Does hospitalist discontinuity during an inpatient stay affect cost, readmissions, or patient satisfaction?
Hospitalist discontinuity may lead to an increase in total hospital costs. The authors of this study postulate that this could be due to additional testing or treatments or new clinical issues identified by hospitalist physicians when they take over care from their colleagues. (LOE = 2b-)
|Study Design: Cohort (retrospective)||Funding: Other|
|Setting: Inpatient (any location) with outpatient follow-up||Allocation: Uncertain|
To assess for physician discontinuity during a hospitalization, these authors used 2 measures: (1) the Number of Physicians Index (NPI), which accounts for the total number of unique hospitalist physicians caring for a patient, and (2) the Usual Provider of Continuity Index (UPC), which calculates the largest number of patient encounters by a single hospitalist physician divided by the total number of encounters. A higher NPI and a lower UPC suggest greater discontinuity. All patients admitted to the hospitalist service during a 3.5-year period, except for those transferred from other services such as the intensive care unit, were included in the study. A physician was defined as caring for a patient if he or she wrote an admission note or progress note during the hospitalization. Nocturnists were not included in this calculation. Outcomes assessed in this study were total costs of hospitalization (excluding physician charges), 30-day readmissions, and patient satisfaction with physician communication using patients’ responses to three questions from the Hospital Consumer Assessment of Healthcare Providers and Systems survey. A positive outcome for patient satisfaction was defined as top box responses to all three questions. The baseline cohort consisted of approximately 18,000 hospitalizations with a mean NPI of 1.9 and a mean UPC of 0.75. Median hospital cost was $6300 per hospitalization and the 30-day readmission rate was 22%. Only 8.5% of patients answered all 3 of the patient satisfaction questions and of these, 63% had top box responses for all 3 questions. After adjusting for covariates, including demographic variables, case mix, Diagnosis Related Group (DRG) weight, and length of stay, a 1-unit increase in NPI was associated with a 0.9% higher median cost (P < .001). Alternative models categorizing DRG weight and length of stay into quartiles also showed higher costs with a 1-unit increase in NPI (8% cost increase; P < .001) and with a 0.1-unit decrease in UPC (12.6% cost increase; P < .001). For the readmissions outcome, although a 0.1-unit decrease in UPC was associated with a small decrease in 30-day readmissions (odds ratio = 0.97; 95% CI, 0.95-0.99), the other 3 models did not corroborate this finding. No statistically significant correlation was found between hospitalist discontinuity and patient satisfaction; however, the study may have been underpowered to detect this given the low number of respondents to the patient satisfaction survey.
Nita Shrikant Kulkarni, MD
Assistant Professor in Hospital Medicine