Background: Our institution is a Level 1 regional trauma center with a large inpatient volume. A large percentage of these patients are elderly, with the traumatic events occurring as sequelae of their age, debility or medical complications. Last year, we developed a Trauma/Acute Care Surgery Hospitalist Co-Management program, due to, in large part, the increasing demand for medical care of elderly acute surgical patients. A select group of specialized hospitalists aid in the management of anticoagulation, diabetes, and pain control in the perioperative period, as well as fall-risk assessment and prevention, and ensure appropriate transitions of care. We sought to quantify the effects of the co-management service in terms of throughput, patient safety, and other quality measures, which is often challenging in co-management arrangements.
Purpose: We aimed to monitor key metrics to assess the effect of co-management on this specific population of patients. Metrics decided upon by key stakeholders included average length of stay (LOS), case mix index (CMI), excess days per case, readmission index, mortality index, CAUTI, CLABSI, and VTE rates, and patient satisfaction; these metrics are in-line with our institutional goals.
Description: Patients who were co-managed by our hospitalist group were assigned a tag in our EMR. Quarterly reports were generated, comparing co-managed surgical patients with patients who were not tagged, as well as baseline data from 2016. Furthermore, our tagged patients were subdivided by admission diagnosis into groups of patients with trauma diagnoses and those with acute care surgery diagnoses, to allow for more refined cohort matching and comparison. Through Quarter 2 of 2017, we have demonstrated a shorter length of stay for co-managed trauma patients compared to one year prior (5.21 days vs. 6.16 days) with similar case mix indices, as well as improvements in patient mortality index (0.23 vs. 1.24) and HCAHPS survey results for all patients co-managed (Likelihood to recommend the hospital always: 76% vs. 61%; Doctor explains things in way you understand always: 78% vs. 50%). Readmission index remained grossly unchanged from year prior. We have now developed a tool within our EMR to track this data in real-time, as well as other demographics and metrics, such as LACE score, Charlson co-morbidity index, time to physical therapy evaluation, and lists of high-risk medications that can increase morbidity in the elderly population; our goal is to now work prospectively to identify high-risk patients to ensure prompt intervention.
Conclusions: By utilizing our EMR and tagged patient population, with cohort matching based on diagnosis, we have been able to successfully evaluate our co-management program. Through this effort, we have demonstrated significant gains in throughput, quality and patient satisfaction for those patients that are co-managed. We plan to continue to track this data, as well as more granular metrics to ascertain our effect and identify potential opportunities in real-time.