We present findings from a 2001 survey of U.S. hospital
coagulation laboratories. (See the Appendix for a copy of the actual survey
used.) We sent this questionnaire to 800 hospital laboratories, stratified
into large (> 200 beds) and small (<200 beds)
hospitals. Study sample was a random selection of the large and small
hospitals from the 1999 directory of the American Hospital Association
(AHA). The selected large and small hospitals constituted 26% and 9% of the
large and small hospitals listed in this directory, respectively. We
administered this survey and collected results between June and October
2001. Respondents had the option of mailing a completed survey or submitting
one via the Internet. We received 632 responses (corresponding to a response
rate of 79% including 20 responses submitted electronically).
Our findings show great variability in certain laboratory
practices. Although in most cases, response patterns from the large and
small hospital laboratories were not significantly (P > 0.050)
different, several questions solicited significantly different responses
from these 2 groups. For most of these questions, a greater proportion of
large hospitals adhered to published laboratory practice recommendations and
guidelines. The following is a summary of major findings.
Performance of Coagulation Testing
Ninety-seven percent of the respondents performed
Test Requisition and Specimen Management
Test requisition. The respondents noted the
following usage information on their test requisition forms: coumadin, 53%;
unfractionated heparin, 39%; heparinoid, 33%; low molecular weight heparin (LMWH),
23%; and salicylate, 16%. The large hospital respondents noted significantly
more requests for information on requisition forms for using unfractionated
heparin (P = 0.003), heparinoid (P = 0.044) and LMWH (P
Rejection of specimens. The proportions of the
respondents below did not note the following reasons for rejecting a
coagulation specimen in their laboratories: specimen collected via
indwelling catheter, 68%; label not having hospital medical record number,
55%; specimen stored at an inappropriate temperature, 15%; specimen
hemolyzed, 14%; requisition form and specimen label having conflicting
patient information, 8%; and specimen transport time exceeding recommended
time frame, 8%. The College of American Pathologists (CAP) has recommended
that specimens used for monitoring heparin therapy be collected from a
different extremity than the one used for heparin infusion (Arch Pathol
Lab Med. 1998;122:782798). A significantly greater proportion of the
large hospital respondents (59%) rejected coagulation specimens because of
the lack of hospital medical record number compared to the small hospital
respondents (31%); P < 0.001. These results suggest a need for
improvement in certain laboratory practices relating to test requisition and
specimen collection procedures.
Practices Relating to Prothrombin Time (PT) Assay
Anticoagulant concentration. Based on the
recommendation of the World Health Organization (WHO) and the NCCLS
guidelines, 3.2% (109 mmol/L) citrate is the anticoagulant of choice for
coagulation testing (Arch Pathol Lab Med. 1998;122:768781). Eighty
percent of the large hospital respondents stated exclusively using 3.2%
sodium citrate as the anticoagulant compared to 66% of the small hospital
respondents (P < 0.001).
Reporting of results. Reporting PT results in
seconds may lead clinicians to inappropriately compare results between
institutions (Am J Clin Pathol. 1998;109:589594) and relying on PT
ratio has been documented to cause errors in anticoagulant therapy (Arch
Intern Med. 1992;152:278282). Almost all the respondents (99.8%) used
international normalized ratio (INR) to report PT; however, 97% also
reported PT in seconds and/or as therapeutic PT ratio. Three percent of the
respondents reported PT results in INR only.
Reference interval. Ninety-two percent of the
respondents conducted in-house evaluations to establish the reference
interval for their PT assay. Most respondents (46% of the large hospitals
and 74% of the small hospitals, P < 0.001) used less than 40 subjects
to establish their PT reference intervals. According to the NCCLS, a minimum
of 120 subjects for each reference population or subclass is recommended to
establish reference intervals for quantitative laboratory tests (NCCLS
approved guideline2nd edition. Document C28-A2. Vol 15; No.
4). Five percent of the respondents noted using at least 120 subjects to
establish their reference ranges for PT assay.
Sensitivity of PT assay to heparin. According to the
CAP, laboratories should determine sensitivity of their PT
assays to heparin (Arch Pathol Lab Med. 1998;122:782798).
