Transcript Post-Intervention - Anticoagulation Centers of Excellence

Abstract # 49
Category C1
Collaboration between inpatient
anticoagulation service and hospital reference
laboratory to reduce misdiagnosis of
Heparin Induced Thrombocytopenia
Harmony Bowles, PharmD
PGY1 Pharmacy Resident
University of New Mexico Hospital
IRB Approval Received
Disclosure Statement
• Harmony Bowles
• Investigators have no conflict of interest to
disclose
• This study was not funded
• Proprietary information or results may be subject
to different interpretations
• Speaker’s presentation of this slide indicates
agreement to abide by the non-commercialism
guidelines provided in the CE Requirements page
Learning Objectives
• Objective #1:
– Describe the role of a multidisciplinary collaboration in
reducing overdiagnosis of HIT
• Objective #2:
– List financial and clinical benefits of having a
multidisciplinary team involved in diagnosis and
management of HIT
Target Audience: Pharmacists
• Level 1 trauma center
• 646 bed tertiary care teaching hospital
• ~50 full-time pharmacists
• Pharmacy-driven inpatient anticoagulation
service seven days a week
Pre-Test Assessment Questions
• T/F: The use of a validated pre-test probability tool known as
the 4T’s score has been shown to improve the process of
identifying patients at low risk for HIT.
• T/F: Collaboration between inpatient anticoagulation service
and the reference laboratory will lead to a reduction in the
number of inappropriate HIT antibody ELISA tests
performed.
• T/F: Multidisciplinary collaboration leads to cost savings by
decreasing the number of patients being inappropriately
tested and treated for HIT.
Background
• Heparin Induced Thrombocytopenia (HIT) is a prothrombotic, immune- mediated adverse reaction to heparin
or low-molecular weight heparin (LMWH) therapy1
• Accurate diagnosis requires:
– Patient clinical presentation
– Laboratory results
• Most institutions utilize an enzyme-immunoassay (ELISA)
to detect HIT antibodies (Ab)
– High sensitivity: >90% negative predictive value
– Low specificity: high potential for false positives2
1. Arepally G, Ortel T. N Engl J Med 2006;355:809-17.
2. Lo G, Sigouin C, Warkentin T. Am J Heamatol. 2007;82:1043-1043.
Background
• Hospitalized patients have multiple reasons for developing
thrombocytopenia
• Inappropriately ordered non-specific ELISA Ab increases
potential for overdiagnosis of HIT
•
The 4T’s score3
– Validated assessment tool to identify patients at risk for HIT
– Based on 4 patient parameters:
• Magnitude of platelet drop
• Timing of thrombocytopenia in relation to heparin exposure
• Presence of new thrombosis
• Other possible reasons for thrombocytopenia
– High negative predictive value
3. Lo G, Juhl D, Warkentin T, et al. J Thromb Heamost 2006; 4:759-765.
Study Objectives
• Primary objective:
– Evaluate impact of multidisciplinary collaboration in
reducing number of inappropriate Ab tests performed,
thereby reducing the misdiagnosis of HIT
• Secondary objective:
– Evaluate potential cost savings of decreasing Ab testing
and use of direct thrombin inhibitors (DTIs) for
inappropriate treatment of HIT
Methods – Study Design
• Study design: Single center, pre-post
prospective study with retrospective chart review
• Inclusion criteria: All patients for whom Ab was
ordered during study period
– ≥ 18 years of age
– Sufficient data available to calculate 4T’s score
Methods – Intervention Outline
Methods – Data Collection
• At risk population
– Patients exposed to unfractionated heparin (UFH) or low
molecular weight heparin (LMWH) products
– Patients for whom an Ab tests ordered during pre- and postintervention period obtained from reference laboratory
• Patients with Ab ordered
– Manual chart review for patient demographics, characteristics,
pre-defined bleeding and thrombotic events
• 4T’s score independently and retrospectively calculated
by 3 evaluators to establish inter-rater reliability (IRR)
Methods – Data Collection
• Cost:
– Hospital cost for performing Ab test obtained from
reference laboratory
– Drug treatment cost per patient for those receiving
direct thrombin inhibitor (DTI) in pre-intervention
period evaluated by:
• DTI acquisition cost
• Number of patients treated
• Number of doses dispensed
Methods – Data Analysis
• Fleiss kappa performed to assess IRR
between 3 evaluators
• Chi-square used to analyze categorical
variables
• Student t-test used to analyze continuous
variables
• All data entered and analyzed with SPSS
software version 19.0
Results – IRR
• IRR
– Fleiss kappa = (0.643)
