Background:
We are a 29 member academic hospitalist group at a large urban/suburban academic tertiary care center. Like other health systems, we are under tremendous pressure to decrease length of stay (LOS) and excess days (EDs). While patients with LOS greater than 15 days comprise only 5.52% of our total admissions, they contribute to greater than half of our total EDs. In the past, a retrospective chart review on these patients failed to elucidate any actionable items for change. A more detailed and innovative prospective survey was proposed in the hopes that real‐time attending physician input might better identify areas amenable to intervention.
Purpose:
To prospectively identify factors that contribute to long LOS/ED, which would otherwise be difficult to obtain in a retrospective chart review.
Description:
We designed a 30 item questionnaire including patient demographics, living situations, insurance, diagnoses and characteristics such as psychiatric issues, intensive care utilization and use of palliative care services. Admitted patients with a LOS greater than 10 days were identified using the electronic medical record. We made the assumption that patients admitted 10 days were likely to stay beyond 15 days. The hospitalists caring for these patients were interviewed weekly using the questionnaire until the patient was discharged. For the 50 patients surveyed, the average LOS was 21.9±10.9 days. The average age was 58.5 years (range 18‐93). The most important factors identified that contributed to long LOS were psychiatric problems (38%), infections (48%), critical care utilization (30%) and malignancies (22%). The majority of infections were not hospital acquired but out of the 10 that were, 5 were due to Clostridium difficile, 2 to urinary catheters and 2 to line infections. A palliative care consultation was obtained in 34% of cases and the average LOS in this group was extended by 2.5 days. Most patients (48%) were discharged home. 18% of the patients were felt to have family members that contributed adversely to the treatment or disposition plan. Social issues such as lack of social support or insurance were not found to be contributing factors.
Conclusions:
A prospective study to evaluate long LOS can improve the quantity and quality of data obtained, particularly as it pertains to a patient’s psychiatric/behavioral disorders and the role of family members. Our analysis suggests that we may have been overemphasizing factors such as type of insurance and social situation. For the future, the group will look at ways to identify patients with psychiatric diagnoses and challenging families earlier in the admission, and intervene using a multidisciplinary approach that involves psychiatry and social work. Further work needs to be done to streamline treatment of infectious diseases. Efforts will be made to engage our intensivist colleagues to start thinking about disposition even while the patients are still under their care.