Seventeen percent of the respondents determined the sensitivity of their PT
assays to heparin. Also according to the CAP guideline, laboratories should, where possible, select a
thromboplastin that is insensitive to heparin in the therapeutic range (Arch
Pathol Lab Med. 1998;122:768781). Fifty-nine percent of the large
hospital respondents selected a PT-thromboplastin reagent that was
insensitive to heparin in the heparin therapeutic range, compared to 40% of
the small hospital respondents (P < 0.001).
The CAP recommends that thromboplastins with a manual ISI
between 0.9 and 1.7 be used (Arch Pathol Lab Med. 1998;122:768781).
The large hospital respondents reported an average ISI of 1.52 (median,
1.56) while the small hospitals reported an average of 1.70 (median, 1.89).
Of the large hospital respondents, 50% reported ISI values of
< 1.70 compared to 36% of the small hospital
respondents (P = 0.001).
Forty-two percent of the large hospital
respondents reported ISI values of < 1.20
compared to 24% of the small hospital respondents (P < 0.001) as
recommended by the American College of Chest Physicians (Chest.
Practices Relating to Activated Partial Thromboplastin
Time (aPTT) Assay
Therapeutic range. According to the CAP guideline,
adjusted dose and therapeutic heparin require anticoagulant monitoring using
a method with a defined therapeutic range (Arch Pathol Lab Med.
1998;122:782798). Seventy-three percent of the large hospital respondents
noted having an aPTT therapeutic range for heparin compared to 53% of the
small hospital respondents (P < 0.001). While 64% of the respondents
stated they reported the aPTT therapeutic range for heparin when monitoring
heparin therapy, 9% included the corresponding heparin concentration with
How the aPTT therapeutic range for heparin was determined.
The CAP recommends that therapeutic range of unfractionated heparin for the
aPTT reagent-instrument system should be determined with each change in
reagent (lot number or manufacturer) or instrument (Arch Pathol Lab Med.
1998;122:782-798). This may be accomplished by (1) comparison of ex vivo
specimens with an appropriately validated heparin assay or (2) comparison of
ex vivo specimens to a previously calibrated aPTT using a method to
control for reagent drift. The respondents adhered to the following
practices to determine the aPTT therapeutic range for heparin:
- using samples from patients on heparin therapy to compare a new to an
old reagent lot, 59% (66% of the large hospital respondents versus 50% of
the small hospital respondents, P = 0.007);
- using heparin-spiked samples to compare a new to an old reagent lot,
- using heparin-spiked samples to compare a new to an old heparin lot,
15% (12% of the large hospital respondents versus 21% of the small
hospital respondents, P = 0.038);
- using samples from patients on heparin therapy to compare a new to an
old heparin lot, 12%;
- performing anti-Xa assay, 37% (47% of the large hospital respondents
versus 18% of the small hospital respondents, P < 0.001); and
- performing protamine sulfate titration, 9%.
When the aPTT therapeutic range for heparin was reconfirmed.
The respondents reconfirmed the aPTT therapeutic range for heparin under the
- when new instrumentation is used, 79%;
- when new reagent lots are used, 75%;
- when new reagents are used, 51%; and
- after a specified time period, 22%.
Specimen management. According to the NCCLS, samples
can be assayed up to 4 hours after phlebotomy if centrifuged within 1 hour
of collection (NCCLS approved guideline3rd edition.
Document H21-A3. Vol 18; No. 20). The respondents indicated they adhered to
the following practices to manage specimens before aPTT analysis:
- specimens assayed within 4 hours after phlebotomy, 96%;
- specimens centrifuged within 1 hour of collection, 88% (84% of the
large hospital respondents versus 92% of the small hospital respondents, P
- specimens kept at room temperature prior to testing, 82%; and
- specimens kept at 4 C prior to testing, 22%.
Practices Relating to Assays for von Willebrand Disease (vWD)
Performance of von Willebrand factor (vWF) assays.
Ten percent of the large hospital respondents stated that they provided
results for von Willebrand factor antigen (vWF Ag) compared to 0.4% of the
small hospital respondents (P < 0.001). Fourteen percent of the large
hospital respondents noted that they provided results for von Willebrand
factor activity (Ristocetin cofactor activity) compared to 0.3% of the small
hospital respondents (P < 0.001). Finally, 3% of the large hospital
respondents stated that they provided results for vWF multimers compared to
0.4% of the small hospital respondents (P = 0.007). The following
proportions of the respondents performed the 3 vWF assays:
- 38% for vWF Ag and vWF activity,
- 25% for vWF activity only,
- 15% for vWF Ag, vWF activity and vWF multimers,
- 15% for vWF Ag only, and
- 6% for vWF multimers only.