– Interpretation: Good IRR4
4. Fleiss J. Statistical Methods for Rates and Proportions, 2nd ed. New York: John Wiley and Sons; 1987.
Results – Study Population
At Risk Population
(# of patients)
PreIntervention
PostIntervention
p-value
Admitted to UNMH
19,312
19,422
-
Exposed to UFH/LMWH
6,867
7,809
NS
Patients with Ab ordered
(% of total exposed patients)
176
(2.5%)
108
(1.4%)
<0.001
*NS: No statistical difference
Results – Demographics
Patients with Ab ordered
Characteristic
Age, mean (SD), years
Male
*NS: No statistical difference
Pre-Intervention
(n=176)
Post-Intervention
(n=108)
56.9
(14.5)
56.4
(16.2)
NS
87 (49.4%)
55 (50.9%)
NS
p-value
Results – Ab Testing
Results – Ab Testing
n=127
n=127
n=74
n=33
Results – Cost
Patients with Ab ordered
Pre-Intervention
(n=176)
Post-Intervention
(n=108)
Patients treated for HIT
16
(9.1%)
9
(8.3%)
0.912
# of Ab tests cancelled
0
46
<0.001
p-value
Cost incurred
Hospital cost per Ab test
Hospital drug cost per patient treated for HIT
(Pre-intervention)
$35.00
$6400.00
Results – Cost Savings
Average $ per patient exposed
$18
$16
$15.92
$14
$12
$10
$7.66
$8
$6
$4
$2
$0
Pre
Post
Estimated annual cost savings=($15.92 – $7.66) x 7809= $64,502.34
Results – Adverse Events
Patients with Ab ordered
Pre(n=176)
Post(n=108)
Bleeding event
10.2%
6.5%
0.279
Thrombotic event
0.6%
0%
0.433
Adverse outcome
p-value
• No bleeding events directly associated with HIT treatment during
hospitalization
• Single thrombotic event in the pre-intervention period in patient with
negative HIT Ab result
Study Limitations
• Calculation of 4T’s score performed retrospectively
• Data collected from chart review dependent on accurate
documentation
• Patients could have received treatment for bleeding or
thrombosis events at another facility
• Total cost evaluation did not include pharmacist time to
perform intervention
• Limited external validity as not all hospitals have
dedicated inpatient anticoagulation service
Conclusions
• Impact of multidisciplinary collaboration
– Reduction in inappropriate HIT testing
– Reduction in misdiagnosis of HIT without causing
adverse clinical outcomes
• Use of the 4T’s score by a pharmacists is a means to
improving accuracy of HIT testing and diagnosis and
improving patient outcomes
• Substantial cost savings may be realized through
decreased lab testing, decreased DTI use and
avoidance of unnecessarily treating patients for HIT
Post-Test Assessment Questions
• T/F: The use of a validated pre-test probability tool known as
the 4T’s score has been shown to improve the process of
identifying patients at low risk for HIT
• T/F: Collaboration between inpatient anticoagulation service
and the reference laboratory will lead to a reduction in the
number of inappropriate HIT ELISA tests performed
• T/F: Multidisciplinary collaboration leads to cost savings by
decreasing the number of patients being inappropriately
treated for HIT
Acknowledgements
•
•
•
•
Allison Burnett, PharmD
Richard D’Angio, PharmD, BCPS
Matthew Borrego, Ph.D., M.S
Tiffany Montoya, PharmD