Publications

BACKGROUND: Emergency medicine residents need to be staffed in a way that balances operational needs with their educational experience. Key to developing an optimal schedule is knowing a resident's expected productivity, a poorly understood metric.

OBJECTIVE: We sought to measure how a resident's busiest (peak) workload affects their overall productivity for the shift.

METHODS: We conducted a retrospective, observational study of resident productivity at an urban, tertiary care center with a 3-year Accreditation Council for Graduate Medical Education-approved emergency medicine training program, with 55,000 visits annually. We abstracted resident productivity data from a database of patient assignments from July 1, 2010 to June 20, 2015, utilizing a generalized estimation equation method to evaluate physician shifts. Our primary outcome measure was the total number of patients seen by a resident over a shift. The secondary outcome was the number of patients seen excluding those in the peak hour.

RESULTS: A total of 14,361 shifts were evaluated. Multivariate analysis showed that the total number of patients seen was significantly associated with the number of patients seen during the peak hour, level of training, the timing of the shift, but most prominently, lower variance in patients seen per hour (coefficient of variation < 0.10).

CONCLUSIONS: A resident's peak productivity can be a strong predictor of their overall productivity, but the substantial negative effect of variability favors a steadier pace. This suggests that resident staffing and patient assignments should generally be oriented toward a more consistent workload, an effect that should be further investigated with attending physicians.

OBJECTIVE: To demonstrate the incremental benefit of using free text data in addition to vital sign and demographic data to identify patients with suspected infection in the emergency department.

METHODS: This was a retrospective, observational cohort study performed at a tertiary academic teaching hospital. All consecutive ED patient visits between 12/17/08 and 2/17/13 were included. No patients were excluded. The primary outcome measure was infection diagnosed in the emergency department defined as a patient having an infection related ED ICD-9-CM discharge diagnosis. Patients were randomly allocated to train (64%), validate (20%), and test (16%) data sets. After preprocessing the free text using bigram and negation detection, we built four models to predict infection, incrementally adding vital signs, chief complaint, and free text nursing assessment. We used two different methods to represent free text: a bag of words model and a topic model. We then used a support vector machine to build the prediction model. We calculated the area under the receiver operating characteristic curve to compare the discriminatory power of each model.

RESULTS: A total of 230,936 patient visits were included in the study. Approximately 14% of patients had the primary outcome of diagnosed infection. The area under the ROC curve (AUC) for the vitals model, which used only vital signs and demographic data, was 0.67 for the training data set, 0.67 for the validation data set, and 0.67 (95% CI 0.65-0.69) for the test data set. The AUC for the chief complaint model which also included demographic and vital sign data was 0.84 for the training data set, 0.83 for the validation data set, and 0.83 (95% CI 0.81-0.84) for the test data set. The best performing methods made use of all of the free text. In particular, the AUC for the bag-of-words model was 0.89 for training data set, 0.86 for the validation data set, and 0.86 (95% CI 0.85-0.87) for the test data set. The AUC for the topic model was 0.86 for the training data set, 0.86 for the validation data set, and 0.85 (95% CI 0.84-0.86) for the test data set.

CONCLUSION: Compared to previous work that only used structured data such as vital signs and demographic information, utilizing free text drastically improves the discriminatory ability (increase in AUC from 0.67 to 0.86) of identifying infection.

The objective of this study was to understand the social elements of clinical and organizational interactions of the key stakeholders in the specific context of an electronic dashboard used by the emergency department (ED) and inpatient medicine teams at the time of clinical referral and handover. An electronic handover function is utilised at the ED-inpatient interface at this institution and has given clinicians the ability to better communicate, monitor the department and strive to improve patient safety in streamline the delivery of care in the acute phase. This study uses an ethnographic qualitative research design incorporating semistructured interviews, participant observation on the ED floor and fieldwork notes. The setting for this research was in the ED at a tertiary University affiliated hospital. Triangulation was used to combine information obtained from multiple sources and information from fieldwork and interviews refined into useable chunks culminating in a thematic analysis. Thematic analysis yielded five central themes that reflected how the clinical staff utilised this IT system and why it had become embedded in the culture of clinical referral and handover. Efficient time management for improved patient flow was demonstrated, value added communication (at the interpersonal level), the building trust at the ED-inpatient interface, the maintenance of mutual respect across medical cultures and an overall enhancement of the quality of ED communication (in terms of the information available). A robust electronic handover process, resulted in an integrated approach to patient care by removing barriers to admission for medical inpatients, admitted via ED. The value proposition for patients was a more complete information transfer, both within the ED and between departments.

BACKGROUND: Patient handoffs between units can introduce risk and time delays. Verbal communication is the most common mode of handoff, but requires coordination between different parties.

OBJECTIVE: We present an asynchronous patient handoff process supported by a structured electronic signout for admissions from the emergency department (ED) to the inpatient medicine service.

METHODS: A retrospective review of patients admitted to the medical service from July 1, 2011 to June 30, 2015 at a tertiary referral center with 520 inpatient beds and 57,000 ED visits annually. We developed a model for structured electronic, asynchronous signout that includes an option to request verbal communication after review of the electronic handoff information.

RESULTS: During the 2010 academic year (AY) all admissions used verbal communication for signout. The following academic year, electronic signout was implemented and 77.5% of admissions were accepted with electronic signout. The rate increased to 87.3% by AY 2014. The rate of transfer from floor to an intensive care unit within 24 h for the year before and 4 years after implementation of the electronic signout system was collected and calculated with 95% confidence interval. There was no statistically significant difference between the year prior and the years after the implementation.

