BACKGROUND:Improved patient experience is associated with enhanced clinical outcomes. The Hospital Consumer Assessment of Healthcare Providers and Systems Overall Rating of the Hospital (ORH) can be considered a quality metric for health systems. We hypothesized that machine learning models could identify actionable predictors of a less-than-top-box ORH score across a multihospital system and sought to determine whether these predictors were consistent across model architectures and hospital size. METHODS:We analyzed 86 351 Hospital Consumer Assessment of Healthcare Providers and Systems surveys from 12 hospitals within a New Jersey health system (January 2023 to May 2026). Four machine learning models (logistic regression, decision tree, random forest, and XGBoost) were developed to predict a less-than-top-box ORH score. Feature importance was compared across the 3 highest-performing models, and SHapley Additive exPlanations (SHAP) analysis was performed on the XGBoost model. All analyses were repeated separately across hospitals stratified by size (large, medium, and small). RESULTS:XGBoost, logistic regression, and random forest demonstrated excellent, comparable discrimination (area under the curve 0.871-0.883); the decision tree model underperformed and was excluded. Feature importance and SHAP analysis converged on the same 3 predictor domains: nursing communication, physician communication, and hospital environment, which jointly accounted for 67% of total SHAP importance. This 3-domain pattern remained stable across hospital-size strata, with only modest shifts in relative ranking. CONCLUSIONS:Across multiple machine learning architectures, analytic methods, and hospital sizes, nursing communication, physician communication, and hospital environment consistently emerged as the primary, modifiable drivers of overall hospital rating. These convergent, generalizable findings identify clear strategic priorities for health systems seeking to improve patient experience.
BACKGROUND:The Universal Protocol has standardized pre-procedural verification since 2004, yet procedural sentinel events persist; 127 wrong-surgery events were reported nationally in 2024. Compliance-based checklists may be insufficient for the human-factors environment of complex procedural care. METHODS:In a prospective 3-phase quality improvement evaluation at a rural health system, we tested an enhanced timeout (HIPESAC) in which the proceduralist presents and justifies each verification element to the full team, any member empowered to question, correct, or halt the procedure. For high-risk cases, an expanded Patient Safety Advocate (PSA) protocol adds voicing of concerns from least to most senior, with a designated patient advocate issuing consent to proceed. Phase 1 piloted 4 PSA-activated procedures; Phase 2 audited 20 with surveys; and Phase 3 evaluated the standard timeout across 22 procedures. RESULTS:The platform identified latent threats before initiation: material readiness events in 8 of 20 high-risk procedures (40%; one critical absent item, 7 proactive adjustments) and 9 process deficiencies across 22 standard procedures. The timeout added a median of 54 seconds. Surveys indicated high acceptance among respondents, with unanimous endorsement of speaking-up empowerment. In 2 cases-one per tier-a non-proceduralist surfaced information that altered the procedure. CONCLUSIONS:The platform was feasible, well accepted, and surfaced latent procedural threats in this single-center evaluation. By requiring the proceduralist to present to the team and, at higher risk, inverting the order of voice, it may operationalize authority-gradient inversion as an intrinsic feature. Multi-site controlled evaluation is warranted.
Using artificial intelligence to identify themes in interview data from a quality improvement program evaluation produced 4 replicable themes grounded in the data. However, 2 consistently identified themes resulted from subtle misrepresentations and could have easily misled results without thorough data knowledge and output audit by the human research team.
INTRODUCTION:University Hospitals transformed its quality strategy by transitioning to Det Norske Veritas accreditation and seeking International Organization for Standardization (ISO) 9001 certification to support its zero-harm framework and establish a sustainable quality management system. INTERVENTION:Through standardized governance, leadership engagement, risk-based thinking, internal auditing, and an ISO 9001 implementation toolkit, hospitals embedded continuous readiness, accountability, and improvement into daily operations while aligning board-level oversight with frontline practice. CLOSING GAPS:Det Norske Veritas accreditation and ISO 9001 certification are not well-known models in the United States. This journey describes implementation of a scalable model, moving beyond episodic compliance to continuous learning and improvement.
Background: Mental health and substance use can coexist and worsen outcomes for youth. The goals of the Transformative and Evidence-based Approaches to Mental Health and Substance Use Screening (TEAMSS) quality improvement virtual learning collaborative were to enhance the implementation of screening and follow-up practices for adolescent substance use and mental health issues in primary care. Methods: TEAMSS engaged 160 clinicians at 27 pediatric primary care clinics in 18 states. Interventions focused on enhancing clinical processes related to screening, brief intervention, and follow-up for substance use, depression, anxiety, and suicide risk. The learning collaborative model was adapted for a virtual environment with learning sessions and action periods of shorter duration interspersed with individualized coaching. Results: Clinics standardized processes for screening and intervention, incorporated use of validated screening tools, increased the use of electronic health records and mobile devices in screening, engaged interprofessional team members, identified community resources, and increased clinician and staff knowledge of evidence-based guidelines and quality improvement methods. Substance use screening significantly increased from 35% to 75%. Follow-up of adolescents with positive substance use screening increased from 69% to 86%. Depression screening increased from 82% to 90%, anxiety screening increased from 32% to 69%, and suicide risk screening increased from 18% to 41%. Conclusions: TEAMSS significantly improved care for substance use, depression, anxiety, and suicide risk for adolescents at participating primary care clinics. Adapting a learning collaborative approach to a virtual setting requires close attention to maintaining engagement with clinical care teams, especially those in busy primary care clinics.
