T. R. Grenda1,2, M. R. Hemmila1,2, S. L. Wong1,2, A. Mikhail2, S. E. Regenbogen1,2 1University Of Michigan,Center for Healthcare Outcomes And Policy,Ann Arbor, MI, USA 2University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA
Introduction: As healthcare reimbursement reform has increasingly penalized hospitals for unplanned readmissions, there is widespread interest in developing interventions to prevent them. In a high-volume colorectal surgery service, we designed, implemented, and evaluated a pre-discharge screening program aimed at preventing readmission following inpatient colorectal surgery.
Methods: We composed a 10-item screening tool to identify patients at increased risk for postoperative readmission. At discharge, mid-level providers or residents completed the screening and identified patients received a follow-up phone call from clinic nursing staff 48-72 hours after discharge, to identify and redirect patients with problems to early outpatient attention. We obtained data on comorbidities and outcomes from supplemental review of the electronic medical record. Statistical analysis was performed to compare early (<7 days) and 30-day readmission rates between patients with positive and negative screens, and among those with positive screens, between those who did and did not receive follow-up phone calls.
Results: 290 consecutive patients undergoing colorectal surgery were screened for readmission risk. 193 of these patients (66.5%) screened positive using the tool (Table 1). The 30-day readmission rate was 12.4% for patients screening positive and 3.1% for those screening negative (relative risk 4.0, p=0.009). The screening tool had a sensitivity of 91% for early readmission and 88% for 30-day readmission. The positive predictive value of the tool was 5.6% and 12.3% for early and 30-day readmission, respectively. Of those patients screening positive, only 52% were successfully contacted by nursing staff for follow-up phone call. There were no significant differences in readmission rates, at either 7 days (phone call: 3.9% vs. no phone call: 7.7%, p=0.4) or 30 days (11.8% vs. 13.2%, p=0.8) associated with receiving an intervention phone call. Issues screened for during the phone call did not predict subsequent readmission.
Conclusion: Our study identifies a 10 question tool with high sensitivity for detecting patients at highest risk for readmission after colorectal surgery. However, a targeted early follow up phone call intervention did not appear to prevent readmissions. Future efforts aimed at understanding the specific factors predictive of readmission are needed to guide implementation of effective interventions to prevent postoperative readmissions.
