|
|
 |
|
|
|
Posted 09/29/2005
Kathleen M. LaBeau, MT(ASCP); Shahram Shahangian, PhD, DABCC, FACB
Abstract and Introduction
Abstract
-
In October 2003, the Washington State Department of Health and the Centers for
Disease Control and Prevention used prothrombin time testing to develop a model
to reduce medical errors by identifying steps that are vulnerable to errors and
comparing current laboratory practices with voluntary practice standards.
-
Laboratory personnel can use the study results and references to voluntary
practice standards to reduce their opportunities for error for prothrombin time
testing.
-
Professional societies can use this model to systematically assess other
error-prone tests to both establish and harmonize best practices among
laboratories.
Introduction
Medical mistakes and errors are unacceptably high, despite a longstanding focus
on activities carried out in the name of quality assurance, quality
improvement, total quality management, and quality assessment. In 1999, a
report by the Institute of Medicine revealed the magnitude of medical errors
and concluded that most were the result of systematic failures and were
preventable.[1]
One approach to reducing serious medical errors is by identifying quality
indicators and developing systems for best practices. Practice standards and
guidelines are developed through a consensus process that identifies specific
essential requirements for materials, methods, and practices. They are designed
to both establish and harmonize best practices among the health care community.
However, studies have shown that despite required and voluntary standards of
practice, many laboratory professionals fail to use them.[2]
|
Purpose of the Study
In October 2003, the Washington State Office of Laboratory Quality Assurance
(LQA) and the Centers for Disease Control and Prevention (CDC) created a model
to collect and monitor laboratory quality indicators from a broad spectrum of
clinical laboratories.
To develop our model, we established the following objectives:
-
Select a common, error-prone laboratory test.
-
Identify the steps that are vulnerable to errors.
-
Investigate voluntary practice standards to determine best practices.
-
Gather information about current practices from a variety of testing sites.
-
Share the findings.
-
Recognize inherent differences between testing settings and methods.
-
Make recommendations about quality indicators and best practices.
We selected the prothrombin time (PT) test to develop this model since it is a
very common test that is vulnerable to errors and adverse patient outcomes.
Patients on oral anticoagulation therapy must be monitored carefully to prevent
dangerous complications of bleeding or thrombosis.
|
Methods
To gather information about current laboratory testing practices, a
questionnaire was developed in October to December 2003 by the LQA and CDC, and
was pilot-tested in 5 laboratories in Washington State in December 2003. We
researched numerous voluntary practice standards addressing PT testing that
served as the basis for our questionnaire. Questions were developed to address
the areas we identified to be vulnerable to errors for PT testing. These
included:
-
Selection of the thromboplastin reagent.
-
Concentration of the anticoagulant in collection tubes.
-
Specimen acceptance and rejection policies.
-
Implementation of new lots of reagents.
-
Contents of patient test report to clinicians.
Using the Washington Medical Test Site (MTS) data base and licensure
application forms, testing sites performing PT by either waived or non-waived
test complexity methods were identified and targeted to receive the
questionnaire. Laboratories located in Alaska, Idaho, and Oregon, performing
proficiency testing for PT, were identified using the CLIA (OSCAR) data base.
Questionnaires were mailed to 591 laboratories in the Pacific Northwest region
on January 27, 2004.
|
Results
Respondents
Two hundred ninety-seven completed questionnaires were returned by March 19,
2004, resulting in an overall response rate of 50% (
Table 1 ).
We further categorized respondents according to the test reagent/system used.
Seventy percent indicated they used a reagent associated with a traditional PT
test method and 30% indicated the use of reagent test strips or cartridges
associated with point of care (POC) testing devices (
Table 2 ).
A wide variety of backgrounds for testing personnel were given. Different
patterns were seen in testing personnel between the sites using traditional
test methods and those using POC devices (
Table 3 ).
Selection of the Thromboplastin Reagent
When using the PT test to monitor oral anticoagulation therapy, the sensitivity
of the thromboplastin reagent to the depletion of vitamin K dependent
coagulation factors is reflected as the international sensitivity index (ISI).
All thromboplastins are calibrated against standards with sensitivities
comparable to the WHO International Reference Plasma, which is assigned an ISI
of 1. Commercial manufacturers of thromboplastin reagents calculate the ISI and
include it in the product package insert.
