CLIN.CHEM.38/7,1365-1370(1992)

CLINICALCHEMISTRY,Vol.38, No. 7, 1992 1365

Testing for Creatine Kinase and Creatine Kinase-2 in Ontario: Reference Rangesand Assay TypesA. R. Henderson,’ M. J. McQueen, R. L. Patten, S. Krishnan, D. E. Wood,2 and S. Webb

In 1991,246 and 136 OntarIo laboratories performed totalcreatine kinase (CK; EC 2.7.3.2) and creatine kinase-2(CK-2) assays, respectively. A questionnaire mailed tothese laboratories requested information about the typesof assay used, the origin of their reference ranges, andthe source of their instruments and reagents. All labora-tones used current test formulations for CK, afthoughseven laboratories did not assay at 37#{176}C.For CK, 69% ofall laboratories reported different upper reference limitsfor men and women (5th-95th percentiles: 160-250 and115-215 U/L, respectively); 31% reported similar rangesfor both sexes. Fifty-six percent derived their own ranges;the remainder used either kit inserts or literature refer-ences, and nearly 60% of this latter group claimed to havevalidated these suggested ranges before use. For 6% ofall laboratories, their pediatric ranges were similar to theiradult ranges. For CK-2, only 32% used their own refer-ence range; the remainder used kit inserts or literaturereferences, but only 49% of this group validated theseranges before use. Reference limits (5th-95th percen-tiles) for CK-2 were as follows: actMty 6-24 U/L; fractionof CK, 0022-0.06; and, for mass assays, 5-10 tg/L andrelative index 0.015-0.04.

Additional Keyphras..: isoenzymes - proficiency testing

Mandatory external quality assurance for Ontariolaboratories comprises licensing and inspection by theOntario Ministry of Health and, since 1974, mandatoryproficiency testing by the Ontario Medical Association

Enzyme and Lipids Committee, Laboratory Proficiency TestingProgram, Ontario, Canada.

‘Address for correspondence Department of Clinical Biochem-istry, University Hospital, Windermere Road, London, Ontario,Canada, N6A 5A5.

2Address reprint requests to this author at Laboratory Profi-ciency Testing Program, 250 E. Bloor St., Suite 501, Toronto,Ontario, Canada M4W 1E6.

3The following manufacturers are indicated: Abbott: AbbottLaboratories, Abbott Park, IL; American Monitor American Mon-itor Corp., Indianapolia, IN; Baxter Baxter Healthcare Corp.,Irvine, CA; Beckman: Beckman Instruments, Inc., Carlsbad, CA;Boehringer: Boebringer Mannhelm Canada Ltd., Laval, Quebec;Ciba-Corning Ciba-Corning, see Gilford; Coulter Coulter Elec-tronics, Inc., Hialeah, FL;Dade: Dade Division, Baxter HealthcareCorp., Irvine, CA; Du Pont: Du Pont Co., Wilmington, DE; EM:Electro-Nucleonics, Inc., Fairfield, NJ; Gilford Gilford Instru-ments Lab., Oberlin, OH; Hitachi: see Boehringer Hybritech:Hybritech Inc., San Diego, CA; IL Instrumentation Laboratory,Inc., Lexington, MA; Kodak: Eastman Kodak Co., Rochester, NY;Olympus: Olympus Corp., Lake Success, NY; Roche: Roche Diag-nostic Systems, Nutley, NJ; Sigma: Sigma Diagnostics, St. Louis,MO; Technicon: Technicon Instruments Corp., Tarrytown, NY.

Province of Ontario uses SI units. Accordingly, we reportpercentage of total CK activity as a fraction; i.e., 5%is reported as0.05.

Received September 13, 1991; accepted March 3, 1992.

under a contract from the Ontario Ministry of Health.The Laboratory Proficiency Testing Program (LPTP), aunit of the Ontario Medical Association (1), has as itsprimary functions monitoring the performance of allOntario patient-testing laboratories, whether owned byhospitals or commercial enterprises, and providing ed-ucational assistance to all such laboratories. Testing isassigned to 14 working committees, one of which is theEnzyme and Lipids Committee.

