Figure 1.

Relative odds of mortality associated with statin use in the first 2 hospital days. *Adjusted for: age, gender, smoking, congestive heart failure, lymphoma, metastatic cancer, other neurologic disorders, obesity, pulmonary circulation disease, renal failure, solid tumor without metasasis, valvular disease, weight loss, depression, hypertension, psychoses, primary diagnosis, pneumonia type, initial antibiotic(s) received, early treatment (day 1 or 2) with angiotensin-converting enzyme inhibitor, angiotensin receptor blocker, Foley catheter, restraints, nutritional supplements, total parenteral nutrition, gastrostomy or jejunostomy tube, thiamine, calcium or mechanical ventilation. †Variables in propensity model: All variables above, plus admission from skilled nursing facility, insurance type, marital status, race/ethnicity, hospital geographic region, number of beds, teaching hospital, and setting (urban/rural), attending specialty, acquired immune sufficiency syndrome, alcohol abuse, deficiency anemia, collagen vascular disease, chronic blood loss anemia, chronic pulmonary disease, diabetes, drug abuse, hypothyroidism, ischemic heart disease, paralysis, peripheral vascular disease, peptic ulcer disease and bleeding, aspirin, bisphosphonates, clopidogrel, folic acid, glucosamine, multivitamin, vitamin B2, B5, B6, C, D or E, dementia medications, collagenase, prealbumin, psychotropic drugs, silvadene, antidepressants, beta blockers, calcium channel blocker, diuretics, ezetimibe, fibrates, amiodorone, inhaled bronchodialators, inhaled corticosteroids, insulin, immunosuppressants, niacin, nitroglycerin, non-steroidal anti-inflammatory drugs, proton pump inhibitors, resins, steroids, thyroid replacement therapy, and warfarin. ‡All stratified analyses adjusted for propensity score and co-variates.