Objective: To assess the effects of clinical pharmacist participation in the care of critically ill Medicare patients with thromboembolic or infarction-related events (TIE) on clinical and economic outcomes.
Methods: In this retrospective database review (September 1, 2004-August 31, 2005), patient data were retrieved from the 2004 Expanded Modified Medicare Provider Analysis and Review database. Outcomes data evaluated included mortality rates, length of intensive care unit (ICU) stay, total Medicare charges, drug and laboratory charges, and rates of bleeding complications. In addition, outcomes related to the bleeding complications (transfusions, mortality rate) were assessed. Patient outcomes in ICUs with clinical pharmacy services were compared with patient outcomes in ICUs without these services. Clinical pharmacy services were defined as direct patient care services provided by a pharmacist specifically devoted to the ICU; other services such as order processing or drug distribution were not part of these services. A description of ICU pharmacy services was obtained from a 2004 national survey.
Results: We identified 141,079 patients with TIE, of whom 7987 also had bleeding complications. In hospitals with ICU clinical pharmacy services, mortality rates in patients with TIE only and TIE with bleeding complications were higher by 37% (odds ratio [OR] 1.41, 95% confidence interval [CI] 1.36-1.46) and 31% (OR 1.35, 95% CI 1.13-1.61), respectively, than in ICUs with clinical pharmacy services. Lengths of ICU stay were longer by 14.8% (mean +/- SD 7.28 +/- 8.17 vs 6.34 +/- 7.80 days, p<0.0001) and 15.8% (12.4 +/- 13.28 vs 10.71 +/- 9.53 days, p=0.008), respectively. The lack of clinical pharmacist participation in a patient's care was associated with extra Medicare charges of $215,397,354 (p<0.001) and $63,175,725 (p<0.0001) and extra drug charges of $26,363,674 (p<0.0001) and $2,610,750 (p<0.001) for TIE only and TIE with bleeding complications, respectively. Without clinical pharmacy services, bleeding complications increased by 49% (OR 1.53, 95% CI 1.46-1.60), resulting in 39% more patients requiring transfusions (OR 1.47, 95% CI 1.28-1.69); these patients also received more blood products (mean +/- SD 6.8 +/- 10.4 vs 3.1 +/- 2.6 units/patient, p=0.006).
Conclusion: Involving clinical pharmacists in the direct care of intensive care patients with TIE was associated with reduced mortality, improved clinical and charge outcomes, and fewer bleeding complications. Hospitals should promote direct involvement of pharmacists in the care of patients in the ICU.