During surveys assessing performance of determina-tions of creatine kinnae (CK; EC 2.7.3.2) isoenzyme 2(CK-2, or CK-MB), in which all participating laborato-ries were asked to provide their upper reference rangesfor both total CK and CK-2 in men (2), the Enzyme andLipids Committee noticed a wide range of reportedvalues for these analytes. In addition, our experience ofon-site inspections of laboratories having an unsatisfac-tory performance in proficiency testing suggested thatlaboratory staff, even senior staff, were not always fullycognir.ant of the details of a particular assay system.Given that the current formulation of the CK assay isparticularly complex (3), we believed that inquiry intothe principles of that assay might also be of value.Therefore, we issued questionnaires to all laboratorieslicensed to perform total CK and CK.2 assays, to obtainfurther information about these assays with regard to

#{149}the methods in use and the level of laboratorians’understanding of the principles on which the methodsare based

#{149}the ranges used for male, female, and pediatricpopulations, and

#{149}the way in which these ranges were obtained (4), i.e.,by the laboratory’s own determinations, by the manu-facturers analytical method (package insert), or bymethods found in the literature; if the latter two sourceswere used, did the laboratory validate these derivedranges locally?

The health-care system of Ontario serves a populationof 9.5 million people (-36% of Canada’s population),similar to the population size of Michigan (9.2 million)or Ohio (10.8 million); therefore, the results of thissurvey are likely to be of interest to other North Amer-ican health-care organizations. To our knowledge, this

is the first such survey of an entire health-care systemunder a single administrative organization.

Materials and Methods

LaboratOrieL The 246 and 136 Ontario laboratorieslicensed, respectively, to determine total CK and CK-2were surveyed by an LPTP postal inquiry to eachlaboratory director in April 1991.

CK questionnaire. The LPTP Committee asked for the

assay temperature, thiol activator(s), adenylate kinaseinhibitor(s) used in the assay, and the manufacturer’smethod insert.3 All likely thiol activators and adenylatekinase inhibitors were listed on the questionnaire, sothe respondent merely had to check the appropriateboxes. In addition, we requested the current referenceranges for male, female, and pediatric populations, anddetails of how and when these ranges were obtained(population, number of subjects, preanalytical condi-tions, etc.). Alternatively, if the reference range was notderived by the laboratory, we asked for its source-manufacturer’s kit insert or a literature reference. Wealso asked whether the laboratory had validated such areference range in any way. Finally, comments wererequested on any difficulties encountered while obtain-ing such ranges. Those 196 (79.7%) laboratories thatprovided unsatisfactory or incomplete data received asecond request for further data.

CK-2 questionnaire. The content of this questionnairewas similar to that for total CK, described above. Inaddition, we asked whether the laboratory used thefraction of CK activity due to CK-2, or the relativeindex in the case of mass (immunological) assays [rela-tive index = CK-2 (tg/L)/total CK (U/L)J, or CK-2activity alone.4 If this last mode of reporting was used,the respondent was asked to justify this procedure witha literature reference. Here, 114 (83.8%) laboratoriesprovided unsatisfactory or incomplete data and receiveda second request for further data.

Analysis of questionnaire responses. There was aneventual 100% response to the questionnaires. All re-sponses were entered into a MuM1’S database program,written in-house in the Department of Clinical Bio-chemistry, University Hospital, London, Ontario, whichwas used for all subsequent analyses. Each laboratorywas coded by its three-digit LPTP number and by thethree-character method code assigned to the assay oftotal CK (i.e., thiol activator, analyzer, reagent source)and CK-2 (i.e., measurement principle, analytical de-vice, reagent source).

LPTP total CK survey results. The results of an LPTPsurvey, conducted in January 1991, for two specimenswith total CK activities near the upper reference limitfor men (all-method means of 183 and 229 U/L), werecategorized by analyzer type.

ResultsTotal CK Assay

Of the 246 laboratories, 97% used an assay tempera-

ture of 37#{176}C;the remaining 7 laboratories used 30#{176}C.The characteristics of the total CK assays are listed inTable 1.