Several voluntary practice standards and other publications recommend the use
of thromboplastin reagents that have a low ISI value.[3-7]
Thromboplastins with low ISIs are more responsive or "sensitive." The
variability of international normalized ratio (INR) values produced by
different test systems is reduced by the universal use of highly responsive
reagents. While sensitive thromboplastin reagents with lower ISI values may
offer the potential for improved precision in determining the INR [due to the
fact that INR = (PT ratio)ISI where PT ratio = patient PT/mean
normal PT], some studies have suggested that low-ISI reagents may be less
precise.[6]
The following recommendations have been made for the selection of reagent ISI
values:
| |
• |
0.9 to 1.70 |
College of American Pathologists[3] |
| |
• |
<1.50 |
NCCLS[4]
(Clinical and Laboratory Standards Institute)
|
| |
• |
Close to 1.00 |
American Society of Health System Pharmacists[5] |
| |
• |
<1.20
|
Hirsch[6] |
| |
• |
<1.50 |
American Heart Association[7] |
For sites using traditional methods (
Table 4 ), the range of ISI values for their reagents was 0.85 to 2.33,
with an average ISI of 1.34 and median ISI of 1.15.
Table 5 shows a frequency distribution of the reagents used according
to ISI values.
Sites using POC devices and the ISI values of their reagents are summarized in
Table 6.
Concentration of the Anticoagulant in Collection Tubes
Several voluntary practice standards or guidelines recommend the use of
collection tubes containing sodium citrate in the concentration of 3.2%.[3,8-10]
The INR can be affected by the citrate concentration. The specimen osmolarity
is closer to plasma and decreases the variability in clotting times related to
the variability in the hematocrits and filling volumes of the tubes when using
3.2% sodium citrate. Many of the manufacturers determine their ISI values using
3.2% citrate and the same citrate concentration should be used in individual
laboratories. Low ISI reagents yield higher INR values when under-filled
samples are collected in 3.8% citrate.
Participants were asked if they collected samples for PT by venipuncture. Of
the 260 respondents to this question, 216 (83%) indicated they did. The
majority of respondents (92%) used only the recommended citrate concentration
of 3.2%. Four percent used a concentration of 3.8%, and 1% used both 3.2% and
3.8% sodium citrate concentrations.
Specimen Acceptance and Rejection Policies
There are several voluntary practice standards that address the proper
collection and handling of specimens for coagulation testing and PT testing in
particular.[4,8,11]
Of the 216 respondents that collected samples by venipuncture, 202 (94%) said
they had a written policy addressing specimen acceptability and rejection for
PT testing. Participants were given a list of issues that are commonly
recommended for inclusion in specimen acceptability and rejection policies for
coagulation testing. They were asked to acknowledge those they included in
their written policy (
Table 7 ).
It should be noted that depending on the setting or the methodology, some of
these issues may not apply. For example, the collection of samples from patient
lines and heparinized specimens may be applicable for patients in hospitals but
not for most patients in outpatient settings. Specimens that are icteric or
lipemic may affect test methods based on optical clot detection but may not be
a concern for mechanical clot detection methodologies.
Of the 259 respondents, 85 (33%) stated that they collected samples by finger
stick or capillary collection. Of those, 87% indicated that they had a written
policy addressing the proper collection of capillary specimens for PT testing.
Implementing New Lots of Reagents
Various practice standards address issues associated with implementing new lots
of testing reagents. Some address general activities such as establishing or
verifying patient reference ranges and mean of normal, and some are specific
for handling new lots of thromboplastin reagents.[3-5,11,12]
Given a list of 8 indicators of quality practices associated with the
evaluation of new lots of thromboplastin reagents for PT testing, participants
were asked which they performed (
Table 8 ).
Contents of Patient Test Report to Clinicians
The ISTH and the WHO recommend that reporting of PT results for patients on
oral anticoagulation therapy include the use of INR values.[10,13]
Other practice standards and publications suggest this as well.[3,4,6,11]
Given a list of choices, participants were asked which test values and other
information they provide in the patient report to clinicians. Nearly all
respondents reported the INR value (
Table 9 ).
|
Discussion
Adherence to the standards we studied was relatively high for sites performing
traditional PT methods. For those collecting venipuncture samples, 92% used the
generally recommended concentration of sodium citrate, and 94% had a written
specimen acceptance/rejection policy. The majority of these sites used a
reagent with an ISI of <1.70 (76%), verified their reference range
(92%), and established their mean of normal (95%) for new lots of
thromboplastin reagents. For sites using POC devices, specimens were primarily
obtained by capillary collection methods. Therefore, issues related to
collection tubes and transport, processing, and storage of samples are not
applicable in those cases. Respondents using POC devices relied more on
information provided by the manufacturer for reference ranges and mean of
normal values, rather than establish their own. Nearly all testing sites
reported INR values.
Because clinicians compare INR values against standardized therapeutic ranges
and monitor trends in an individual patient's INR values over time, consistency
in test values from an individual laboratory and agreement in values between
different laboratories are issues of key importance. Errors can occur when a
laboratory changes to a new lot of testing reagents. Testing personnel may not
recognize that their reagent sensitivity has changed and may not do studies to
verify their test results are consistent and calculations are accurate.