Reporting of ABO specific reference interval for vWF antigen
assay. Nineteen percent of the respondents that performed vWF Ag
assay reported an ABO specific reference interval for this assay.
Provision of vWF multimers results. Eighty-nine
percent of the respondents noted that they performed vWF multimers assay
only when ordered by a clinician; 38% did so when Ristocetin cofactor was
decreased; 29% performed this assay when Ristocetin cofactor was
disproportionately decreased relative to vWF Ag; 25% did so when vWF Ag and
vWF activity were both low; and 13% did so only if Ristocetin-induced
platelet aggregation indicated a Type II B vWD.
Practices Relating to Thrombosis/Hypercoagulability Workup
Protein S assays. Ten percent of the large hospital
respondents usually performed the assay for protein S activity (functional
test) before the antigenic assay compared to 0.3% of the small hospital
respondents (P < 0.001). If the results of the functional test were
decreased, 17% performed antigenic assay to differentiate Type I deficiency
from Type II while 20% performed free and total protein S antigen assay.
Performance of activated protein C (APC) resistance and
factor V Leiden mutation assays. Eleven percent of the large hospital
respondents performed activated protein C (APC) resistance assay compared to
1% of the small hospital respondents (P < 0.001). If, after
performing the APC resistance assay, results indicated resistance to APC,
61% obtained results for factor V Leiden mutation. Long-term anticoagulation
in carriers of factor V Leiden, on the basis of the carrier state alone, is
not indicated (Blood. 1997;89:19631967).
Algorithm for Diagnosing a Lupus Anticoagulant (LA)
Offering an LA profile. Eighteen percent of the
respondents stated that they offered an LA profile.
Practices leading to mixing studies. When a PT
result was prolonged, 37% of the respondents did not offer mixing studies
for PT, and 56% did so only if there was an additional order for the mixing
study. When an aPTT result was prolonged, 33% of the respondents did not
offer mixing studies for aPTT, and 59% did so only if there was an
additional order for the mixing study.
Workup to diagnose an LA. If the results of the
mixing study for aPTT did not correct to normal, 17% of the respondents
initiated a workup to diagnose an LA. Those routinely initiating a workup to
diagnose an LA most commonly performed the following tests:
- dilute Russell viper venom time, 79%;
- hexagonal phase phospholipid (Staclot LA) assay, 51%;
- lupus sensitive aPTT, 40%; and
- platelet neutralization procedure, 35%.
Practices Relating to Monitoring for Low Molecular Weight
Heparin (LMWH) Therapy
Monitoring of LMWH therapy. Fourteen percent of the
respondents (19% of the large and 10% of the small hospitals, P =
0.002) noted that they monitored LMWH therapy.
Assays used. The CAP recommends the chromogenic
antifactor Xa method for monitoring LMWH (Arch Pathol Lab Med.
1998;122:799807). In this survey, those monitoring LMWH therapy did so most
commonly by using aPTT (72%) and anti-Xa (53%) assays. While 65% of the
large hospital respondents used an anti-Xa assay to monitor for LMWH
therapy, 18% of of the small hospital respondents did so (P = 0.001).
Fifty-eight percent of the large hospital respondents used an aPTT assay to
monitor LMWH therapy compared to 96% of the small hospital respondents (P
= 0.001). The observation that 17% of the large hospital respondents
reportedly performed in-house anti-Xa assay compared to 2% of the small
hospital respondents (P < 0.001) may help to explain these findings.
Calibrators used. According to the CAP guideline, a
hospital pharmacy should dispense heparin of a single manufacturer and lot
number (Arch Pathol Lab Med. 1998;122:782798). The most common
calibrators for anti-Xa assay were reportedly
- LMWH supplied by pharmacy, 53%;
- internal standard LMWH, 22%;
- unfractionated heparin, 14%; and
- internal standard unfractionated heparin, 11%.
The CAP recommends that a calibrated LMWH be used to
establish the standard curve for an assay to measure LMWH and that
unfractionated heparin not be used to establish the standard curve for
monitoring LMWH (Arch Pathol Lab Med. 1998;122:799807). While 74% of
the respondents used different calibration curves for LMWH and
unfractionated heparin, 42% did so for each type of LMWH.
