Effect of an Artificial Intelligence Decision Support Tool on Palliative Care Referral in Hospitalized Patients: A Randomized Clinical Trial.

CONTEXT: Palliative care services are commonly provided to hospitalized patients, but accurately predicting who needs them remains a challenge. OBJECTIVES: To assess the effectiveness on clinical outcomes of an artificial intelligence (AI)/machine learning (ML) decision support tool for predicting patient need for palliative care services in the hospital. METHODS: The study design was a pragmatic, cluster-randomized, stepped-wedge clinical trial in 12 nursing units at two hospitals over a 15-month period between August 19, 2019, and November 17, 2020. Eligible patients were randomly assigned to either a medical service consultation recommendation triggered by an AI/ML tool predicting the need for palliative care services or usual care. The primary outcome was palliative care consultation note. Secondary outcomes included: hospital readmissions, length of stay, transfer to intensive care and palliative care consultation note by unit. RESULTS: A total of 3183 patient hospitalizations were enrolled. Of eligible patients, A total of 2544 patients were randomized to the decision support tool (1212; 48%) and usual care (1332; 52%). Of these, 1717 patients (67%) were retained for analyses. Patients randomized to the intervention had a statistically significant higher incidence rate of palliative care consultation compared to the control group (IRR, 1.44 [95% CI, 1.11-1.92]). Exploratory evidence suggested that the decision support tool group reduced 60-day and 90-day hospital readmissions (OR, 0.75 [95% CI, 0.57, 0.97]) and (OR, 0.72 [95% CI, 0.55-0.93]) respectively. CONCLUSION: A decision support tool integrated into palliative care practice and leveraging AI/ML demonstrated an increased palliative care consultation rate among hospitalized patients and reductions in hospitalizations.

Can a Novel ICU Data Display Positively Affect Patient Outcomes and Save Lives?

The aim of this study was to quantify the impact of ProCCESs AWARE, Ambient Clinical Analytics, Rochester, MN, a novel acute care electronic medical record interface, on a range of care process and patient health outcome metrics in intensive care units (ICUs). ProCCESs AWARE is a novel acute care EMR interface that contains built-in tools for error prevention, practice surveillance, decision support and reporting. We compared outcomes before and after AWARE implementation using a prospective cohort and a historical control. The study population included all critically ill adult patients (over 18 years old) admitted to four ICUs at Mayo Clinic, Rochester, MN, who stayed in hospital at least 24 h. The pre-AWARE cohort included 983 patients from 2010, and the post-AWARE cohort included 856 patients from 2014. We analyzed patient health outcomes, care process quality, and hospital charges. After adjusting for patient acuity and baseline demographics, overall in-hospital and ICU mortality odds ratios associated with AWARE intervention were 0.45 (95% confidence interval 0.30 to 0.70) and 0.38 (0.22, 0.66). ICU length of stay decreased by about 50%, hospital length of stay by 37%, and total charges for hospital stay by 30% in post AWARE cohort (by $43,745 after adjusting for patient acuity and demographics). Better organization of information in the ICU with systems like AWARE has the potential to improve important patient outcomes, such as mortality and length of stay, resulting in reductions in costs of care.

Testing modes of computerized sepsis alert notification delivery systems.

BACKGROUND: The number of electronic health record (EHR)-based notifications continues to rise. One common method to deliver urgent and emergent notifications (alerts) is paging. Despite of wide presence of smartphones, the use of these devices for secure alerting remains a relatively new phenomenon. METHODS: We compared three methods of alert delivery (pagers, EHR-based notifications, and smartphones) to determine the best method of urgent alerting in the intensive care unit (ICU) setting. ICU clinicians received randomized automated sepsis alerts: pager, EHR-based notification, or a personal smartphone/tablet device. Time to notification acknowledgement, fatigue measurement, and user preferences (structured survey) were studied. RESULTS: Twenty three clinicians participated over the course of 3 months. A total of 48 randomized sepsis alerts were generated for 46 unique patients. Although all alerts were acknowledged, the primary outcome was confounded by technical failure of alert delivery in the smartphone/tablet arm. Median time to acknowledgment of urgent alerts was shorter by pager (102 mins) than EHR (169 mins). Secondary outcomes of fatigue measurement and user preference did not demonstrate significant differences between these notification delivery study arms. CONCLUSIONS: Technical failure of secure smartphone/tablet alert delivery presents a barrier to testing the optimal method of urgent alert delivery in the ICU setting. Results from fatigue evaluation and user preferences for alert delivery methods were similar in all arms. Further investigation is thus necessary to understand human and technical barriers to implementation of commonplace modern technology in the hospital setting.

Automatic quality improvement reports in the intensive care unit: One step closer toward meaningful use.

AIM: To examine the feasibility and validity of electronic generation of quality metrics in the intensive care unit (ICU). METHODS: This minimal risk observational study was performed at an academic tertiary hospital. The Critical Care Independent Multidisciplinary Program at Mayo Clinic identified and defined 11 key quality metrics. These metrics were automatically calculated using ICU DataMart, a near-real time copy of all ICU electronic medical record (EMR) data. The automatic report was compared with data from a comprehensive EMR review by a trained investigator. Data was collected for 93 randomly selected patients admitted to the ICU during April 2012 (10% of admitted adult population). This study was approved by the Mayo Clinic Institution Review Board. RESULTS: All types of variables needed for metric calculations were found to be available for manual and electronic abstraction, except information for availability of free beds for patient-specific time-frames. There was 100% agreement between electronic and manual data abstraction for ICU admission source, admission service, and discharge disposition. The agreement between electronic and manual data abstraction of the time of ICU admission and discharge were 99% and 89%. The time of hospital admission and discharge were similar for both the electronically and manually abstracted datasets. The specificity of the electronically-generated report was 93% and 94% for invasive and non-invasive ventilation use in the ICU. One false-positive result for each type of ventilation was present. The specificity for ICU and in-hospital mortality was 100%. Sensitivity was 100% for all metrics. CONCLUSION: Our study demonstrates excellent accuracy of electronically-generated key ICU quality metrics. This validates the feasibility of automatic metric generation.