Forecasting emergency department hourly occupancy using time series analysis

Study objectiveTo develop a novel predictive model for emergency department (ED) hourly occupancy using readily available data at time of prediction with a time series analysis methodology.MethodsWe performed a retrospective analysis of all ED visits from a large academic center during calendar year 2012 to predict ED hourly occupancy. Due to the time-of-day and day-of-week effects, a seasonal autoregressive integrated moving average with external regressor (SARIMAX) model was selected. For each hour of a day, a SARIMAX model was built to predict ED occupancy up to 4-h ahead. We compared the resulting model forecast accuracy and prediction intervals with previously studied time series forecasting methods.ResultsThe study population included 65,132 ED visits at a large academic medical center during the year 2012. All adult ED visits during the first 265 days were used as a training dataset, while the remaining ED visits comprised the testing dataset. A SARIMAX model performed best with external regressors of current ED occupancy, average department-wide ESI, and ED boarding total at predicting up to 4-h-ahead ED occupancy (Mean Square Error (MSE) of 16.20, and 64.47 for 1-hr- and 4-h- ahead occupancy, respectively). Our 24-SARIMAX model outperformed other popular time series forecasting techniques, including a 60% improvement in MSE over the commonly used rolling average method, while maintaining similar prediction intervals.ConclusionAccounting for current ED occupancy, average department-wide ESI, and boarding total, a 24-SARIMAX model was able to provide up to 4 h ahead predictions of ED occupancy with improved performance characteristics compared to other forecasting methods, including the rolling average. The prediction intervals generated by this method used data readily available in most EDs and suggest a promising new technique to forecast ED occupancy in real time.     

The Challenge of Predicting Demand for Emergency Department Services

Objectives:  The objective was to develop methodology for predicting demand for emergency department (ED) services by characterizing ED arrivals.
Methods:  One year of ED arrival data from an academic ED were merged with local climate data. ED arrival patterns were described; Poisson regression was selected to represent the count of hourly ED arrivals as a function of temporal, climatic, and patient factors. The authors evaluated the appropriateness of prediction models by whether the data met key Poisson assumptions, including variance proportional to the mean, positive skewness, and absence of autocorrelation among hours. Model accuracy was assessed by comparing predicted and observed histograms of arrival counts and by how frequently the observed hourly count fell within the 50 and 90% prediction intervals.
Results:  Hourly ED arrivals were obtained for 8,760 study hours. Separate models were fit for high- versus low-acuity patients because of significant arrival pattern differences. The variance was approximately equal to the mean in the high- and low-acuity models. There was no residual autocorrelation ( r  = 0) present after controlling for temporal, climatic, and patient factors that influenced the arrival rate. The observed hourly count fell within the 50 and 90% prediction intervals 50 and 90% of the time, respectively. The observed histogram of arrival counts was nearly identical to the histogram predicted by a Poisson process.
Conclusions:  At this facility, demand for ED services was well approximated by a Poisson regression model. The expected arrival rate is characterized by a small number of factors and does not depend on recent numbers of arrivals.     

ARIMA乘积季节模型在山西省结核病预测中的应用

目的 应用自回归移动平均（ARIMA）乘积季节模型对山西省2022和2023年结核病发病率进行预测,为结核病防控提供参考依据。方法 收集《中国疾病预防控制信息系统-结核病管理信息系统》2010—2021年山西省结核病月发病率数据,进行模型构建和检验。基于2010—2020年结核病月发病率数据使用R 4.1.0软件构建ARIMA乘积季节模型,并用2021年月发病率检验模型,同时预测山西省2022和2023年结核病流行趋势。结果 2010—2021年山西省共报告结核病患者191 517例,发病率由68.29/10万下降到23.74/10万,总体呈下降趋势。每年的1、2、10月发病率较低,3—6月发病率较高,尤其以冬春交替之际发病率最高。根据2010年1月至2020年12月结核病月发病率拟合出ARIMA(0,1,1)(1,1,1)12模型,该模型的赤迟信息量准则、均方根误差、平均绝对百分比误差和平均绝对误差分别为202.07、0.49、9.19、0.33。通过检验发现该模型的平均绝对百分比误差为11.34%,预测2022年山西省结核病发病率在0.51/10万～2.12/10万,2023年在0...     

