The Direct Medical Costs Associated with Suspected and Confirmed Cases of Heparin-Induced Thrombocytopenia.

Introduction: Heparin-induced thrombocytopenia (HIT) is a serious adverse reaction associated with heparin exposure. Few studies have assessed the economic burden associated with HIT and none have assessed the costs of suspected HIT. The objective of our study was to identify and quantify the direct medical costs associated with suspected and confirmed cases of HIT from the Canadian hospital perspective.

Methods: Over the last 7 years SHSC, a tertiary care adult academic hospital, experienced on average 125 cases of suspected HIT per year. A retrospective burden of illness analysis was conducted to examine the costs associated with suspected and confirmed cases of HIT at SHSC in 2005. Suspected HIT was defined by the performance of a HIT enzyme linked immunosorbent assay (ELISA) from January 1, 2005 to December 31, 2005. Confirmed HIT cases were defined as those who had a positive serotonin release assay (SRA) or a positive HIT ELISA plus a high clinical probability for HIT. In addition, HIT-related resource utilization variables were assessed by chart review and included:

1. inpatient medication use;

2. laboratory and imaging tests; and

3. duration of hospital stay.

Physician visits and post-discharge costs were not included in the cost analysis. The average cost per case of confirmed HIT, confirmed HIT with thrombosis (HITT), and negative HIT was calculated in 2007 Canadian dollars. Cost information was obtained from multiple sources, including the pharmacy database (WORx©) and Ontario Health Insurance Program (OHIP) Schedule of Benefits. For this analysis, we did not assume that length of stay (LOS) was prolonged by HIT unless the patient was readmitted with a HIT-related complication. The costs attributable to prolongation of LOS due to HIT or its treatment are being determined, although it is difficult to allocate some costs to the presence of HIT alone. The cost for a negative HIT case only includes the cost of inpatient medication use, laboratory and imaging tests.

Results: There were 110 suspected HIT cases in 2005. We excluded 2 patients because their HIT status could not be determined after careful review. There were 88 HIT negative cases, 8 cases with confirmed HIT and 12 with HITT (Table). In this conservative analysis, patients with HITT incurred substantially greater costs than those with HIT. The average attributable cost of managing a HIT negative case was $116. Cardiovascular surgery (CVS) patients accounted for 48% of the suspected HIT cases and 65% of the confirmed cases. This represents 5% and 1% of the 960 CVS patients seen over the year.

Conclusions: This is the first study to identify and quantify the direct medical costs associated both with confirmed HIT and HITT and also negative HIT cases. Clearly, if LOS is prolonged by HIT, as has been shown by others, the costs per case will escalate. Cardiovascular surgery cases are being analyzed separately.