred thirty-six (84%) of 162 laboratories re-
ported that they received some assistance from instrument
manufacturers in establishing, reference intervals. Of lab-
‘oratories that received assistance from manufacturers, 111
(62%) reported receiving statistical support, 91. (67%) re-
Ceived consultative support, 35 40%) followed 2 i
dlr for establishing reference intervals that had been pro-
vvided by the manufacturer, and 27 (20%) received speci
mens for testing from the manufacturer.
‘Ach Patho Lab Med—Vol 131, March 2007
We sh pts sve deal ution abet
tne proses they ud oe potas reference
ters for adults. OF the respondents that indicted they
had conducted an intemal reference interval study of
healthy individuals, half (65 laboratories) indicated that
they ad tested_specimens from between 21" and 50
I nda and ene unter (2 abortr in
dicated they had tested more than 10) specimens. OF the
Sites that feted heathy individuals to help establish ret
tence intervals, 56 sites (9%) set thei reference intervals
ting the mean ofthe tested reference population plus or
minus 2 standard deviations, while the remainder of re
Spondents (73 sites; 37%) used test resulls 0 “verify” an
tktemal reference interval obtained from the manufactur:
lished textbook, or some other source. These data
Ste shown in Table 3 Laboratories that set their reference
intervals from the results of intemal testing (mean = 2
$1) did not fend fo fest more healthy patients than labo
ratoies that used internal testing 10 "erly" at external
interval (Table 4 chisquare = 257; P = 33)
Interlaboratory Variation in Reference Intervals
There was slight variation in reference intervals among
the central 80% of study laboratories (Table 5). Upper lim=
its of reference intervals tended to vary slightly more from
laboratory to laboratory than lower limits did. There was
rno dramatic difference between the amount of interlabo-
ratory variation in adult reference intervals and pediatric
reference intervals, even though adult reference intervals
‘were more often validated by testing specimens from
healthy individuals and pediatric reference intervals were
Reference Intervals—Friedberg et al 351Table 3._ Methods Used to Establish Potassium Reference Interval
Twatatons
ota evel ran on Which specimen type
Both plasma and serum 100 ou
Plasma 2 154
Serum ” 228
1 potassium levels run om both plasma and serum, laboratory
“Converts” a result from one specimen type to the equivalent level fom the other type 28 622
Esablishes and reports? diferent reference intervals base on specimen type 0 7.
If an intemal reference interval study was performed for potassium values, how many healthy individuals were studied?
20 4 aa
21-50 6 soa
51-100 2 217
100. 2 28
If an intemal reference interval study was performed for potassium values, how did laboratory’ set the reference interval?
Satstcal analysis of collected data (eg, mean * 2 SD) 56 Ba
‘Manufacturers recommendation after comparing to collected data fom an interal reference interval
‘tu 55 126
Texthook data after comparing to collected data from an internal reference interval study 0 78
Another laboratory's reference intenal after comparing to collected dat rom an internal reference in
tensa study 23
other 5 Yo
most often obtained from external sources. Reference in- Institutional Factors and Practices Associated With
tervals forthe 7 study analytes showed similar levels of
interlaboratory variation
Outside of the central 80% of laboratories there was
more substantial variation in reference intervals. A Tew
laboratories had surprisingly low and high limits for their
reference intervals. For example, ae laboratory reported
that its reference interval for adult polasium extended
down to'3.0 mmol/L, while another feported thatthe up-
per limit of is potassium reerence interval extended up
{0 57 mmol/L In some cass the reference intervals used
by 2 laboratories did not overlap, and the upper lint of
che laboratory’ reference interval was lower than the lov-
er limit of another’. For example, one laboratory consil-
ted an aPTT of 30 seconde to be “low” while another
considered an aPTT of 30 seconds tobe “high.” Using the
Tukey procedure, 40 (3.1%) of 1271 adult reference inter
vals contained at least 1 limit that was an outer. The 2-
pass/3 SD procedure idenified the same total number of
Sule. There were no significant differences in the fac-
tin of outlier reference interval limits among the analytes
wwe studied,
Reference Intervals,
We examined all ofthe institutional factors and practic:
es in Tables Tand 2 to elucidate variables that Were as
Soviated with reference intervals. For several analytes, par
fcular instrument manufacturers were associated with
higher of lower reference intervals, Tables 6 and 7 iis:
trate this relationship, showing statistically significant as
Soriatons between instrument manufacturer and the me-
dian value ofthe upper and lower limits of adult reference
interval, respectively: Tables 8 and 9 show the same Te
lationship forthe upper and lower limits of pediatic re
thence iMerals: Investing, specimen Spe dil not af
fect the potassium reference intervals used by partic
pants even though the potassium concentration i plasma
{lower than in serum™® The method used to establish a
reference interval (adoption from manufacturer versus on-
site testing of a healthy population) was not sssocated
with the interval in se.