Forecasting Daily Patient Volumes in the Emergency Department

Background: Shifts in the supply of and demand for emergency department (ED) resources make the efficient allocation of ED resources increasingly important. Forecasting is a vital activity that guides decision‐making in many areas of economic, industrial, and scientific planning, but has gained little traction in the health care industry. There are few studies that explore the use of forecasting methods to predict patient volumes in the ED. Objectives: The goals of this study are to explore and evaluate the use of several statistical forecasting methods to predict daily ED patient volumes at three diverse hospital EDs and to compare the accuracy of these methods to the accuracy of a previously proposed forecasting method. Methods: Daily patient arrivals at three hospital EDs were collected for the period January 1, 2005, through March 31, 2007. The authors evaluated the use of seasonal autoregressive integrated moving average, time series regression, exponential smoothing, and artificial neural network models to forecast daily patient volumes at each facility. Forecasts were made for horizons ranging from 1 to 30 days in advance. The forecast accuracy achieved by the various forecasting methods was compared to the forecast accuracy achieved when using a benchmark forecasting method already available in the emergency medicine literature. Results: All time series methods considered in this analysis provided improved in‐sample model goodness of fit. However, postsample analysis revealed that time series regression models that augment linear regression models by accounting for serial autocorrelation offered only small improvements in terms of postsample forecast accuracy, relative to multiple linear regression models, while seasonal autoregressive integrated moving average, exponential smoothing, and artificial neural network forecasting models did not provide consistently accurate forecasts of daily ED volumes. Conclusions: This study confirms the widely held belief that daily demand for ED services is characterized by seasonal and weekly patterns. The authors compared several time series forecasting methods to a benchmark multiple linear regression model. The results suggest that the existing methodology proposed in the literature, multiple linear regression based on calendar variables, is a reasonable approach to forecasting daily patient volumes in the ED. However, the authors conclude that regression‐based models that incorporate calendar variables, account for site‐specific special‐day effects, and allow for residual autocorrelation provide a more appropriate, informative, and consistently accurate approach to forecasting daily ED patient volumes.     

基于自回归移动平均模型的浙江省肺结核发病趋势预测

  目的   应用自回归移动平均（ARIMA）模型对浙江省肺结核疫情预测分析,为浙江省肺结核精准化防控工作提供科学依据。   方法   收集2011年1月至2021年8月的浙江省肺结核发病率数据, 基于R软件（4.0.3）利用2011 — 2020年肺结核发病率数据建立ARIMA模型,比较2021年1—8月预测数据和实际数据并选择最优模型。   结果   2011年1月至2020年12月浙江省报告新发肺结核患者总计374 718例,呈逐年下降趋势,每年12月至次年2月发病率较低,3 — 5月相对较高。 确定最优模型为ARIMA（2,1,0）（1,1,2）₁₂,该模型拟合的2021年1 — 8月浙江省肺结核发病率预测值与真实值的平均相对误差为8.87%,赤池信息准则值、贝叶斯信息准则值、均方根误差值和平均绝对百分比误差值分别为95.02、111.05、0.30和4.39。  结论   ARIMA（2,1,0）（1,1,2）₁₂模型能较好地拟合预测浙江省肺结核发病率在时间序列上的变动趋势,但需根据实际情况动态调整,提高预测精度。     

自回归移动平均模型在全国手足口病疫情预测中的应用

目的探讨应用时间序列基于季节性差分的自回归移动平均模型(autoregressive integrated moving average,ARIMA)预测全国手足口病的发病情况。方法利用"中国疾病预防控制信息系统"中的"疾病监测信息报告管理系统"(又称"传染病疫情信息网络直报系统")的资料,应用SPSS 19.0统计软件、采用ARIMA,对全国2009年1月至2012年12月手足口病逐月发病情况进行建模和拟合,利用所得到的模型对2013年1-6月的发病情况进行预测,并评价其预测效果。结果分析结果显示,手足口病发病以年为周期,1年中5-6月为高发月。非季节移动平均参数滞后两次后为0.532,t检验的P值为0.003,差异有统计学意义。BIC=21.955,Ljung-Box统计量检验残差序列为白噪声序列。最佳ARIMA(0,1,2),(0,1,0)12预测的平均相对误差为0.52,预测效果一般。按照不同发病模式分为两层后分别建立ARIMA,平均相对误差为0.12,预测效果好。结论对监测数据进行时间序列分析是用于传染病预测的一个重要的工具。分析发现中国不能用一个ARIMA拟合手足口病资料,因地区间发病的变异和模式不同;按手足口病的发病模式将各省分为单峰和双峰两层,分别拟合ARIMA,模型拟合效果更好。     