Computed tomography-guided percutaneous lung biopsy (CT-PLB) is a cornerstone diagnostic procedure for pulmonary lesions. While traditionally managed via inpatient admission, this model imposes considerable clinical and operational burdens. Day-surgery management offers a streamlined alternative, yet comparative evidence on its safety and efficiency remains scarce. The authors conducted a retrospective cohort study of patients undergoing CT-PLB in 2024, comparing a day-surgery pathway (Day-Surgery Center) with the conventional inpatient pathway (oncology ward). Propensity score matching ensured group comparability. Outcomes focused on clinical management metrics: safety (complication/readmission rates), diagnostic performance, and health care efficiency (length of stay, cost). Among 154 patients, overall outcomes were favorable. After matching, the day-surgery cohort demonstrated comparable safety (complication rate 2.22% vs 6.67%, risk difference -4.45% [95% confidence interval, -12.0% to 3.4%]; P = 0.62) and comparable diagnostic accuracy (82.22% vs 79.17%, P = 0.84) relative to inpatients. Crucially, the day-surgery model significantly optimized efficiency, drastically reducing hospitalization duration and costs ( P < 0.001), with 100% patient satisfaction in both groups. Day-surgery CT‑PLB appears comparable to inpatient procedure in safety and diagnostic yield, with significantly lower costs and shorter hospital stay. Given the retrospective design and small sample size, these findings are hypothesis‑generating. Prospective, adequately powered validation is needed.
BACKGROUND:The steady increase in laboratory testing volume burdens the health care system with substantial costs, contributes to patient dissatisfaction, and generates adverse environmental effects. Beyond financial expenditure, unnecessary testing poses medical risks and affects patient well-being. Our quality improvement (QI) project aimed to employ a multimodal intervention to enhance appropriate laboratory utilization amongst residents and hospitalists. METHODS:This QI initiative used a multimodal intervention to enhance appropriate lab utilization among residents and hospitalists. It involved evaluating baseline knowledge gaps, implementing targeted education, and integrating daily structured discussions on lab necessity. In addition, an electronic dashboard was utilized with real-time data on lab ordering volume and labs ordered per attending. RESULTS:The intervention led to a net decrease of 4657 lab orders over the 3-month period. Specifically, there were significant reductions in comprehensive metabolic panel (CMP) (19.5%), serum Mg (11.8%), serum Phos (14.1%), and complete blood count (CBC) (11.2%) orders, alongside a 20% increase in basic metabolic panel (BMP) orders. Patient satisfaction improved, with affirmative "rest and recover" responses rising from 51.8% to 63.2%. This reduction generated an estimated $10,645 in cost savings and a 1304.61 kg reduction in CO2 emissions. CONCLUSION:This multifaceted QI approach, integrating systemic interventions and education, effectively reduced unnecessary lab testing. It simultaneously improved patient experience, yielded cost savings, and mitigated environmental impact, offering a transferable framework.
In July 2021, our institution launched a program to begin documenting the medically ready for discharge date (MRDD) and the discharge delays (DDs) in the electronic medical record (EMR) to better allocate resources to streamline patient flow. The MRDD is the date at which a patient is deemed clinically stable for discharge by their medical team. DDs describe the reasons why a patient remains in the hospital beyond the MRDD. Our aim was to increase entry of DDs from 1.7% to ≥10% in hospitalized patients on a 26-bed Medicine unit between November 2023 and October 2025. The DD information would enable us to identify trends among vulnerable patient populations. We implemented Plan-Do-Study-Act cycles utilizing in-person and email communication by interdisciplinary teams as well as incorporating specific columns into the residents' EMR. We additionally stratified DD types based on the patient social vulnerability index. Since November 2023, the percentage of documented DDs for patients discharged beyond the MRDD was sustained above the target of 10% and peaked at 24% in October 2025. The most common themes of DDs were Placement, Disposition, Consult, and Home Coordination. We also found patients with higher social vulnerability index were disproportionately affected by DDs, including Consult, Home Coordination, and Transportation.