We examined whether any ofthe characteristics in Ta-
bles and were associated With the use of aberrant adult
Table 4. Number of Healthy tadivi
als Tested for Potassium and Method Used to Establish Reference Intervals®
Interval Calalated Using Statistical
Taboratoris Determining Reference
Tnteral Adopted From Manufacturer,
No.of Healthy ‘Analysis of Tested Reslls Alter Comparison to Total
Individuals Tested “ean 28D) Internal Tested Results Laboratories, No.)
50 w a a a)
51-100 a 15 28.23)
“100 16 2 28.23)
Total Laboratories 56 (46) 6754) 123.100),
No association was detected between The number of ested individuals and the method used to tern the
a2 Poa
352. Arch Pathol Lab Med—Vol 121, March 2007
france terval, aque
Reference Intervals—Fredberg etalTable 5. Variation in Reference Intervals of Selected Analyles Among 163 Laboratories
‘All stution Percents
Analyte, 50m
Population, and Litt Minimum 100h——25th_—_(Medlan)—_75th_—_—_90th Maximum
Pousiom, mmol
‘Adult, low 12 3033358
‘Adal, igh 1 45 50 0ST 3208357
Pediatric ow wa 30003303535
Pediatric, high 41 50 OST 5205456
(Calcium toa, mgt
‘Adult, low 12 7683S 8S 8D
‘Adal, high tol 96 101 tT tS Sta
Pediatric ow 576 84 BABS 8B 8G
Pediatric, high 439% T2007 tS
Magnesium, mgd
‘Adult, low 7 13S
‘Adal, igh 37200222322
Pediatric, low wa 121566 we
Pediatric, high wa 200212 od
‘Thyroid stimulating hormone, LL
‘Adult, low 154 010030034035 0447050
‘Adal, igh 155 300420468494 5.50, 5.60.00
Pediatric low 133 010-030-034 03504050 0.70
Pediatric, high 136 400440475 5.00 5.50 5.90 8.00
Hemoglobin (male), g/l.
‘dul, low i 115 OOH ttt
‘Adal, igh tel 15016517075 ttt
Pediatric ow Me '8S 03 OS? So
Pediatric, high 8027137118200
Hemoglobin (female, pL
‘Adal, low 39° 14ST OO tO
‘Adal, high 139135150155 te0 tet tO
Pediatric ow 9001055130
Pediatric, high 8 7174S DSTO TBO
Platelets, x10"
‘dul, low 1200030
thigh tor 316 400-400, 4004050500
Pediatric, low 146 100130015050
Pediatric, high 1332532 4004005050539
‘Activated partial thromboplastin time, s
‘Adult, low 13617 2 2 2B 25 2% oR
‘Adal, igh 15928 32 33 35 a BG
Pediatric low 350 2 2 2 25 %
Pediatric, high 13828 2 33 35 a BG
~ Taw indicates the Tower Tit ofthe reference inten high the upper Tint ofthe relerence interval
sfc nal ins i he upper or er dei) I~ Some of th borne th st Hey als cate
Stitutions that established their felerence intervals before their own reference intervals from their findings, but most
2001 were 3.1 times more likely to have aberrant high use inlaboratory testing to “validate” reference intervals
PTT limits than were institutions that established their supplied by manufacturers.
intervals during 2001 to 2005 (P — 02) Institutions that Although different approaches were used to establish
placed test instruments in service before 2002 were 13 reference intervals, no particular approach was associated
fimes more likely to have aberrant low TSH limits than with the reference interval that laboratories ultimately es-
were institutions that placed instruments in service in tablished. In other words, there was no discernable differ
2002 to 2005 (P= 01)" No other factor was statically ence between the relerence intervals of laboratories that
‘ssociated with aberrant reference interval limit. adopted manufacturers’ reference intervals without fur-
ther testing, laboratories that tested healthy subjects to
COMMENT validate manufacturer supplied intervals, and laboratories
To our knowledge, this is the frst large survey describ- that tested healthy subjects to caleulate their own rete
ing how reference intervals are_actualy established by ence intervals
clinical laboratories “in the field” We found that approx: Since the approach laboratories used to determine ref-
imately haf of all laboratories test healthy adults toestab- erence intervals did not appear to influence the interval
lish reference intervals, but few test healthy children. that was ultimately established, we question the conver
‘Ach Pathol Lab Med—Vol 131, March 2007 Reference IntervalsFriedberg et al 353TSH indicates thyroid stimuating hormane; APT, activated paral
thromboplastin time
tional wisdom that itis always necessary for laboratories
to validate manufacturers’ intervals with on-site testing
before adopting the interval locally. I'a manufacturer ad
quately describes the process it used to establish a
hhealth-associated reference interval, and the laboratory is
using the manufacturer's analytic platform in accordance
‘with the manufacturer's instructions and is testing a sim-
ilar population, validation of a manufacturer's interval
might consist of nothing more than the laboratory medical
director making a professional determination that the sup-
plied interval is appropriate for the local laboratory. This
Approach is contemplated in the CLSI standard,” but not
in the current version of the CMS survey procedures."