The major analyzer types, used at 37#{176}C,are describedin Table 2. Eight analyzer types accounted for 79% ofthe analyzers used in the Province for total CK analy-ses. Figure 1 shows the upper reference limit for men byanalyzer type. We also classified upper limits for men bysome locations (Figure 2). Data obtained from Hamilton(Figure 2C), where there is a citywide laboratory sys-tem, should be contrasted with data from London (Fig-

1366 CLINICALCHEMISTRY,Vol.38, No.7, 1992

labia 1. CharacterIstics of CK Assays (n = 246)n

ThiolactivatorNAcetyl-L-cysteine(NAC) 219 89.0Dithioerythritol (DTE) 11 4.5Monothioglycerol 12 4.9NAGand DTE/dithiothreitol 4 1.6

Adenylatekinase inhibitorsAMP and diadenosinepentaphosphatea 191 77.6Diadenosinepentaphosphatealonet) 55 22.4a One laboratoryalso used NsF with this combination.bAll Kodakand Paramac analyzers (see Table 2) and an Olympus AU5031

analyzer.

Table 2. Analyzers (n = 246) Used to Assay TotalCreatlne Klnase Activity

n S

Kodak Ektachem and DT6O 43 17.5Roche Cobas MIRA 39 15.9Hftachisystems 36 14.6Technicon RA-1000 20 8.1Abbott Spectrum 17 6.9Coulter DACOS 14 5.7Abbottblchromatic systems 14 5.7Baxter Paramax 11 4.5ILDimension 10 4.1ILMonarch 10 4.1Otheranalyzersa 32 13.0

a Beckman systems, Roche Cobas FARA,Ciba-Coming EXPRESS, DuPont aca,Olympus systems, TechniconAXON, ILGenesis, AmericanMonitorKDA Gilford systems, and ENIsystems.

ure 21)), where there is not. A group (n = 11) of adjacentcommunity hospitals has partially completed the pro-cess of generating a group reference range for theirentire district (Figure 2F.

Upper reference limits (37 #{176}C).Of the 239 laboratories,74 (31%) had the same reference ranges for both menand women (a further 6% did not report any referencevalue for women at all): the 5th-95th percentiles were130-253 UIL (median 206 U/L). Of the 239 laboratories,41 (17.2%) used their own reference range, 36 used themanufacturer’s kit insert range, and 4 used a rangefrom a literature source; 7 of these laboratories usedmore than one source for their reference ranges.

Of the remaining 165 laboratories (69%), 93 (38.9%)used their own reference range, 72 used the manufac-turer’s kit insert range, and 8 used a range from aliterature source. Twelve laboratories used more thanone source for these ranges. Of the 68 laboratories usinga manufacturer’s or literature reference range, 48(70.6% of 68 laboratories) claimed to have validated theranges they were using. When the ranges for each sexwere dissimilar, the following 5th-95th percentiles (andmedian values) were reported for the total CK assaysperformed at 37 #{176}C:men (n = 237), 160-250 (205);women (n = 142), 115-215 (155); male-female differ-ence (n = 142), 17-90 (50); laboratories reporting theirown range (n = 93), 110-243 (205); and, laboratories

DHAMILTON

5 iI $

D

FIg. 2. DistributIonof total CKupper reference limits for men tor allassays performed at 37#{176}C,by geographical locationA, Greater Toronto;B, Ottawa; C, Hamilton; D, London;E Windsor;F, districthospital group

Table 3. CharacterIstIcs of CK-2 Assays (n = 136)

,Fig. 1. DIstribution of total CKupper reference limits for men forallassays performedat 37#{176}C,by mainanalyzer typeEach manufacturer’s recommended upper limit(U/L) for total CK activity formen Is as followsfor each analyzer A, Kodak analyzers (374); B RocheCobas MIRA (186); C, Hitachianalyzers (195); D. Technlcon RA-1000(243);E Abbott Spectrum (235); F, Coulter DACOS (206); G, Abbott blchromatlcanalyzers (235); Baxter Paramax (224)

n S74 54.429 21.314 10.38 5.96 4.45 3.7

CLINICAL CHEMISTRY, Vol. 38, No.7, 1992 1367

using a reference range obtained from a kit insert (n =

72), 160-267 (195).The Province’s two wholly pediatric hospitals used

their own range (5) or a range from a standard pediatricreference source (6). At other laboratories in the Prov-ince, the majority (80.9%) of laboratories responded thatthey did not have a need for pediatric ranges; however,26 laboratories noted the age-dependence of theseranges, a further 6 laboratories mentioned that pediat-ric ranges were wider than adult ranges, and 15 labo-ratories stated that their pediatric ranges were identicalto their adult ranges.

