The objective of this study was to assess the clinical and financial impact of a quality improvement project that utilized a modified Early Warning Score (mEWS)-based clinical decision support intervention targeting early recognition of sepsis decompensation.
319 Background: Sepsis is a life-threatening organ dysfunction caused by a dysregulated response to infection and a major healthcare problem in cancer patients. Patients with cancer are at increased risk of morbidity and mortality from sepsis. Approaches like the modified Early Warning Score (mEWS) could offer an opportunity to identify patients at risk for sepsis earlier and to prompt a timely evaluation. However, available evidence for cancer outpatient settings remains scarce. The aim of this presentation is to describe first results of the mEWS implementation at Huntsman Cancer Hospital (HCC) outpatient clinics. The question, whether patients with sepsis can be detected using this approach will be answered. Methods: In 2015, the University of Utah Healthcare implemented mEWS as prediction tool for sepsis. The automated calculation of mEWS is based on real-time data from the EMR. For this analysis, we extracted a cancer outpatient population from a 15-month time frame (from December 2016 to February 2018) within the EHR. We selected patients with a mEWS > 4, grouped these patients according to the process of care after mEWS scoring (e.g., admitted to ICU) and analyzed subgroup based upon mEWS score, discharge disposition and sepsis diagnosis. Results: Within the analyzed time frame 502 cancer outpatients had a mEWS score of 4 or higher. 88 of these patients (17.5%) were diagnosed with sepsis after mEWS screening. Out of the patients with sepsis, 22 were admitted to ICU, 63 were admitted to a medical floor and 3 were treated at the Huntsman Acute Care Clinic. Out of the 414 without a diagnosis of sepsis, 13 patients were admitted to ICU, 72 patients were admitted to a medical floor (in each case for other reasons then sepsis) and 329 patients were sent to home. Conclusions: Sepsis is a serious problem in cancer outpatient care. In our analysis, one in six patients with a mEWS score of 4 and higher were diagnosed with sepsis. This analysis has shown that the implementation of a real-time, EMR based scoring system like mEWS can support the early detection and treatment of sepsis in this population.
Physicians often fail to communicate well with patients. The objective of this retrospective controlled interrupted time series study was to evaluate the impact of a standardized communication intervention to improve physician communication. All patients ages 18 years or older (N = 7739 visits) admitted to University of Utah Health Care in Salt Lake City, Utah, from July 1, 2012, to June 31, 2014, were included. Obstetrics, rehabilitation, and psychiatric patients were excluded. The primary outcome was the percentage of patients who answered “Always” to all HCAHPS questions regarding physician-patient communication. Among the intervention group, the primary outcome increased from 56% to 63% ( P = .014, N = 1021) while remaining stable for the control group (65% to 66%, P = .6, N = 6718). The downward trend reversed after the intervention (−0.6% to +1.7% per month, P < .001). Standardized communication was associated with improvement in physician communication HCAHPS scores.
Residents in rural areas face barriers to accessing acute care. Rural home hospital (RHH) or delivery of acute care at home could represent an important clinical care model. This study assessed the feasibility and acceptability of RHH as a substitute to traditional hospital care. Patients were cared for by a remote RHH attending physician and an RHH registered nurse deployed to the home. The study team conducted daily check-ins with RHH clinicians to assess workflows for completion. Surveys assessed patient experience and qualitative interviews assessed perceived acceptability, safety, and quality of care. We completed qualitative analysis of the interviews and coded qualitative data into domains and subdomains through an iterative process. RHH was successfully deployed to three acutely ill patients in rural Utah. RHH admission, daily care, and discharge processes were accomplished for each patient. From qualitative analysis, we identified four domains: (1) Perceived comfort level during RHH admission, (2) Perceived safety during RHH admission, (3) Perceived quality of care during RHH admission, and (4) Perception of RHH workflows. We found acute care was delivered to rural homes with satisfactory patient and clinician experience. Team dynamics, technology build, robust clinical and operational workflows, and care coordination were important to a successful admission. Learnings from this study can inform program design and training for RHH teams and startup for larger RHH evaluation. Home hospital care is expanding rapidly in the United States and RHH could represent an important clinical care model.
Topic Significance & Study Purpose/Background/Rationale: Sepsis is a leading cause of death of hospitalized patients. Patients with cancer are at increased risk of morbidity and mortality from sepsis; infection is the leading cause of death in patients with hematologic malignancies or undergoing transplantation. Studies have shown delays in antibiotic therapy are associated with increased risk of death. In this study, we describe a project to improve early recognition and treatment for patients with sepsis who are admitted to Huntsman Cancer Hospital in the Blood and Marrow Transplant/Hematology Unit. Methods, Intervention, & Analysis: To identify sepsis reliably and avoid alert fatigue, we used a modified Early Warning Score (mEWS).Tabled 1Modified Early Warning Score (mEWS)Measure3210123Temperature≤35.035.1-35.535.6-38.038.1-39.039.1-40.9≥41.0Respiratory Rate≤89-1112-2021-2526-29≥30Pulse≤3031-3940-100101-110111-130≥131Systolic BP≤8081-9091-100101-180181-200201-220>221 Open table in a new tab •A score ≥4 was the trigger to alert providers for possible sepsis.•February-August 2016, we evaluated 55 autologous and allogeneic transplant patient records for time to antibiotic administration, antibiotic change, ICU transfer and mortality.•Results compared with a historical control group of similar patients with comparable mEWS scores prior to mEWS implementation. Findings & Interpretation:•A total of 169 mEWS alerts (score 4 and higher) were generated in 55 patients.•A total of 5/61 and 2/46 patients in the study and control groups were started on antibiotics (P = .04).•Change in antibiotics occurred in 47/61 (74.6%) of the study group compared with 16/46 (25.4%) in the control (P < .0001).•Time to antibiotic administration was significantly shorter in the study versus the control group (37 min versus 68.1 min, respectively, P < .0001).•The number of transfers to the ICU were similar (21 versus 15%, P = .04).•Similar proportion of deaths (5 and 7%, P = .7) in the study and the control groups. Discussion & Implications: mEWS implementation was associated with a significantly shorter time to antibiotic administration and higher number of antibiotic changes. Although other outcomes like mortality didn't seem to be affected in this BMT group, further evaluation of the system in a larger prospective cohort is warranted.