Klein and Junge! sound a note of caution about adopt-
ing manufacturers’ reference intervals without local on-
site testing of healthy individuals; the authors have con-
cers that preanalytic procedures used by manufacturers
(uch as specimen collection and storage procedures) may
not be adequately duplicated by every laboratory, creating
the need for local testing of healthy patients to ensure that
manufacturers’ intervals are locally applicable. While rec-
‘ognizing that the adoption of maniufacturers’ reference in-
tervals without on-site validation poses some risk, our
study suggests these risks are already being assumed by
most laboratories that report pediatric reference intervals,
since little testing of healthy children is taking place in
laboratories even though specimen collection procedures
for pediatric patients show a great deal of variability from
institution 40 institution. ‘The establishment of reference
intervals for “special fluids” poses problems similar to
those sen in pediatrics, a8 spetal ld samples are rarely
354 Arch Pathol Lab Med—Vol 121, March 2007
Table & Relationships Between Inorument Table 7, Relationships Between notameat
Manuel Ada efrence tral Upper Manure at Adare Ira Lower
Te Ti
‘it ‘i
tetence tetence
Ante, ince nay ‘nc
ni a are neat ‘Upper uni ae, woot owe
porte vsti ‘Ei sn anaere instal
Naan, mga (= OO) TH aU P= 900
Mies 4230 toche now
Dae Bring ey fochet lich Hest
Poe aOR Bettina Recess 3 tbe
Balkan Sychron a i Baan Bon
foc haat ae) Be Bo Ons
foche Bho vio bow
TSH, mitt. (P= 001) ‘Abibor 8 048
te no 420 srr s 000)
fiche tachi Se Biotrun Pati Boon
hen ete bade etn 8 i B
Ci ma Festi arts BOS
Dae Bring bin Das Being Actin st BO
Poe ub Bagosterdasmarrra te
BRL Aces B38 Rohruriemelonce syahasi._— 1825
ares P= 001 Ta nts Hon snag Homnone APT TST Pa
Dade Behring Actin FS 10 33 "hrombeplastin time
Botte Pain SR
Dade Beg tn FS es
HeneiterlinceSynias, 183 collected from healthy patients without ciical suspicion
iagnosica Sago STA-PTT A 8 36 i
Blames Sage f 8 of disease and reference intervals tend tobe adopted from
extemal sources without on-site validation.» Reference in
tervals for children and special Muids can beset by mul-
tiple cooperating laboratories but diferences between
test systems make this approach somewhat dificult to m=
lement."
PaWhile we do not believe local testing of healthy patients
is required before adopting a well-documented reference
interval from a manufacturer there s nothing wrong with
2 loca laboratory performing on-site testing of Healthy ne
dividuals using te abbreviated 20specinen CLSI proce:
dure to validate a manufacturers reference interval Ir no
facility performs omsite validation of manufactures rel
ftence interval the entire laboratory community will be
relying on manufacturers and their regulators fo ensure
that intial reference interval studies are performed cor
realy: We must also point out that a dection by 2 labo-
talory ditector to forgo on-site testing of healthy individ
tals forthe purpose of establishing a reference interval
does not relleve the laboratory ofits obligation fo ade-
Guately calibrate a new instrument and test the instru
thent®s analytic accuracy and precision before placing it
into service
We found that for most institutions, interlaboratory var
iabilty inthe reference interval values for our study an-
alytes was fairly low. The type of test instument in use
was the main source of reference interval variability that
wwe could identify explaining much of the variation in el
‘rence intervals, Varatoninfeference intervals among =
alytic platforms is probably appropriate given that major
instrument platforms show analytic bia relative 0 one
another
Were the reference intervals used by most study partic:
ipanis accurate? Without acces fo the patient populations
tring tested by each of our study laboratories, we cannot
answer this question diectly. However, we collected ine
direct evidence that the potassium reference interval used
Reference Intervals—Fredberg etal‘ales (approximately 03 mmol/L, depending on the an
Biylc test System). Only 28 sites “converted” results for 1
Specimen type tothe equivalent level from the other type
before reporting, and only 17 sites established diferent
tec ners fr ea