We did not observe any relationship between the

ImmunolnhibftlonElectrophoresisMicroparticle enzyme immunoassay (Abbott)Du PontcolumnFluoromettic enzyme immunoassay (Dade Stratus)Immunoenzymetric(Hybrftech Inc.)

reported upper reference limit for men and the surveyresults (data not shown).

CK-2 Assay

The types of assay used by the 136 laboratories todetect CK-2 are shown in Table 3. Most of the imniu-noinhibition assays were from Roche (37.8%), Kodak(23%), Boehringer Mannheini (17.6%), and Sigma(10.8%); the remainder were from various other suppli-ers (10.8%). These data indicate a shift from immunoin-hibition methods (63% of the 123 laboratories reportedin reference 2) to the newer imniunoassay (mass assay)techniques-a trend that seems likely to continue, giventhe ease of use of the newer methods.

Ten laboratories (7.4% of 136 laboratories) reported a

- C

I

. . I. Is mae N

- -

EAbbott tIll

1368 CLINICALCHEMISTRY,Vol.38, No. 7, 1992

screen and confirmation strategy for performing CK-2assays. Seven used an immunoinhibition assay forscreening; if the result was positive, six laboratoriesconfirmed by electrophoresis, and one by using anothertype of immunoinhibition assay. The remaining threelaboratories screened with a mass assay and confirmedby electrophoresis.

Upper reference limits. Of the 136 laboratories per-forming the CK-2 assay, 43 used their own values, 94used the manufacturers kit insert range (and 47% ofthese users validated these ranges), and 13 used a rangefrom a literature source, of whom 62% validated theselected ranges. Fourteen laboratories used more thanone source for these ranges. Of the 43 (31.6% of 136laboratories) using their own range, only 2 had differentranges for men and women. Of the 94 (69.1% of 136laboratories) using a kit insert range, only 4 had sex-specific ranges. None of the laboratories using a litera-ture range had sex-specific ranges.

Although 98 laboratories reported upper referenceranges for CK-2 activity in men, only 89 laboratoriesalso provided values for women (six of which differedfrom the range for men). For all assays based on themeasurement of activity, the Sth-95th percentiles formen and women were both 6-24 UIL (fraction: 0.022-0.06), whereas for all mass assays (n = 24) the corre-sponding figures are 5-10 ig/L, with a relative indexrange of 0.015-0.04.

Reference ranges differed within each methodology(Figure 3). Ranges are shown for laboratories that usedelectrophoresis (Figure 3, A and B) or non-Kodak im-munoinhibition methods (Figure 3, C and D). Upperreference ranges for men for Kodak users (n = 17 notshown) were 16 (n = 11), 15 (n = 3), and 10 U/L (n = 3),with a fraction of 0.04 (one each reported 0.05 and 0.06).Responses were sufficient to display mass assay rangesfor only the Abbott IMx (Figure SE).

The methods of reporting CK-2 results are listed inTable 4. Most laboratories used both activity (or mass)and CK-2 fraction of total CK activity (or relativeindex).

Dlscuulon

Understanding of the Principles of the Assay

The responses to the questionnaires gave considerableinsight into the extent of the understanding of theprinciples of the CK assay and of the factors that affectthe assay. Respondents were senior laboratory staff(sometimes the laboratory director); nonetheless, a sur-prising lack of comprehension of the features of anadequate CK assay was often evident. For example,some respondents were content to indicate that theassay did not contain any adenylate kinase inhibitor atall, even though all assays used in the Province actuallycontain such compounds (Table 1). This finding indi-cates a serious deficiency in the knowledge of seniorclinical chemistry staff about this assay. Indeed, 36%and 29% of the respondents omitted to mention theexistence of one ormore of these inhibitors in the totalCK assay and CK-2 activity assays, respectively. Fewer

.,. D

LS

FIg. 3. DIstribution of CK-2 upper reference limits for men byelectrophoresls, Immunolnhibition,and Abbott Mxmass assayA, CK2 activitieedeterminedafterelectrophoreticseparation; B, CK-2fractionof total CK activity determined after electrophoretic separation; C, CK-2activities determined after lmmunoinhlbltion(non-Kodak users); D, CK-2fraction of total CK activity determined after immunolnhibition(non-Kodakusers): E CK-2 mass determined with the Abbott Mx