Therefore, personnel should adhere to the following best practices when
introducing new lots of reagents, test strips, or cartridges:
-
Verify the ISI value in the product insert.
-
Establish their own patient mean of normal using the new reagent.
-
Perform parallel testing between the old and new lots.
-
Assure that ISI value and patient mean of normal value are correctly entered
into their instrument and laboratory information system where the INR
calculation occurs.
Errors can also occur when a patient moves from one setting to another due to a
lack of correlation between methods. Agreement of test results between
laboratories can be improved when personnel follow these best practices:
-
Use reagents with low ISI values.
-
Use specimen collection tubes with 3.2% sodium citrate.
-
Report INRs.
As part of this study, we also determined if laboratory testing personnel used
voluntary practice standards to develop their PT testing policies and
procedures. We found that a minority of respondents used voluntary practice
standards, and that the most common reason given for not using standards was a
"lack of awareness." By publishing this report, sharing it with study
participants, and posting it on the CDC Web site, we hope to raise awareness of
recommended standards of practice and references that may help to harmonize
practices among all sites performing PT testing. It is also hoped that testing
personnel may investigate and adopt new practices based on a comparison to
their peers.
To review the results of the entire study, go to
http://www.phppo.cdc.gov/MLP/SurveyReports/Prothrombin_2004.aspx.
Acknowledgements
Assistance in on-site data collection was done by Susan Walker, MA, MT(ASCP),
Gary Utter, DrPH, MT (ASCP), Lori Hudson, MT (ASCP), and Leonard Kargacin, MA,
CLS (NCA). Thanks to the Office of Laboratory Quality Assurance Washington
State Department of Health.
Funding Information
Funding for this study was provided by Centers for Disease Control and
Prevention, Cooperative Agreement U10/CCU023393-01.
Disclaimer
Use of trade names are for identification purposes only, and it does not imply
any endorsement by the United States Department of Health and Human Services or
the Centers for Disease Control and Prevention.
References
-
Committee on Quality of Health Care in America, Institute of
Medicine. To Err is Human: Building a Safer Health System. Washington,
DC: National Academy Press, 2000.
-
Shahangian S, Stankovic AK, Lubin IM, et al. Results of a
survey of hospital coagulation laboratories in the United States, 2001. Arch
Pathol Lab Med. 2005;129:47-60.
-
Fairweather RB, Ansell J, van der Besselaar AM, et al. College
of American Pathologists Conference XXXI on Laboratory Monitoring of Oral
Anticoagulation therapy: laboratory monitoring of oral anticoagulation therapy. Arch
Pathol Lab Med. 1998;122:768-781.
-
NCCLS. One-Stage Prothrombin Time (PT) Test and Activated
Partial Thromboplastin Time (APTT) Test; Approved Guideline. NCCLS
document H47-A; 1996.
-
American Society of Health System Pharmacists. ASHP
therapeutic position statement on the use of the international normalized ratio
system to monitor oral anticoagulant therapy. Am J Health Syst Pharm. 1995;52:529-531.
-
Hirsh J, Dalen JE, Deykin D, et al. Oral anticoagulants:
mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest.
1995;108(Supplement):231S-246S.
-
Ansell J, Hirsh J, Wenger NK. Management of Oral Anticoagulant
Therapy: Principles and Practice. American Heart Association,
Postgraduate Education Committee, Council on Clinical Cardiology; 1999.
-
NCCLS. Collection, Transport, and Processing of Blood Specimens
for Coagulation Testing and General Performance of Coagulation Assays; Approved
Guideline-Third Edition. NCCLS document H21-A3; 1998.
-
Hassett AC. Optimizing the INR to improve anticoagulant
monitoring: Transfusion Medicine Update. The Institute for Transfusion
Medicine; 1998.
-
Loeliger EA. International Committee for Standardization in
Haematology/International Committee for Thrombosis and Haemostasis
recommendations for reporting prothrombin time in oral anticoagulation control. Thromb
Haemost. 1985;54:155-156.
-
The Ontario Association of Medical Laboratories, Quality
Assurance and Clinical Laboratory Practice Committee. CLP 014 Guidelines for the
Laboratory Monitoring of Oral Anticoagulants; 1997.
-
NCCLS. How to Define and Determine Reference Intervals in the
Clinical Laboratory; Approved Guideline-Second Edition. NCCLS document
C28-A2; 2000.
-
World Health Organization Expert Committee on Biological
Standardization. 28th Report: WHO Technical Report Series 610, Geneva,
Switzerland: WHO; 1977.
|
This page last reviewed: 10/24/2005
|
|