MUST-Plus: A Machine Learning Classifier That Improves Malnutrition Screening in Acute Care Facilities.

OBJECTIVE: Malnutrition among hospital patients, a frequent, yet under-diagnosed problem is associated with adverse impact on patient outcome and health care costs. Development of highly accurate malnutrition screening tools is, therefore, essential for its timely detection, for providing nutritional care, and for addressing the concerns related to the suboptimal predictive value of the conventional screening tools, such as the Malnutrition Universal Screening Tool (MUST). We aimed to develop a machine learning (ML) based classifier (MUST-Plus) for more accurate prediction of malnutrition. METHOD: A retrospective cohort with inpatient data consisting of anthropometric, lab biochemistry, clinical data, and demographics from adult (≥ 18 years) admissions at a large tertiary health care system between January 2017 and July 2018 was used. The registered dietitian (RD) nutritional assessments were used as the gold standard outcome label. The cohort was randomly split (70:30) into training and test sets. A random forest model was trained using 10-fold cross-validation on training set, and its predictive performance on test set was compared to MUST. RESULTS: In all, 13.3% of admissions were associated with malnutrition in the test cohort. MUST-Plus provided 73.07% (95% confidence interval [CI]: 69.61%-76.33%) sensitivity, 76.89% (95% CI: 75.64%-78.11%) specificity, and 83.5% (95% CI: 82.0%-85.0%) area under the receiver operating curve (AUC). Compared to classic MUST, MUST-Plus demonstrated 30% higher sensitivity, 6% higher specificity, and 17% increased AUC. CONCLUSIONS: ML-based MUST-Plus provided superior performance in identifying malnutrition compared to the classic MUST. The tool can be used for improving the operational efficiency of RDs by timely referrals of high-risk patients.

Effects of a Psychosocial Transitional Care Model on Hospitalizations and Cost of Care for High Utilizers.

Evidence of care coordination programs to reduce readmissions is limited. We examined whether a social work transitional care model reduced hospital utilization and costs with a retrospective cohort study conducted from 9/3/2010-8/31/2012. Patients enrolled in the Preventable Admissions Care Team (PACT) program were matched to controls. PACT patients received follow-up from a social worker to address psychosocial strain. PACT reduced thirty-day readmission rate by 34% (p = <0.001), Sixty-day hospitalization rate by 22% (p = 0.004); ninety-day hospitalization rate by 19% (p = 0.006), and but not 180-day hospitalization rate. Inpatient costs thirty days post-index were $2.7 million for PACT patients and $3.6 million for controls.

The Impact of Bedside Interdisciplinary Rounds on Length of Stay and Complications.

BACKGROUND: Communication among team members within hospitals is typically fragmented. Bedside interdisciplinary rounds (IDR) have the potential to improve communication and outcomes through enhanced structure and patient engagement. OBJECTIVE: To decrease length of stay (LOS) and complications through the transformation of daily IDR to a bedside model. DESIGN: Controlled trial. SETTING: 2 geographic areas of a medical unit using a clinical microsystem structure. PATIENTS: 2005 hospitalizations over a 12-month period. INTERVENTIONS: A bedside model (mobile interdisciplinary care rounds [MICRO]) was developed. MICRO featured a defined structure, scripting, patient engagement, and a patient safety checklist. MEASUREMENTS: The primary outcomes were clinical deterioration (composite of death, transfer to a higher level of care, or development of a hospital-acquired complication) and length of stay (LOS). Patient safety culture and perceptions of bedside interdisciplinary rounding were assessed pre- and postimplementation.. RESULTS: There was no difference in LOS (6.6 vs 7.0 days, P = 0.17, for the MICRO and control groups, respectively) or clinical deterioration (7.7% vs 9.3%, P = 0.46). LOS was reduced for patients transferred to the study unit (10.4 vs 14.0 days, P = 0.02, for the MICRO and control groups, respectively). Nurses and hospitalists gave significantly higher scores for patient safety climate and the efficiency of rounds after implementation of the MICRO model. LIMITATIONS: The trial was performed at a single hospital. CONCLUSIONS: Bedside IDR did not reduce overall LOS or clinical deterioration. Future studies should examine whether comprehensive transformation of medical units, including co-leadership, geographic cohorting of teams, and bedside interdisciplinary rounding, improves clinical outcomes compared to units without these features. Journal of Hospital Medicine 2017;12:137-142.

Reducing excessive medication administration in hospitalized adults with renal dysfunction.

Medication errors are common and harm hospitalized patients. The authors designed and implemented an automated system to complement an existing computerized order entry system by detecting the administration of excessive doses of medication to adult in-patients with renal insufficiency. Its impact, in combination with feedback to prescribers, was evaluated in 3 participating nursing units and compared with the remainder of a tertiary care academic medical center. The baseline rate of excessive dosing was 23.2% of administered medications requiring adjustment for renal insufficiency given to patients with renal impairment on the participating units and 23.6% in the rest of the hospital. The rate fell to 17.3% with nurse feedback and 16.8% with pharmacist feedback in the participating units (P<.05 for each, relative to baseline). The rates of excessive dosing for the same time periods were 26.1% and 24.8% in the rest of the hospital. Automated detection and routine feedback can reduce the rate of excessive administration of medication in hospitalized adults with renal insufficiency.