Table 4. ReportIng of CK-2 Results (n = 136)n S

Assays based on actMty measurements (n lll)CK-2reported by activity and fraction of total CK 79 71.2

activityCK-2reported by activity only 19 17.1CK-2reported byfraction only 11 9.9CK-2reported as “notdetected” 2 1.8

Assays based on measurement of mass (n = 25)CK-2reported bymass and relativeIndex 21 84CK-2reported by mass only 3 12CK-2reported by relative index only 1 4tmlmmunoinhibition,electrophoresis, and column methods (see Table 3).

than 3% of all respondents omitted mention of thepresence of thiol agents in both assays. That theselacunae in understanding were not unique was under-scored by the previously mentioned failure to recognizethat male, female, and pediatric reference ranges differand that a single, sex-neutral range is unacceptable.

Total CK

Nearly all laboratories used the currently recom-mended formulation for the total CK assay (7), or

CLINICALCHEMISTRY,Vol.38, No. 7, 1992 1368

modified formulations that correlate well with the rec-ommended formulation, and include adenylate kinaseinhibitors (8). We therefore expected to obtain a fairlyuniform reference range throughout the Province. Thatis not what we found.

The 95th percentile obtained for the upper referencelimit for men at 37 #{176}Cis very close to those reported byMiller et al. (9) in a study of 550 men and 580 womenand Lott et al. (10) from a population of --8OO0 men andwomen. Although there can be no “gold standard” forsuch values, the essential reliability of the majority ofthe Ontario data appears to be confirmed by the agree-ment with these very extensive US data. However, thevalidity of values >250 and <160 U/L reported by 20laboratories for the upper reference limit for men, and>215 and <115 UIL reported by? laboratories for theupper reference limit for women, when using an opti-mized assay at 37#{176}C,must be questioned. Likewise,despite the significant difference between total CK val-ues in men and women (9-14), 37% of Ontario labora-tories licensed to perform total CK assays did notrecognize this difference. Finally, we seriously questionthe validity of the very wide scatter of the reporteddifferences between the values for men and women,which suggests that the original values may not trulyrepresent each sex group. No laboratory described theeffect of age on the reference range for men, althoughboth Miller et al. (9) and Lott et al. (10) clearly demon-strated this effect.

Considerable racial heterogeneity exists in popula-tions, and these racial differences profoundly affect totalCK values (11-13)--to the extent that different racialreference limits have been suggested (12). In general,such heterogeneity should broaden the reported ranges,but in the 56% of Ontario laboratories that actuallyderived their own reference ranges, such an explanationseems unlikely, because several of these laboratoriesreported rather low values.

Almost all of the plots in Figure 1 show, for themajority of users, a central plateau or a series of pla-teaus with a mid-point near 200 UIL. This is also themedian value of the data obtained for men, which agreeswith the extensive US data mentioned above. In otherwords, the central tendency for each instrument andreagent combination is consistent from instrument toinstrument. However, as already mentioned, someranges for men extend to extreme values. How did thesearise? Some values obtained by the laboratory werestated to have been deliberately set low, e.g., in therange 80-130 U/L, to rule out myocardial infarction(this point will be discussed later). Other values, in thesame range, were claimed to have been taken from thekit insert, but the accompanying insert did not supportthese claims and no effort had been made to validatesuch values locally. Of the 11 values >250 UIL, twowere borderline (253 U/L) and the remainder wereclaimed to be based on kit insert values, only some ofwhich could be verified by the LFI’P Committee fromthe attached inserts. The majority of the “rogue” valuesseen in the plots by location (Figure 2) derive from those

already mentioned; we assume that future efforts tostandardize reference ranges within a community willremove these anomalies. The differences between differ-ent laboratories in the same location are of considerableconcern. For example, in one community, a total CKactivity of 220 U/L is abnormal in three laboratories butnormal in another three. (A similar set of circumstanceswas found for the seven laboratories reporting anoma-lous ranges for women-these will therefore not bediscussed further.)