Transfers are high-risk events, and outside hospital admissions lack standardized processes to ensure timely delivery of essential clinical information. We piloted a novel, standardized process for outside admissions to a single academic medical center’s resident- and hospital medicine-attending-based medicine services. Following literature review, Epic UserWeb review, and frontline physician input, a workflow and templated electronic health record accept note were developed to centralize outside admission documentation. Completed by physicians during real-time communication with outside physicians, the template captures key clinical data and incorporates a communication checklist. Physician education was delivered electronically and in person. Subjective attending and resident perceptions were assessed via electronic prepost surveys administered 1 week before and 90 days after implementation. Secondarily, exploratory patient clinical outcomes were assessed through 90-day prepost manual retrospective chart review. Survey responses were obtained from 60/38 residents and 12/7 attendings (pre/post). Documentation sufficiency on patient arrival was directionally promising among residents (21.7% vs 39.5%, P = 0.062). Satisfaction with the admission process and perceived patient safety improved significantly among residents and attendings ( P ≤ 0.036). Most respondents rated the note as useful or very useful (84% residents; 86% attendings), and attendings reported reduced time managing outside admissions. The intervention was associated with exploratory, nonsignificant reductions in rapid response team activation, intensive care unit transfer, 30-day readmission, mortality, and length of stay compared with preintervention patients. A standardized documentation and communication process improved physician satisfaction and perceived patient safety, with promising exploratory patient outcomes. Further study is warranted to assess impact at scale.
The Centers for Medicare & Medicaid Services has recently tied the Severe Sepsis and Septic Shock Early Management Bundle (SEP-1) compliance to financial incentives, increasing pressure on hospital performance. We aim to review the mechanics of SEP-1 and provide a focused guide for best practices to improve compliance from the perspective of a US-based community emergency department. SEP-1 quality improvement initiatives may benefit from focusing on common failure points. Key strategies include multidisciplinary engagement, staff education, workflow changes, and electronic health record optimization. Implementing such programs may improve performance on this increasingly important metric.
In this pre-post Define, Measure, Analyze, Improve, Control quality improvement study, an Epic electronic health record best practice advisory was implemented to prompt guideline-directed medical therapy (GDMT) for heart failure with reduced ejection fraction. Documented GDMT prescribing rose markedly from 9.1% to 42.2%, supporting electronic health record-based prompts as an effective strategy for GDMT optimization.
BACKGROUND:Hepatitis A virus (HAV) infection remains a preventable cause of acute viral hepatitis with disproportionate morbidity among adults with chronic liver disease and people living with HIV. Despite longstanding Advisory Committee on Immunization Practices recommendations, HAV vaccination coverage in high-risk adults remains suboptimal. We conducted a quality improvement initiative to improve HAV vaccination practices in an academic ambulatory clinic. METHODS:This QI project was implemented at an academic ambulatory care center using Plan-Do-Study-Act (PDSA) cycles from October 2024 to April 2025. Interventions included electronic health record smart-phrase modification, resident and nursing education, workflow integration, and patient education. Adults with cirrhosis or HIV seen during preintervention (October 2023-October 2024) and postintervention (December 2024-April 2025) periods were included in the analysis. Primary outcomes were HAV vaccination initiation and completion rates; secondary outcomes included documentation of HAV discussions and patterns of serologic testing. Proportions were compared using 2-proportion z-tests. RESULTS:Among patients with HIV (n = 91), overall HAV serologic testing declined significantly postintervention (64% vs 38%, P < 0.05). The completion rate of the HAV vaccine series increased from 18% to 30%, although this did not reach statistical significance. Documentation of planned vaccination improved modestly within the internal medicine clinic. Among patients with cirrhosis (n = 100), HAV immunity and vaccination rates were high at baseline and remained stable postintervention, with a nonsignificant increase in HAV addressal (78% vs 84%). Across both cohorts, documentation of vaccination plans improved slightly but remained limited. CONCLUSIONS:A multifaceted, electronic health record-integrated QI intervention was associated with reduced HAV serologic testing and modestly improved vaccination practices, particularly in patients with HIV. Persistent gaps in vaccine completion highlight the need for sustained workflow and follow-up strategies to improve HAV immunization among high-risk adults.
BACKGROUND:Thrombophilia testing is frequently overused despite limited clinical utility in most acute thrombotic settings and clear guideline recommendations, contributing to unnecessary health care utilization and costs. OBJECTIVE:To improve the appropriateness of thrombophilia testing through EHR-based clinical decision support. METHODS:Five thrombophilia PowerPlans containing outdated or inappropriate testing were revised into 4 evidence-based PowerPlans using current guidelines and local expert consensus. Orders were analyzed over 30 weeks before and after intervention for PowerPlan usage, inappropriate testing, ordering by provider type, and estimated laboratory charges. RESULTS:Appropriate testing remained stable, whereas nonevidence-based testing decreased from 19.0% to 12.6% ( P = 0.00001), with ordering reduction across all provider groups. Postintervention, inappropriate PowerPlans orders decreased from 41.4% to 3.1%, with laboratory charges decreasing by $117 436 for nonevidence-based testing over 30 weeks. CONCLUSIONS:Updating PowerPlans reduced inappropriate thrombophilia testing and estimated laboratory charges without reducing appropriate testing, demonstrating the value of EHR-based interventions in promoting high-value care.
In this retrospective study of 1217 incidental imaging findings within a large integrated health system, 97% were classified as actionable and required follow-up, and an electronic medical record-based communication workflow contributed to follow-up in 29% of cases.
