Background: The Center for Medicare and Medicaid Services (CMS) uses 30-day readmissions as a measure of quality for a hospital system. However, there is growing literature that 30-day readmissions may not be a good measure of quality since most readmissions towards the end of 30-day window are likely not preventable. Therefore, using a shorter window may correlate better with the avoidability of the readmission. In this study, we used a novel readmission assessment tool and detailed chart reviews of all patients readmitted within 7-days of discharge from our hospitalist service to quantify the preventability of early readmissions, identify the major drivers, and determine if the timing of the readmission is associated with preventability.
Methods: We adapted the Kaiser Permanente Readmission Assessment tool to create a structured electronic process for systematic evaluation of each 7-day readmission from Jan 2018 to August 2018. The tool was modified through focus group discussions with hospitalists. The electronic tool automatically populates administrative data and is sent to the hospitalist who discharged the patient during the index admission. The hospitalist enters a brief description of the patient’s past history, reasons for readmission, and assesses the potential preventability of readmissions. All cases that were designated as not preventable by discharging attendings were reviewed by interdisciplinary team, including a dedicated objective hospitalist reviewer to identify additional readmissions that may have been preventable. The primary outcome was the percent potentially preventable readmissions. Reasons for readmissions were categorized and analyzed to determine the main drivers of early readmissions.
Results: The compliance rate for the electronic tool was 96%. The hospitalists categorized 41/195 cases (21%) as potentially preventable and the chart review identified an additional 37/195 cases (19%) as potentially preventable, yielding a total of 78/195 cases (40%) categorized as potentially preventable. The major drivers for potentially preventable readmissions were related to issues or system problems during pre-discharge planning (30%), care coordination (10%), and medication reconciliation (8%) at the time of discharge. The cause of the readmission for 66% of 7-day readmissions that were determined to be unpreventable was due to the underlying disease. Readmission within 24 hours were more likely categorized to be preventable than readmissions within 2-3 days or 4-7 days: 68%, 37% and 35%, respectively.
Conclusions: We found that 40% of 7-day readmissions from hospitalist services were preventable through an electronic tool completed by the discharging hospitalist for the index admission and an interdisciplinary team review process. Most readmissions occurring within 24 hours were considered as potentially preventable. Medications reconciliation and pre-discharge planning were the most common drivers of preventable readmissions and can be a focus for quality improvement projects to decrease readmissions.