Length of stay and hospital costs among patients admitted to hospital by family physicians

Objective To compare length of stay and total hospital costs among patients admitted to hospital under the care of family physicians who were their usual health care providers in the community (group A) and patients admitted to the same inpatient service under the care of family physicians who were not their usual health care providers (group B). Design Retrospective observational study. Setting A large urban hospital in Vancouver, BC. Participants All adult admissions to the family practice inpatient service between April 1, 2006, and June 30, 2008. Main outcome measures Ratio of length of stay to expected length of stay and total hospital costs per resource intensity weight unit. Multivariate linear regression was performed to determine the effect of admitting group (group A vs group B) on the natural logarithm transformations of the outcomes. Results The median acute length of stay was 8.0 days (interquartile range [IQR] 4.0 to 13.0 days) for group A admissions and 8.0 days (IQR 4.0 to 15.0 days) for group B admissions. The median (IQR) total hospital costs were $6498 ($4035 to $11 313) for group A admissions and $6798 ($4040 to $12 713) for group B admissions. After adjustment for patient characteristics, patients admitted to hospital under the care of their own family physicians did not significantly differ in terms of acute length of stay to expected length of stay ratio (percent change 0.6%, P = .942) or total hospital costs per resource intensity weight unit (percent change -2.0%, P = .722) compared with patients admitted under the care of other family physicians. Conclusion These findings suggest that having networks of family physicians involved in hospital care for patients is not less efficient than having family physicians provide care for their own patients.