CONCLUSIONS: Our handoff model sought to maximize the opportunity for asynchronous signout while still providing the opportunity for verbal signout when deemed necessary. The process was rapidly adopted with the majority of patients being accepted electronically.

STUDY OBJECTIVE: Resident productivity, defined as new patients per hour, carries important implications for emergency department operations. In high-volume academic centers, essential staffing decisions can be made on the assumption that residents see patients at a static rate. However, it is unclear whether this model mirrors reality; previous studies have not rigorously examined whether productivity changes over time. We examine residents' productivity across shifts to determine whether it remained consistent.

METHODS: This was a retrospective cohort study conducted in an urban academic hospital with a 3-year emergency medicine training program in which residents acquire patients ad libitum throughout their shift. Time stamps of all patient encounters were automatically logged. A linear mixed model was constructed to predict productivity per shift hour.

RESULTS: A total of 14,364 8- and 9-hour shifts were worked by 75 residents between July 1, 2010, and June 20, 2015. This comprised 6,127 (42.7%) postgraduate year (PGY) 1 shifts, 7,236 (50.4%) PGY-2 shifts, and 998 (6.9%) PGY-3 nonsupervisory shifts (Table 1). Overall, residents treated a mean of 10.1 patients per shift (SD 3.2), with most patients at Emergency Severity Index level 3 or more acute (93.8%). In the initial hour, residents treated a mean of 2.14 patients (SD 1.2), and every subsequent hour was associated with a significant decrease, with the largest in the second, third, and final hours.

CONCLUSION: Emergency medicine resident productivity during a single shift follows a reliable pattern that decreases significantly hourly, a pattern preserved across PGY years and types of shifts. This suggests that resident productivity is a dynamic process, which should be considered in staffing decisions and studied further.

BACKGROUND: Hypokalemia is a common clinical disorder caused by a variety of different mechanisms. Although the most common causes are diuretic use and gastrointestinal losses, elevated cortisol levels can also cause hypokalemia through its effects on the renin-angiotensin-aldosterone system. Cushing's syndrome refers to this general state of hypercortisolemia, which often manifests with symptoms of generalized weakness, high blood pressure, diabetes mellitus, menstrual disorders, and psychiatric changes. This syndrome is most commonly caused by exogenous steroid use, but other etiologies have also been reported in the literature. Ectopic adrenocorticotropic hormone production by small-cell lung cancer is one rare cause of Cushing's syndrome, and may be associated with significant hypokalemia.

CASE REPORT: We describe the case of a 62-year-old man who presented to the emergency department with weakness and hypokalemia. The patient was initially misdiagnosed with furosemide toxicity. Despite having a 30-pack-year smoking history, this patient's lack of respiratory complaints allowed him to present for medical attention twice before being diagnosed with lung cancer. It was later determined that this patient's hypokalemia was due to Cushing's syndrome caused by ectopic adrenocorticotropic hormone production from small-cell lung cancer. WHY SHOULD AN EMERGENCY PHYSICIAN BE AWARE OF THIS?: This case reminds emergency physicians to consider a broad differential when treating patients with hypokalemia. More importantly, it prompts emergency physicians to recognize comorbid conditions and secondary, less common etiologies in patients with repeated visits for the same complaint.

Idiopathic retinal vasculitis, aneurysms, and neuroretinitis (IRVAN) is a rare syndrome affecting the retinal and optic disc vasculature. Diffuse retinal ischemia, macular edema, and neovascularization may lead to bilateral vision loss. The authors report a case of a 36-year-old woman presenting with branch retinal artery occlusion (BRAO) in her right eye who was subsequently diagnosed with IRVAN syndrome. She was treated with panretinal photocoagulation for peripheral retinal ischemia and pars plana vitrectomy for vitreous hemorrhage. She later developed a BRAO in her left eye. This case demonstrates that BRAO may be a presenting feature of IRVAN syndrome. [Ophthalmic Surg Lasers Imaging Retina. 2017;48:948-951.].

BACKGROUND: Intubation in rural emergency departments (EDs) is a high-risk procedure, often with little or no specialty support. Rural EDs are utilizing real-time telemedicine links, connecting providers to an ED physician who may provide clinical guidance.

INTRODUCTION: We endeavored to describe telemedicine-assisted intubation in rural EDs that are served by an ED telemedicine network.

MATERIALS AND METHODS: Prospective data were collected on all patients who had an intubation attempt while on the video telemedicine link from May 1, 2014 to April 30, 2015. We report demographic information, indication, methods, number of attempts, operator characteristics, telemedicine involvement/intervention, adverse events, and clinical outcome by using descriptive statistics.

RESULTS: Included were 206 intubations. The most common indication for intubation was respiratory failure. First-pass success rate (postactivation) was 71%, and 96% were eventually intubated. Most attempts (66%) used rapid-sequence intubation. Fifty-four percent of first attempts used video laryngoscopy (VL). Telemedicine providers intervened in 24%, 43%, and 55% of first-third attempts, respectively. First-pass success with VL and direct laryngoscopy was equivalent (70% vs. 71%, p = 0.802). Adverse events were reported in 49 cases (24%), which were most frequently hypoxemia.

DISCUSSION: The impact of telemedicine during emergency intubation is not defined. We showed a 71% first-pass rate post-telemedicine linkage (70% of cases had a previous attempt). Our ultimate success rate was 96%, similar to that in large-center studies. Telemedicine support may contribute to success.

CONCLUSIONS: Telemedicine-supported endotracheal intubation performed in rural hospitals is feasible, with good success rates. Future research is required to better define the impact of telemedicine providers on emergency airway management.