Background Physical restraints are frequently used for intensive care patients and are associated with substantial morbidity. The effects of common evidence-based critical care interventions on use of physical restraints remain unclear. Objective To identify independent predictors of new-onset use of physical restraints in critically ill adults. Methods Secondary analysis of a prospective cohort study involving 5 adult intensive care units in a tertiary care medical center in the United States. Use of physical restraints was determined via daily in-person assessments and medical record review. Mixed-effects logistic regression analysis was used to examine factors associated with new-onset use of physical restraints, adjusting for covariates and within-subject correlation among intensive care unit days. Results Of 145 patients who were free of physical restraints within 48 hours of intensive care unit admission, 24 (16.6%) had restraints newly applied during their stay. In adjusted models, delirium (odds ratio [OR], 5.09; 95% CI, 1.83-14.14), endotracheal tube presence (OR, 3.47; 95% CI, 1.22-9.86), and benzodiazepine administration (OR, 3.17; 95% CI, 1.28-7.81) significantly increased the odds of next-day use of physical restraints. Tracheostomy was associated with significantly lowered odds of next-day restraint use (OR, 0.13; 95% CI, 0.02-0.73). Compared with patients with a target sedation level, patients who were in a coma (OR, 2.56; 95% CI, 0.80-8.18) or deeply sedated (OR, 2.53; 95% CI, 0.91-7.08) had higher odds of next-day use of physical restraints, and agitated patients (OR, 0.08; 95% CI, 0.00-2.07) were less likely to experience restraint use. Conclusion Several potentially modifiable risk factors are associated with next-day use of physical restraints.

Background Early warning systems lack robust evidence that they improve patients’ outcomes, possibly because of their limitation of predicting binary rather than time-to-event outcomes. Objectives To compare the prediction accuracy of 2 statistical modeling strategies (logistic regression and Cox proportional hazards regression) and 2 machine learning strategies (random forest and random survival forest) for in-hospital cardiopulmonary arrest. Methods Retrospective cohort study with prediction model development from deidentified electronic health records at an urban academic medical center. Results The classification models (logistic regression and random forest) had statistical recall and precision similar to or greater than those of the time-to-event models (Cox proportional hazards regression and random survival forest). However, the time-to-event models provided predictions that could potentially better indicate to clinicians whether and when a patient is likely to experience cardiopulmonary arrest. Conclusions As early warning scoring systems are refined, they must use the best analytical methods that both model the underlying phenomenon and provide an understandable prediction.

Background Use of the interprofessional Awakening and Breathing Coordination, Delirium Monitoring and Management, and Early Mobility (ABCDE) bundle is recommended practice in intensive care, but its adoption remains limited. Objective To examine the relationship between intensive care unit provider attitudes regarding the ABCDE bundle and ABCDE bundle adherence. Methods A 1-time survey of 268 care providers in 10 intensive care units across the country who had worked at least 4 shifts per month to examine their attitudes toward workload burden, difficulty carrying out the bundle, perceived safety, confidence, and perceived strength of evidence. Logistic regression models were used to examine the relationship of unit-level provider attitudes with ABCDE bundle adherence in 101 patients, adjusted for patients’ age, severity of illness, and comorbidity. Results For every unit increase in workload burden, adherence to the ABCDE bundle decreased 53% (odds ratio [OR], 0.47; 95% CI, 0.28–0.79; P = .004). Bundle difficulty (OR, 0.29; 95% CI, 0.08–1.07), perceived safety (OR, 0.51; 95% CI, 0.10–2.65), confidence (OR, 0.37, 95% CI, 0.10–1.35), and perceived strength of evidence (OR, 0.69; 95% CI, 0.14–3.35) were not associated with ABCDE bundle adherence. For every unit increase in perceived difficulty carrying out the bundle, adherence with early mobility was reduced 59% (OR, 0.41; 95% CI, 0.19–0.90; P = .03). In addition, ABCDE bundle adherence (ie, ventilator bundle) was less than DE bundle adherence (ie, ventilator-free bundle) (97% vs 72%, z = 5.47; P < .001). Conclusions Focusing interventions on workload burden and factors influencing bundle difficulty may facilitate ABCDE bundle adherence.

Background The ABCDE interprofessional bundle (Awakening and Breathing Coordination, Delirium monitoring and management, and Early mobility) reduces delirium and weakness in critically ill patients. Objective To understand the relationship between organizational domains and provider attitudes. Methods A 1-time electronic survey of 315 care providers in 10 intensive care units across the country to examine associations between organizational domains (policy/protocol factors, unit milieu, tasks, labor quality, labor quantity, and physical environment) and provider attitudes about perceived ease of completion, perceived safety, confidence, and perceived strength of evidence regarding the ABCDE bundle. Spearman correlations ( r s ) were used to examine the associations between organizational domains and provider attitude subscales ( r s ≥ 0.32 was considered clinically important). Results Protocol attributes ( r s = 0.37–0.58), role clarity ( r s = 0.38–0.59), training/understanding ( r s = 0.33–0.46), coordination ( r s = 0.32–0.46), and peer advocates ( r s = 0.37–0.48) were associated with less difficulty performing the bundle and better confidence, perceived safety, and strength of evidence. Participants also reported less difficulty carrying out the bundle when the team worked well together. Task autonomy was associated with better perceived safety ( r s = 0.35) and confidence ( r s = 0.47) related to the bundle. Conclusions Focusing interventions on policy and protocol factors, unit milieu, and task autonomy, which have the strongest associations with providers’ attitudes, may facilitate ABCDE bundle uptake.