Timing of anti-Xa assay. The CAP recommends that
when LMWH is monitored, the sample be obtained 4 hours after subcutaneous
injection (Arch Pathol Lab Med. 1998;122:799807). Forty-six percent
of the respondents did not recommend a time for anti-Xa testing after
subcutaneous administration of LMWH, 32% performed anti-Xa testing 4 hours
after injection, and 14% did so between 2 and 4 hours after injection.
Availability of Specific Coagulation Tests
The top 8 most commonly performed coagulation tests were PT
(100%), aPTT (99%), bleeding time (90%), fibrinogen (78%), D-dimer (66%),
fibrin(ogen) degradation products (52%), activated clotting time (43%), and
thrombin time (38%). We found no significant differences between the large
and small hospital respondents in terms of the proportions performing
in-house testing of the 3 most commonly assayed coagulation tests: PT, aPTT
and bleeding time. Except for plasminogen antigen assay (performed by 2
large hospitals), a significantly greater proportion of the large hospital
respondents performed all other tests in-house compared to the small
hospital respondents (P < 0.001, P = 0.035 for vWF multimers
with 5 large hospitals performing this assay).
Test Result Information, Interpretations and
From 90% to 98% of the respondents provided measurement
units and 7687% provided needed specimen comments for PT, aPTT, vWF Ag and
protein C assays. From 93 to 97% of the respondents supplied reference
intervals for these assays. The following proportions of respondents
provided therapeutic ranges: PT, 54%; aPTT, 38%; aWF Ag, 5%; and protein C,
From 4 to 6% of the respondents noted that they specified
testing methodology/reagent on coagulation test reports for these 4 tests.
One percent noted possible drug interactions for PT and aPTT assays; 7% did
so for vWF Ag assay; and 21% stated drug interactions in reporting protein C
Two percent of the respondents suggested diagnoses for PT
and aPTT assays compared to 10% and 12% for protein C and vWF Ag assays,
respectively. The proportions providing written interpretations were as
follows: aPTT, 4%; PT, 6%; vWF Ag, 21%; and protein C, 22%.
While 2% of the respondents provided recommendations for
further testing for PT and aPTT assays, 12% and 14% did so for vWF Ag and
protein C assays, respectively. The proportions providing recommendation for
treatment were as follows: PT, 1%; aPTT, 1%; protein C, 3%; and vWF Ag, 5%.
The proportions providing recommendation to test family members were as
follows: PT, 0.2%; aPTT, 0.2%; protein C, 8%; and vWF Ag, 10%.
From 29% to 33% provided no result interpretation for any of
these 4 tests, and 4651% provided no testing/treatment recommendations.
Process of Reporting Results
Reporting of critical values. Ninety-nine percent of
the respondents reported critical values. Of those noting that they reported
critical values, the respondents adhered to the following practices:
- critical values telephoned to the clinician and the call documented,
- critical values repeated and documented as confirmed, 91%;
- critical values telephoned to the clinician and the call not always
documented, 6%; and
- critical values indicated on the report, but no further action taken,
Repeating a coagulation test. Circumstances under
which a coagulation test was usually repeated were reportedly as follows:
- control(s) out of range, 98%;
- results outside instrument technical ranges, 98%;
- results being critical values, 95%;
- results not agreeing with previous results, 73%; and
- results outside of the reference interval, 16%.
These data suggest a need for improved reporting of
laboratory test results.
Quality Assurance (QA) Procedures
Respondents usually took the following QA steps:
- critical values brought to immediate attention of the clinician, 99%;
- calibration of all instruments periodically verified, 99%;
- critical values reviewed, 99%;
- new analytical methods validated, 98%;
- patient information on specimen tube and laboratory-generated labels
- specimen label and requisition form matched, 90%;
- instrument printout compared to reported value, 82%;
- patients previous results checked, 76%;
- specimens run in duplicate, 39%;
- controls run in duplicate, 38%; and
- plasma checked for platelet count after centrifugation, 23%.
According to the Clinical Laboratory Improvement Amendments
of 1988 (CLIA) regulations, a laboratory report must be sent promptly to the
authorized person, the individual responsible for using the test results or
the laboratory that initially requested the test (CLIA Subpart K Sec.
493.1253 Condition: Hematology.
One percent of the respondents did not adhere to this practice (
Patient and control specimens must be tested in duplicate
for manual coagulation tests; duplicate testing is not required for
automated coagulation tests (CLIA
Subpart K Sec. 493.1253 Condition: Hematology). Patient and control
specimens were reportedly run in duplicate by 39% and 38% of the
respondents, respectively. We did not ask whether this duplicate testing was
performed using manual or automated methods.
