Pharmacy & Therapeutics - February 2008 - (Page 101)

Antifibrinolytic Therapy in Cardiothoracic Surgery None 6,000 5,000 No. of cases 4,000 3,000 2,000 1,000 0 4 2002 1 2 3 2003 4 1 2 3 4 1 2 2005 3 AP AA variates. Subsequently, we included significant covariates from the single variable screen in a stepwise multiple logistic regression analysis that evaluated the association between treatment group and outcomes. The Hosmer–Lemeshow test was used to measure goodness of fit for each model.21 We performed statistical comparisons and data management using SAS version 9.1.3 (Cary, N.C.) running on a Windows XP platform. The significance level was set a priori at 0.05 (two-tailed) and was not adjusted for multiple comparisons during the logistic regression modeling. 2004 Year–quarter Figure 2. Use of fibrinolytic agents in patients from 20 hospitals undergoing cardiothoracic surgery over three years. AA = aminocaproic acid; AP = aprotinin. Data involving medication use included the dates of administration for the three antifibrinolytic agents (aprotinin, AA, and TA) and for angiotensin-converting enzyme (ACE)–inhibitors with their dates of administration. We attempted to capture the antifibrinolytic dose; however, we considered this information from some participating hospitals to be unreliable and did not pursue the effort further. Other outcomes of interest included: • the in-hospital mortality rate. • the length of stay in the hospital and intensive-care unit (ICU). • a diagnosis of acute renal failure (ICD-9-CM code 584.x). • postoperative stroke (ICD-9-CM secondary diagnostic codes 433.x1, 434.x1, and 997.02). • postoperative acute myocardial infarction (MI) (ICD-9CM secondary diagnostic code 410.xx). • the need for blood transfusion (ICD-9-CM procedure code 99.0x) or hemodialysis code (39.95). RESULTS Descriptive statistics for the various groups are summarized in Table 2. We observed considerable differences in average age, mortality, average ICU length of stay, and severity scores. Quarterly analyses of antifibrinolytic therapy for CTS showed a gradual decline in the use of AA, with a gradual increase in the use of aprotinin over the three years and a relatively stable non-use of either drug (Figure 2). We observed considerable variations among individual hospitals in their use of any antifibrinolytic option (range, 0–50%) (Figure 3). To better characterize the link between antifibrinolytic use and various treatment outcomes, we controlled for comorbidities and concurrent drug use that might be associated with the treatment outcomes of interest. Simple logistic regression was used in their selection. Influential variables selected during the initial screen were then included in the full model. Using stepwise multiple logistic regression, we controlled for demographic variables (age, sex, and race) and comorbidity variables (ACE-inhibitors, diaAP AA Statistical Analysis We compared baseline continuous demographic and medical characteristics using parametric analysis of variance. We used Bonferroni’s multiple comparison test to compare any subsequent pairwise baseline continuous demographic variables and Fisher’s Exact test to compare baseline categorical demographic and medical characteristics for all treatment groups and for any subsequent pairwise comparisons. Using multiple logistic regression analysis, we performed the association-of-treatment group (aprotinin, AA, or no therapy) on primary and secondary dichotomous outcomes, controlling for covariates. Initial comparisons were made across treatment groups, followed by pairwise, between-treatment group comparisons. Using simple logistic regression, we performed an initial single variable screen of co100% 80% 60% 40% 20% 0 None I SCOQDPNBTGJMEHFKLRA Organization code Figure 3 Percentage of antifibrinolytic drug use in patients undergoing cardiothoracic surgery by organization over three years. AA = aminocaproic acid; AP = aprotinin. Vol. 33 No. 2 • February 2008 • P&T® 101

Table of Contents Feed for the Digital Edition of Pharmacy & Therapeutics - February 2008

Pharmacy & Therapeutics - February 2008 Contents Editorial Medication Errors Prescription: Washington The Language of (Forgive Us) Change, As P&T Enters the Digital Age New Drugs/Drug News/New Medical Devices Drug Forecast Use and Outcomes of Antifibrinolytic Therapy in Patients Undergoing Cardiothoracic Surgery at 20 Academic Medical Centers in the United States Evaluation of the Management of Acute Venous Thromboembolism and Its Outcomes: One Institution's Experience American Society of Hematology, 49th Annual Meeting Pharmaceutical Approval Update

Pharmacy & Therapeutics - February 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.