Pharmacist-provided anticoagulation management in United States hospitals: Death rates, length of stay, medicare charges, bleeding complications, and transfusions

We explored the associations between pharmacist-provided anticoagulation management in hospitalized Medicare patients and several major heath care outcomes: death rate, length of stay, Medicare charges, bleeding complications, and transfusions. Using the 1995 National Clinical Pharmacy Services database and the 1995 Medicare database for hospitals, data were retrieved for 717,396 Medicare patients treated in 955 hospitals for conditions requiring anticoagulant therapy. In hospitals without pharmacist-provided heparin management, death rates were 11.41% higher (X-2 (1) = 122.84, p<0.0001), length of stay was 10.05% higher (Mann-Whitney U test = 40039529342, p<0.0001), Medicare charges were 6.60% higher (U = 41004749266, p<0.0001), bleeding complications were 3.1% higher (X-2 (1) = 10.996, p=0.0009) and the transfusion rate for bleeding complications was 5.47% higher (X-2 (1) = 11.24, p=0.0008) than in hospitals with pharmacist-provided heparin management. In hospitals without pharmacist-provided warfarin management, death rates were 6.20% higher (X-2 (1) = 19.20, p<0.0001), length of stay was 5.86% higher (U = 25730993838, p<0.0001), Medicare charges were 2.16% higher (U = 259955112970, p<0.0001), bleeding complications were 8.09% higher (X-2 (1) = 49.259, p<0.0001), and the transfusion rate for bleeding complications was 22.49% higher (X-2 (1) = 78.68, p<0.0001). Study hospitals without pharmacist-provided heparin management had 4664 more deaths, 494,855 more patient-days, 145 more patients with bleeding complications, and $651,274,844 more in patient charges; 9784 more units of whole blood were used in patients requiring transfusions for bleeding complications. Hospitals without pharmacist-provided warfarin management had 2786 more deaths, 316,589 more patient-days, 429 more patients with bleeding complications, and $234,275,490 more in patient charges; 8991 more units of whole blood were used in patients requiring transfusions for bleeding complications. The implications of these findings are significant for the health care system, especially considering that the study population represents 28.25% of hospitalized Medicare patients who should receive anticoagulants, and that total Medicare admissions represent 35.02% of total admissions to United States hospitals.