We also asked for details of the process by whichlaboratories obtained their own reference ranges. Severalreported that their records were missing, or staff hadchanged, so that details could not be provided other thanthe actual ranges themselves. Record keeping of suchdata obviously requires much more attention from hos-pital authorities and regulatory or accreditation organi-zations. Of the 134 laboratories who obtained their ownreference ranges (including the 41 laboratories that hadsimilar reference ranges for men and women), 83(61.9%)either provided histograms of their data showing theirselected cutoff values or explicitly stated their method ofderiving their reference range; the majority took a non-parametric approach and selected the central 95th per-centile range. Most laboratories observed a positivelyskewed distribution of values. Many of the laboratoriesstated the numbers of subjects used-ranging from 15 ina small community hospital to 1490 in bigger commercialand hospital laboratories, where data from several hos-pitals had been pooled. The population used usuallyincluded outpatients, staff, some inpatients, or groups ofsubjects undergoing insurance examinntions. Many lab-oratories attempted to have equal numbers of male andfemale subjects, but others did not state their sex ratio.One community hospital mentioned that, because manyof their subjects performed heavy manual labor, they hadto adjust the resulting reference range downwards to bemore representative of the general population this wasone of the few laboratories to appreciate the effect ofexercise on the CK values (15). One laboratory enclosed acopy of their preanalytical protocol for deriving a refer-ence range: volunteers were asked about their state ofhealth; the time of last meal and fluid intakes; anymedication use, including aspirin, sedatives, tranquiliz-ers, or birth control drugs; and the performance of vigor-ous exercise within the last 48 h. A small laboratory,with a correspondingly small operating budget, com-mented on the expense of deriving a reference range(because they were using an expensive immunoassay).Several respondents mentioned that obtaining a refer-ence range was beyond their capabilities: they werealready understaffed and simply did not have the time.Many laboratories, both hospital-based and commercial,pooled data and resources to obtain a more useful esti-mation of CK ranges in a locale (see, e.g., Figure 2C). Themost commonly cited literature reference used by labo-ratories as a source for their reference range was Tietz(16).

Many of the responses to the questionnaire indicatedthat a major use of the total CK reference limit is to rule

in, or out, myocardial infarction-although this usedemonstrates a naive understanding of the entire pro-cess. Such a process actually requires the careful priorselection of subjects (17), a process that is not usuallyperformed and one that produces decision thresholdsthat may differ widely from the conventionally derivedupper reference limit. For example, whereas a conven-tionally derived upper reference limit for men (based onresults for >200 healthy hospital stafi) was 174 U/L, thedecision thresholds for ruling in myocardial infarctionwere 250,400, and 800 U/L for test specificities of 90%,95%, and 99%, respectively, when an appropriatelymatched nondiseased population was used (18). By con-trast, the decision thresholds for ruling out myocardialinfarction, for test sensitivities of 90%, 95%, and 99%,were 40,30, and 20 U/L, respectively, 6 h after the onsetof chest pain (18).

CK-2

Table 3 shows the present distribution of CK-2 meth-ods in the Province, 82% of which are activity-basedassays. The data in Table 3 should be compared withrecent US evidence, i.e., a survey of the College ofAmerican Pathologists (19) in which 62% of the partic-ipants used mass assays, whereas immunoinhibition,electrophoresis, and ion-exchange (DuPont) assays eachaccounted for <15% of participants.

The CK-2 upper reference limits are shown in Figure3 for selected methods: there appear to be grounds for areview of some of these upper reference limits or forconsidering a change in methodology. When the 5th-95th percentiles are taken into account (see Resultssection), the mass assays clearly demonstrate a muchtighter distribution. These Ontario data are reflected inpilot studies conducted by this LPTP Committee andalso in the results of the survey already referred to (19).

About half of those laboratories that used kit insert orliterature CK-2 reference values validated them byer1nhining diagnostic records in consultation with clin-ical staff. However, such a process could introduce aselection bias, which might explain some of the ex-tremely high activities in Figure 3 (A and C).

To our knowledge, this is the first survey of thereference ranges in use in a single health-care admin-istrative unit, involving nearly 10 million people. Ourfindings are not reassuring. As a result of this survey,the LPTP Enzyme and Lipids Committee has formu-lated, and has proceeded to implement, the followingobjectives:

derivation of acceptable CK reference ranges by anylaboratory that (a) has ranges falling outside the 5th-95th percentiles (men: 160-250 U/L, women: 115-215U/L) established by this survey or (b) does not haveappropriate differences between men, women, and pedi-atric reference ranges (as a corollary, we wish to encour-

1370 CLINICAL CHEMISTRY, Vol. 38, No. 7, 1992

age those laboratories that derived their ranges from kitinserts or literature references to validate these rangeswithin a local setting).