Two percent of the respondents reportedly did not validate
new analytical methods (P < 0.001); and 1% of the respondents
indicated that they did not periodically verify calibration of all of their
instruments (P = 0.003).
These data suggest a need for improvement in performing
certain QA procedures.
Testing location. Respondents performed coagulation
tests in the following locations: core laboratory, 55%; hematology
laboratory, 38%; coagulation laboratory, 16%; point of care, 11%; and stat
laboratory, 5%. Two percent noted that coagulation testing was done at none
of these locations.
Number of full time equivalents (FTEs). Seventy-one
percent employed less than 4 FTEs to perform coagulation testing, while 17%
employed 49 FTEs and 12% employed > 10 FTEs.
Components of competency assessment program. The
respondents included the following components in their competency assessment
program for coagulation testing personnel:
Educational degrees of coagulation laboratory director.
Ninety-one percent of the respondents noted that the laboratory director had
an M.D., while 7% stated that this individual had a Ph.D. Both M.D. and
Ph.D. degrees were noted by 4% of the respondents. Eight percent of the
respondents noted other degrees, and 27% of this group also noted that the
laboratory director possessed an M.D.
Certifications of coagulation laboratory director.
The proportions of the laboratory directors with specific professional
board/society certifications were as follows:
Coagulation service capacities. A clinician was
reportedly available for consultation having expertise in coagulation
disorders in 57% of the responding hospitals (74% of the large versus 38% of
the small hospitals, P < 0.001). An anticoagulation outpatient clinic
specializing in the adjustment of oral anticoagulants reportedly existed at
the institutions of 20% of the respondents (27% of the large versus 12% of
the small hospitals, P < 0.001). Nine percent of the respondents
stated that they had an outpatient clinic specializing in the diagnosis and
treatment of coagulation disorders (17% of the large versus 2% of the small
hospitals, P < 0.001).
Availability of POCT for PT assay. Nine percent of
the respondents (15% of the large versus 3% of the small hospitals, P
< 0.001) had POCT for PT assay.
Laboratory oversight of coagulation POCT. The
laboratory reported having oversight of coagulation POCT (including
certification and regulatory compliance) in 93% of cases (98% of the large
versus 67% of the small hospital respondents, P = 0.001).
Location of coagulation POCT. The respondents noted
the following sites for performing coagulation POCT: coagulation clinic,
64%; cardiac catheterization laboratory, 27%; satellite laboratory, 23%;
operating rooms, 21%; bedside, 18%; dialysis clinic, 13%; and other sites,
Integration of POCT results. Forty percent of the
respondents stated that coagulation POCT results were integrated into the
laboratorys results reporting system. Of these, 95% noted that coagulation
POCT results were integrated into the laboratorys reporting system in the
order of collection time.
Reference interval for POCT of PT. Reference
interval for the POCT PT assay was reported to be the same as the laboratory
PT reference interval for 45% of the respondents. Of the remaining, 36%
noted that the POCT reference interval was established by the same method
used to establish the PT reference interval for the laboratory.
Type of quality control (QC) material/method.
Seventy-four percent of the respondents used electronic QC, 52% employed
lyophilized QC material, and 44% used liquid QC material. Sixty-seven
percent used 2 or more types of QC. Those using 2 or more types of QC
reported they used electronic QC methods along with liquid QC material
(33%), lyophilized QC material (30%), or both (4%).
Frequency of QC runs. According to the CLIA
regulations, for all non-manual and CLIA non-waived coagulation testing
systems, the laboratory must include 2 levels of control each 8 hours of
operation, i.e. once per an 8-hour shift, and each time a change in reagents
occurs (CLIA Subpart K. Sec. 493.1253 Condition: Hematology.
Thirty-nine percent of the respondents performed QC once per shift, while
54% did so once per day. Ninety-three percent of the respondents performed
QC either once per day or once per shift.
Our findings show substantial variability exists in certain
coagulation laboratory practices. To our knowledge, this is currently the
only report of a broad and comprehensive survey of nationally based
coagulation-specific and general laboratory practices in hospitals. We
believe these results present a representative and accurate snapshot of
hospital coagulation laboratory practices in the United States in 2001.