PURPOSE: To examine the associations between (1) the economic indicators of hospital costs, charges, and length of stay (LOS) for inpatient cases and (2) the use of MEDLINE searches for such cases. METHOD: An outcome-based, objective, prospective study with an economic evaluation was conducted from September 1989 to September 1990 at three metropolitan Detroit teaching hospitals representing both allopathic and osteopathic care. The study consisted of (1) 192 test cases, derived from a consecutive sample of inpatients of all ages for whom MEDLINE searches were requested at the participating medical libraries, and (2) 10,409 control cases, which were of the same diagnosis-related groups (DRGs) as the test cases but did not involve identified MEDLINE searches. Statistical analysis included the use of multivariate analyses of variance and correlation coefficients. Comparisons of cases were made on case-by-case and DRG bases regarding total patient costs, charges, and lengths of stay for cases with or without MEDLINE searches. RESULTS: The test cases were found to have a higher severity of illness. Among test cases, statistically significant relationships existed between (1) hospital expenses and LOS and (2) hospital expenses and the timing of the search during hospitalization when controlling for LOS. When cases were matched for DRG and LOS, the cases with early searches (i.e., conducted during the first half of hospitalization) had significantly lower expenses. CONCLUSION: Of the test-case patients (for whom MEDLINE searches were conducted during hospitalization), those whose searches were conducted earlier had statistically significantly lower costs, charges, and lengths of stay than those whose searches were conducted later.
PURPOSE: To examine the associations between (1) the economic indicators of hospital costs, charges, and length of stay (LOS) for inpatient cases and (2) the use of MEDLINE searches for such cases. METHOD: An outcome-based, objective, prospective study with an economic evaluation was conducted from September 1989 to September 1990 at three metropolitan Detroit teaching hospitals representing both allopathic and osteopathic care. The study consisted of (1) 192 test cases, derived from a consecutive sample of inpatients of all ages for whom MEDLINE searches were requested at the participating medical libraries, and (2) 10,409 control cases, which were of the same diagnosis-related groups (DRGs) as the test cases but did not involve identified MEDLINE searches. Statistical analysis included the use of multivariate analyses of variance and correlation coefficients. Comparisons of cases were made on case-by-case and DRG bases regarding total patient costs, charges, and lengths of stay for cases with or without MEDLINE searches. RESULTS: The test cases were found to have a higher severity of illness. Among test cases, statistically significant relationships existed between (1) hospital expenses and LOS and (2) hospital expenses and the timing of the search during hospitalization when controlling for LOS. When cases were matched for DRG and LOS, the cases with early searches (i.e., conducted during the first half of hospitalization) had significantly lower expenses. CONCLUSION: Of the test-case patients (for whom MEDLINE searches were conducted during hospitalization), those whose searches were conducted earlier had statistically significantly lower costs, charges, and lengths of stay than those whose searches were conducted later.