#{149}standardization of CK reference ranges within acommunity.

#{149}exploration of educational initiatives for improvingknowledge of the essential elements of a satisfactory CKassay.

References1. Lohmnn RC, WoodDE, Jacobs WI, Elliott ML. Reliability ofwhite blood cell counting. Arch Pathol Lab Med 1989;13:989-94.2. McQueen MJ, Patton EL, Henderson AR, Krishnan 5, WoodDE, WebbS. Proficiency testing for creatune kunase isoenzymeCK-2 (CK-MB) in Ontario. Clin Chem 1990;36:2102-5.3. Moss DW, Henderson AR, Kachinar JF. Enzymes 1n TietzNW, ad. Textbook of clinical chemistry. Philadelphia WB Saun-ders Co., 1986:619-774.4. Solberg HE. Establishment and use of reference values. In:Tieta NW, ed. Textbook of clinical chemistry. Philadelphia: WBSaunders Co., 1986:356-86.5. Cherian AG, Hill JG. Age dependence of serum enzymaticactivities (lk*line phosphatase, aspartate aminotransferase, andcreatine kinase) in healthy children and adolescents. Am J ClinPathol 1978;70:783-9.6. Meites S. Creatune kinase. In: Meites 5, ed. Pediatric clinicalchemistry: reference (normal) values, 3rd ed. Washington, DC:AACC Press, 1989:103-8.7. H#{248}rderM, Elser RC, Gerhardt W, Mathieu M, Sampson EJ.Approved IFCC recommendation on methods for the measurementof catalytic concentration of enzymes: Part 7. IFCC method forcreatune kinase (ATP:creatine N-phosphotransferase, EC 2.7.3.2).Eur J Cliii Chem Cliii Biochem 1991;29:435-56.8. Nealon DA. Relative inhibition of human adenylate kinase andcreatine kinase isoenzymes by adenosine 5’-monophosphate anddiadenosune pentaphosphate. Clin Chem 1985;31:333-4.9. Miller WG,Chinchilli VM, Gruemer ND, Nance WE. Samplingfrom a skewed population distribution as exemplified by estima-tion of the creatine kinase upper reference limit. Clin Chem1984;30:18.-23.10. LottJA, Tholen DW, Masaion CG. Determination of referenceranges for serum enzymes via a large interlaboratory survey. ArchPathol Lab Med 1987;111:9-15.11. Meltzer 1W. Factors affecting serum creatine phosphokinaselevels in the general population: the role of race, activity and age.Clin Chim Acta 1971;33:165-72.12. Wong ET, Cobb C, Umehara MK, et a!. Heterogeneity ofserum creatune kinase activity among racial and gender groups ofthe population. Am J Clin Pathol 1983;79:582-6.13. Black HE, Quallich H, Gareleck CB. Racial differences inserum creatune kunase levels. Am J Med 1986;81:479-87.14. Swaminathan R, Ho CS, Donnan SPB. Body composition andplasma creatine kinase activity. Ann Clin Biochem 1988;25:389-91.15. Nicholson GA, McLeod JG, Morgan G, et al. Variable distri-butions of serum creatune kunase reference values: relationship toexercise activity. J Neurol Sci 1985;71:233-45.16. Tietz NW, ed. Clinical guide to laboratory tests, 2nd ed.Philadelphia: WB Saunders Co., 1990:1-931.17. RansohoffDF, Feinstein AR. Problems of spectrum and bias inevaluating the efficacy of diagnostic tests. N Engl J Med 1978;299:926-30.18. Leung FY, Galbraith LV, Jablonsky G, Henderson AR. Re-evaluation of the diagnostic utility of serum total creatine kinaseand creatine kinase-2 in myocardial infarction. Clin Chem 1989;35:1435-40.19. 1991 College of American Pathologists surveys. Electrophore-sia/chromatography survey set EC-A, specimen EC-02, April,1991. Skokie, IL: CAP, 1991.