Effective quality improvement (QI) education should improve patient care, but many curriculum studies do not include clinical measures. The research team evaluated the prevalence of QI curricula with clinical measures and their association with several curricular features. MEDLINE, Embase, CINAHL, and ERIC were searched through December 31, 2013. Study selection and data extraction were completed by pairs of reviewers. Of 99 included studies, 11% were randomized, and 53% evaluated clinically relevant measures; 85% were from the United States. The team found that 49% targeted 2 or more health professions, 80% required a QI project, and 65% included coaching. Studies involving interprofessional learners (odds ratio [OR] = 6.55; 95% confidence interval [CI] = 2.71-15.82), QI projects (OR = 13.60; 95% CI = 2.92-63.29), or coaching (OR = 4.38; 95% CI = 1.79-10.74) were more likely to report clinical measures. A little more than half of the published QI curricula studies included clinical measures; they were more likely to include interprofessional learners, QI projects, and coaching.

Introduction: Acute physiological deterioration is a major contributor to in-hospital morbidity and mortality. Early detection and intervention of deteriorating patients is key to improving patient outcomes. Prior research has demonstrated the effectiveness of Early Warning Systems and other algorithmic approaches in automatically identifying these patients from passively monitoring vital signs. Methods: In this work, we conduct a prospective pilot study of clinical deployment of the Mayo Clinic Bedside Patient Rescue (BPR) system using an escalating alerting logic enabled by machine learning. Among four units where the BPR system was deployed, time to response and time to intervention for deteriorating patients were significantly reduced relative to matched control units. Results: In pilot units, time to response decreased by 35.4% (from 63.2 minutes to 40.8 minutes) and time to intervention decreased by 48.5% (from 106.3 minutes to 55.9 minutes). No significant differences were observed in counterbalance metrics of mortality, ICU transfer rate, and Rapid Response Team activation rate. Furthermore, the automated alerting system was well-received by clinicians participating in the pilot study, as assessed by survey. Discussion: These results demonstrate a successful clinical deployment of a practice-changing machine learning alert system with demonstrable impact on improving patient care.

A 53-year-old man was admitted to the hospital medical service after an outpatient evaluation for encephalopathy revealed a serum sodium level of 166 mmol/L. The duration of the patient's hypernatremia was unknown. The patient had been previously healthy but in the past year had experienced progressive fatigue and neurologic disturbances including somnolence, constructional apraxia, cognitive impairment, narcolepsy, cataplexy, and rapid eye movement sleep behavior symptoms. Several months before the current admission, he was noted to be hyponatremic, likely as a consequence of syndrome of inappropriate antidiuretic hormone secretion, and he was treated with water restriction and demeclocycline. In the course of evaluating his symptoms, he was found to have a single lymph node on the right side of his neck that was positive for metastatic squamous cell carcinoma; a tonsillar origin was suspected, but ultimately no primary lesion was identified. Additional imaging evaluation at that time (approximately 2 months before admission) prompted a hypothalamic biopsy that revealed noncaseating granulomas in the absence of infection. In the time leading up to admission, his symptoms continued to progress, with fluctuating hypernatremia and hyponatremia, polydipsia, polyuria, and worsening hypersomnia.The patient's outpatient medications included treatments for neurosarcoidosis (initiated 2 months before the current admission), including prednisone (40 mg daily), methotrexate, cyclosporine, hydroxychloroquine, and infliximab. In addition, he was receiving demeclocycline (600 mg twice daily) for treatment of hyponatremia. He was not on fluid restriction but had unreliable oral intake because of his progressive encephalopathy. When aroused and prompted, he had an intact thirst mechanism and was able to drink.Laboratory evaluation at admission yielded the following (reference ranges provided parenthetically): serum sodium, 166 mmol/L (135-145 mmol/L); serum potassium, 4.0 mmol/L (3.6-5.2 mmol/L); serum glucose, 92 mg/dL (70-140 mg/dL); creatinine, 1.4 mg/dL (0.8-1.3 mg/dL); and serum urea nitrogen, 31 mg/dL (8-24 mg/dL).On examination, the patient was somnolent but arousable to voice (Glasgow Coma Scale score, 13). He was oriented to person but not place or time. His mentation precluded a comprehensive review of systems, but collateral information from his family was positive for progressive inability to perform activities of daily living and unintentional 11.25-kg weight loss.

No abstract