Validation of a pneumonia prognostic index using the MedisGroups comparative hospital database

摘要: Abstract Purpose: Our purpose was to validate a previously developed pneumonia-specific prognostic index in large, multicenter population. Patients and methods: We prospective, study of 346 patients with clinical radiographic evidence pneumonia admitted 3 Pittsburgh hospitals (the derivation cohort), validated the 14,199 principal ICD-9-CM diagnosis 78 1989 Medis-Groups Comparative Hospital Database validation cohort). The classified into five ordered risk classes based on six predictors mortality: age greater than 65 years, pleuritic chest pain, vital sign abnormality, altered mental status, neoplastic disease, high-risk etiology. Each patient cohort assigned class by obtaining values for index's MedisGroups performance cohorts assessed comparing hospital mortality rates within each classes. Results: rate 13.0% cohort, 11.1% (p = 0.26). agreement class-specific striking exception V: I, 0% versus 1% cohort; II, 1.1%; III, 10.9% 8.6%; IV, 21.8% 26.2%; V, 73.7% 37.7%. There were no statistically significant differences first four classes, which represented vast majority (94%) as well (98%) cohorts. Conclusion: These data support generalizability index. This index, performs exceptionally classifying low-risk patients, may help physicians identify community-acquired who could safely be managed ambulatory setting, or if hospitalized, that treated abbreviated inpatient care.