In 2012, the Fairview Health System implemented a formal care transitions process that included referrals to outpatient services provided by medication therapy management (MTM) pharmacists, among other clinical services. This analysis evaluates the impact of the MTM-provided comprehensive medication management (CMM) service on readmission rates. Retrospective electronic medical record (EMR) data were used to identify hospital admissions between December 1, 2012, and July 31, 2015. Thirty- and 60-day readmission rates were calculated in both a CMM and comparator cohort. Readmission rates also were stratified by readmission risk category. A total of 43,711 patients, contributing 57,673 hospitalizations, were included in the analysis. Of those, 1291 hospitalizations had a CMM visit within 30 days of discharge (median 6 days) and were considered the CMM cohort. Patients who received a CMM visit had a significantly lower rate of 30-day readmissions (8.6% vs. 12.8%, P < 0.001). The 60-day readmission rate remained lower among CMM patients but did not reach statistical significance (15.6% vs. 17.6%; P = 0.0528). When patients in each cohort were stratified by readmission risk category, the CMM cohort had a statistically significant lower rate of 30-day readmission in the highest risk groups (Average: 7.1% vs. 9.5%, P = 0.025; Elevated: 9.9% vs. 21.4%, P < 0.001; High: 18.3% vs. 35.9%, P < 0.001; Extreme: 36.4% vs. 77.7%, P = 0.006). CMM performed by an MTM pharmacist reduces the rate of readmission at 30 days post discharge and may have the largest impact among patients at highest risk of readmission.

No outside funding supported this work. Affeldt reports advisory board membership with Janssen, and Skiermont reports membership with Amgen and McKesson. The other authors have nothing to disclose. Peterson and Budlong contributed the study concept and design and wrote the manuscript. Affeldt, Skiermont, Kyllo, and Heaton reviewed and revised the manuscript.

AC, p,0.001). In the last 12 months 36% SC patients had been hospitalized vs 16% AC (p,0.001) and had mean of 2.3 vs 2.0 (p=0.007) comorbidities respectively. Poorvery poor Q-LES-Q overall life satisfaction was reported by 24% SC patients and 11% AC (p,0.001) with overall scores of 49.7 and 56.2 (p,0.001) respectively. Poorer quality of life was also reported through mean EQ5D VAS scores (64.6 SC vs 68.9 AC p=0.01). Patients reported mean WPAI percentages of 34.2 SC vs 31.0 AC for overall work impairment (p=0.44) and 47.5 SC vs 42.7 AC (p=0.038) for daily activities impairment. Conclusions: Poorer outcomes and increased likelihood of hospitalization are evident in Schizophrenia patients who are sometimes compliant with current treatment compared to those who are always compliant.

e18298 Background: Febrile neutropenia (FN) is a key driver of morbidity and mortality in oncology patients receiving myelotoxic chemotherapy. Identification and prophylaxis per guidelines can result in a substantial reduction of FN risk and is a key oncology quality of care measure. This project evaluated the feasibility of using electronic medical records (EMR) to quantify FN rates and characterize patients at risk for FN. Methods: Search algorithm was used in EMR to identify patients initiating oncologic care between 1/1/2016 and 12/1/2017 at the Masonic Cancer Clinic, University of Minnesota—Fairview Health System, for a lymphoid, gastrointestinal, breast, female genital, or thoracic malignancy. FN event was identified using a diagnosis of neutropenia (ICD-10: D70.x) with associated fever (ICD-10: R50.2-R50.9) or infection (ICD-10: A40.x, A41.x, B95.0-B95.8, B96.0-B96.8, T80.21-T80.29) during the myelotoxic chemotherapy course. FN rates were evaluated overall and by cancer type, disease status, prior chemotherapy use, select comorbidities, age, curative or palliative intent, and treatment plan FN risk. Results: A total of 1,123 patients receiving 1,663 myelotoxic chemotherapy plans were identified. Patients were predominately female (60%) with a mean age of 61 years (SD: 14.1). A total of 66 patients experienced 79 FN events. Of these events, 70 (88.6%) resulted in a hospitalization. The FN rate during the study period was 5.9% (95% CI: 2.3%-10.2%). Factors associated with the highest FN rates were: lymphoid malignancy (9.9%), non-metastatic disease (7.2%), no prior chemotherapy (6.5%), ≥ 3 comorbidities. Conclusions: Our findings indicate it is feasible to estimate FN rates using a search algorithm with a known FN definition in the EMR of a large integrated delivery network. While additional analyses are planned to account for concomitant patient characteristics to fully understand the occurrence of FN, the current project nevertheless demonstrates the ability of estimating FN rates to both measure care quality and allow pursuing quality improvement (QI) initiatives for appropriate FN prophylaxis. The results also have the potential to serve as a benchmark for similar QI projects in other large healthcare systems.

No abstract