Objective: To determine whether the absence or presence of clinical pharmacists in intensive care units (ICUs) results in differences in mortality rates, length of ICU stay, and ICU charges for Medicare patients with nosocomial-acquired infections, community-acquired infections, and sepsis.
Design, setting, and patients: The type and level of pharmacy services provided to ICUs were obtained from a 2004 national survey. Clinical pharmacy services were defined as having at least a partial pharmacist full-time equivalent specifically devoted to the ICU for the purpose of direct involvement in patient care. Infections were defined using International Classification of Diseases, Ninth Revision, Clinical Modification codes. ICU outcome data were drawn from the 2004 modified Medicare provider analysis and review. Depending on the infection studied, the involvement of clinical pharmacists was evaluated in 8,927-54,042 patients from 265 to 276 hospitals.
Interventions: None.
Measurements and main outcomes: Mortality rates, length of ICU stay, Medicare charges, drug charges, and laboratory charges for each of the infections categorized according to the absence or presence of clinical pharmacists. Compared to ICUs with clinical pharmacists, mortality rates in ICUs that did not have clinical pharmacists were higher by 23.6% (p < 0.001, 386 extra deaths), 16.2% (p = 0.008, 74 extra deaths), and 4.8% (p = 0.008, 211 extra deaths) for nosocomial-acquired infections, community-acquired infections, and sepsis, respectively. Similarly, ICU length of stay was longer by 7.9% (p < 0.001, 14,248 extra days), 5.9% (p = 0.03, 2855 extra days), and 8.1% (p < 0.001, 19,215 extra days) for nosocomial-acquired infections, community-acquired infections, and sepsis, respectively. ICUs that did not have clinical pharmacists had greater total Medicare billings of 12% (p < 0.001, $132,978,807 extra billing charges), 11.9% (p < 0.001, $32,240,378 extra billing charges), and 12.9% (p < 0.001, $224,694,784 extra billing charges) for nosocomial-acquired infections, community-acquired infections, and sepsis, respectively. Similar findings were observed for Medicare drug charges and laboratory charges.
Conclusion: The involvement of clinical pharmacists in the care of critically ill Medicare patients with infections is associated with improved clinical and economic outcomes. Hospitals should consider employing clinical ICU pharmacists.