The universal definition of myocardial infarction provides criteria for the diagnosis and classification of the type of acute myocardial infarction (AMI) and recommends the measurement of cardiac troponin (cTn) for the detection of myocardial necrosis. Despite this, many institutions still use creatinine kinase-MB as part of the standard laboratory evaluation for AMI. The aim of this study was to ascertain the cost benefit of using only cTn for the diagnosis of acute coronary syndromes. A survey was done of Kansas City hospitals and the University HealthSystem Consortium list server for comparison of whether other institutions use 1 or both biomarkers. Locally, most of the responding hospitals in Kansas City and 57% of responding university hospitals surveyed use the 2 biomarkers for the diagnosis of AMI. After initiating medical staff education, the investigators’ institution’s standard order sets were changed to use only cTnI for the assessment of AMI. The cost savings by eliminating creatine kinase-MB on the order sets was based on the annual volume of processed specimens. The cost of supplies, equipment, and materials was determined on a per test basis. After initiation of the new standard order sets, the creatine kinase-MB testing rate decreased to <50 per month, from 1,098, for an annualized savings of $86,786. In conclusion, as the health care system is challenged to improve care while containing cost, the use of a single biomarker (cTn) in the diagnosis of AMI fulfills this promise without sacrificing patient care.
The diagnostic value of cardiac troponin (cTn) measurement is well established. However, less well appreciated is the potential economic benefit from using a single marker instead of a panel of biomarkers that traditionally have included creatine kinase-MB (CK-MB). Expert consensus underscores using cTn determinations only for better clinical care and a reduction in cost. The elimination of multiple biomarkers, including CK-MB, in the diagnosis of acute coronary syndromes (ACS) relies on evidence-based medicine consistent with the universal definition of acute myocardial infarction (AMI). We report our experience in cost savings after developing standard order sets with cTnI as the only marker on chest pain, ACS, and post–percutaneous coronary intervention orders. Our hypothesis was that a significant cost savings would be realized without compromising patient care.
Methods
Truman Medical Center is an inner-city, university-affiliated, 278-bed teaching hospital in Kansas City, Missouri. In January 2009, the Section of Cardiology within the Department of Internal Medicine elected to adopt only cTnI as the standard of measurement for AMI. The standard order sets for all chest pain, ACS, and post–percutaneous coronary intervention orders used only cTnI for the assessment of AMI, and concomitant CK-MB laboratory tests were removed. The ordering physician could add CK-MB if so desired.
The cost savings by eliminating the CK-MB on the order sets was based on the annual volume of 13,176 processed specimens. The cost of supplies, equipment, and materials was determined to be $7.05 per test, thus incurring an annual expense of $92,891. Although the activities of specimen management (processing, results reporting, and quality control) require human resources and might result in further cost, we did not factor this into our analysis.
Modification of physician practice and acceptance of the standard order sets was facilitated by lectures and informal presentations at student, resident, and staff conferences starting 6 months before the orders were changed. A survey was done of area hospitals for their use of biomarkers in the diagnosis of AMI. In addition, the University HealthSystem Consortium list server was used for comparison, and 8 of the 14 responding hospitals use both biomarkers.
Results
Locally (not including Truman Medical Centers), 75% of the responding hospitals in Kansas City use cTnI and CK-MB for the diagnosis of AMI, and 57% of the responding university hospitals in the University HealthSystem Consortium list server use the 2 markers. Figure 1 shows the acceptance of the order sets, with a rapid decrease in the number of CK-MB tests ordered per month. The number continued to decrease over the 12 months, with an average of only 46 tests per month in the last 6 months.
