HSR&D Home » Research » RRP 12-199 – HSR&D Study
Improving Anticoagulation Practices for Patients Presenting with ACS
Mary E Plomondon, PhD MSPH
Rocky Mountain Regional VA Medical Center, Aurora, CO
Funding Period: October 2012 - September 2013
Anticoagulant therapy for acute coronary syndrome (ACS) patients is recommended by clinical practice guidelines. Further, appropriate dosing of anticoagulant therapy is necessary to ensure effectiveness and safety and is a current ACC/AHA STEMI/NSTEMI test performance measure. This study describes the variability in excess dosing of unfractionated heparin (UFH), low-molecular weight heparin (LMWH), and glycoprotein IIb/IIIa inhibitors across VA hospitals and the association between patient and provider characteristics with excessive dosing.
The objectives of this study are to 1) describe the variability of anticoagulation dosing among patients presenting with ACS in the VA and assess the patient and provider factors associated with excessive dosing; 2) describe barriers/facilitators to appropriate anticoagulant dosing through qualitative interviews at sites that dose excessively and sites that dose per guideline recommendations.
A national retrospective cohort study of all veterans presenting with ACS at VA facilities from 2009-2011 was obtained through collaboration with the Inpatient Evaluation Center (IPEC). Inclusion criteria for our cohort were veterans who have been hospitalized with ACS, including myocardial infarction and unstable angina, from 2009-2011 using ICD-9 codes 410 and 411. We excluded patients who were missing anticoagulant dosing, weight and renal function measures.
Continuous variables are reported as medians or means and interquartile ranges or standard deviations. Categorical variables are reported as percentages. Significance testing included Chi-squared tests for categorical variables and wilcoxon rank sum tests or t-tests for continuous variables.
The patient and provider factors were tested for association with excessive dosing using hierarchical logistic regression, adjusting for clustering at the hospital levels. Because the primary goal was to identify factors associated with excessive dosing, a dichotomous outcome (excessive dosing versus not excessive dosing) was used. The analysis adjusted for the factors known to be associated with excessive dosing and for clustering at the site level. SAS 9.3 was used for all analyses.
For the qualitative analysis, individual telephone interviews were conducted with cardiologists and in-patient pharmacists to describe the barriers and facilitators to appropriate anticoagulation dosing at select VA hospitals. The sites were chosen based on the quantitative results which identified sites who appeared to consistently dose appropriately and those who did not consistently dose per guideline recommendations. The interview guide used in this inquiry consisted of 14 semi-structured, open-ended questions to explore the anticoagulation processes at each site. This study used an iterative, inductive and deductive toolkit of analytical strategies, drawing primarily on content analysis methodology. To ensure analytical rigor, our qualitative analyst began with repeated readings to achieve immersion followed by initial open coding using an emergent rather than an a priori approach to emphasize respondents' perspectives and de-emphasize researchers' assumptions. Subsequent analyses utilized a deductive approach based on the literature. As patterns, relationships, and themes emerged and were analyzed, they were reviewed by members of the multidisciplinary research team in order to assess the evocativeness, thoroughness, and comprehensiveness of the findings. To organize the coding process computer software Atlas.ti (Atlas.ti Scientific Software Development GmbH) was utilized.
This was a national cohort study of 53,489 patients admitted with ACS at 135 VA hospitals between FY09-FY11 and administered anticoagulation therapy during their hospital stay. Specifically we assessed the variability of anticoagulation dosing among patients and the patient and provider factors associated with excessive dosing. Excess dosing was defined based on the ACC/AHA test performance measures: unfractionated heparin > 70U/kg or infusion >15U/kg per hour, low molecular weight heparin (LMWH) >1.05mg/hr and glycoprotein IIb/IIa inhibitors, epifibitide: >180ug/kg and infusion of >2.0 ug/kg per minute among patients with creatinine clearance <50ml/min; tirofiban >0.4ug/kg and infusion of >0.1ug/kg among patients with creatinine clearance <30ml/min.
Dosing information was available on 37,633(75%) patients. Of these patients, 60.5% (22756) received unfractionated heparin, 37.0% (13926) received low molecular weight heparin and 2.5% (951) received glycoprotein IIb/III. The average percentage of patients receiving excess dose of any anticoagulant across hospitals was 7.2. By therapy, the mean excess unfractionated heparin dose averaged 2.1%, low molecular weight heparin 12.9% and glycoprotein IIb\IIIa 5.3%. Excessive dosing among hospitals ranged from 0% to 30% of patients.
Characteristics associated with excess dosing included age, body mass index (BMI), creatinine, location of dose administration and person administering the medication. Patients with older age and higher BMI were at lower risk of having excess dose. Patients with higher creatinine were at lower risk of having excess dose when the therapy was unfractionated heparin and LMWH and at higher risk of having excess dose of GIIB/IIIA. Among the provider and facility characteristics, patients who were first administered anticoagulant in ICU and administered therapy by a resident (rather than emergency department physician, specialist, nurse) were more likely to receive an excess dose.
Given the wide range of excessive dosing across hospitals (0%-30%), the qualitative interviews focused on a sample of sites where appropriate dosing was most consistently performed and a sample of sites where appropriate dosing was least consistently performed. We spoke with pharmacists and/or cardiologists at 2 high performing sites and 4 low performing sites. Sites that more often dosed appropriately had ACS-specific dosing algorithms and order sets, whereas sites where appropriate dosing was least consistent relied on individuals to order and follow-up on the labs and dosing. Comments from sites with algorithms included, "our algorithm is set such that the follow-up is automatically set when they order the drug". The sites without algorithms discussed using "mostly education" for dosing ACS patients and they felt that, in the setting of ACS, dosing is difficult and "more specialized education" is needed.
All sites recognized the physician as the decision maker for the type of anticoagulant and dose of anticoagulant. Sites with algorithms commented on the how the process still must rely on physician availability. If the physician is not available when the first labs come back, "it is a matter of finding somebody to report the value to and the time it takes for that individual to actually follow-up on that value". Further, if a resident is responsible for dosing and they have not been trained where the algorithm/order set resides in the electronic medical record, they are unable to use it to dose the patients. Sites without algorithms commented on residents using tools brought from the university (e.g., notes, handbooks) to help them choose the type and dose of anticoagulant.
When we asked sites for suggestions for improvement the comments from all sites, those with and without algorithms, focused on continuity of care; for instance, have one person responsible to order, wait for labs, and adjust the dosing. Sites suggested having more pharmacy or nursing resources to allow one person to be dedicated to consider all the necessary factors that must be considered when dosing anticoagulants for ACS patients.
In summary, using quantitative analyses we identified high/low performing sites to target our qualitative interviews and found sites that more often dosed within the recommend range were more often the sites that have ACS-specific dosing algorithms and order sets. Some sites without algorithms saw there was a need commenting, "it [algorithm] was discussed as a future option" or "it [algorithm] is on our perpetual to-do list".
There is wide hospital variability in excess dosing of anticoagulants for patients treated for ACS. The qualitative interviews found that ACS-dosing algorithms and orders sets were helpful toward more consistently dosing within the recommended ranges. Suggested next steps include surveying all VA sites to identify their use and training of ACS-specific dosing algorithms with the goal of implementing quality improvement programs to improve appropriate anticoagulant dosing among ACS patients nationally.
External Links for this Project
Dimensions for VADimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.
If you have VA-Intranet access, click here for more information vaww.hsrd.research.va.gov/dimensions/
VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project
DRA: Cardiovascular Disease
DRE: Treatment - Observational
MeSH Terms: none