The Effect of Hospital Occupancy on Emergency Department Length of Stay and Patient Disposition

Emergency department (ED) overcrowding is a common problem. Despite a widespread belief that low hospital bed availability contributes to ED overcrowding, there are few data demonstrating this effect. OBJECTIVES: To identify the effect of hospital occupancy on ED length of stay for admitted patients and patient disposition. METHODS: This was an observational study design using administrative data at a 500-bed acute care teaching hospital. All patients presenting to the ED between April 1993 and June 1999 were included in the study. The predictor variable was daily hospital occupancy. Outcome measures included daily ED length of stay for admitted patients, daily consultation rate, and daily admission rate. The models controlled for the average daily age of ED patients and the average daily "arrival density" index, which adjusts for patient volume and clustering of patient arrivals. RESULTS: The average hospital occupancy was 89.7%. On average 155 patients visited the ED daily; 21% were referred to hospital physicians and 19% were admitted. The median ED length of stay for admitted patients was 5 hours 54 minutes (interquartile range 5 hr 12 min to 6 hr 42 min). Daily ED length of stay for admitted patients increased 18 minutes (95% CI = 12 to 24) when there was an absolute increase in occupancy of 10%. The ED length of stay appeared to increase extensively when hospital occupancy exceeded a threshold of 90%. Consultation and admission rates were not influenced by hospital occupancy. CONCLUSIONS: Increased hospital occupancy is strongly associated with ED length of stay for admitted patients. Increasing hospital bed availability might reduce ED overcrowding.     

Distances to Emergency Department and to Primary Care Provider's Office Affect Emergency Department Use in Children

Objectives:  Patients of all ages use emergency departments (EDs) for primary care. Several studies have evaluated patient and system characteristics that influence pediatric ED use. However, the issue of proximity as a predictor of ED use has not been well studied. The authors sought to determine whether ED use by pediatric Medicaid enrollees was associated with the distance to their primary care providers (PCPs), distance to the nearest ED, and distance to the nearest children's hospital.
Methods:  This historical cohort study included 26,038 children age 18 and under, assigned to 332 primary care practices affiliated with a Medicaid health maintenance organization (HMO). Predictor variables were distance from the child's home to his or her PCP site, distance from home to the nearest ED, and distance from home to the nearest children's hospital. The outcome variable was each child's ED use. A negative binomial model was used to determine the association between distance variables and ED use, adjusted for age, sex, and race, plus medical and primary care site characteristics previously found to influence ED use. Distance variables were divided into quartiles to test for nonlinear associations.
Results:  On average, children made 0.31 ED visits/person/year. In the multivariable model, children living greater than 1.19 miles from the nearest ED had 11% lower ED use than those living within 0.5 miles of the nearest ED (risk ratio [RR] = 0.89, 95% CI = 0.81 to 0.99). Children living between 1.54 and 3.13 miles from their PCPs had 13% greater ED use (RR = 1.13, 95% CI = 1.03 to 1.24) than those who lived within 0.7 miles of their PCP.
Conclusions:  Geographical variables play a significant role in ED utilization in children, confirming the importance of system-level determinants of ED use and creating the opportunity for interventions to reduce geographical barriers to primary care.     

基于自回归移动平均模型各亚型流行性感冒流行特征与趋势预测

目的 分析我国2010—2019年流行性感冒的流行特征和分布规律,预测各亚型流感发病趋势。方法 采用ARIMA乘积季节模型,对流感数据进行原始序列预处理、模型识别、参数估计和统计建模,预测流感发病趋势。结果 构建流感自回归移动平均模型(ARIMA)乘积季节模型,预测模型为ARIMA(1,2,1)(0,1,1)₁₂,数据信息提取充分(Q=14.257,P>0.05),相对误差约10%;甲型流感预测模型为ARIMA(2,1,1)(0,2,2)₁₂,数据信息提取充分(Q=13.236,P>0.05),预测2018年12月至2019年3月的甲型流感发病率较高,4月开始,发病率迅速下降,与实际情况相似,相对误差控制在10%以内;乙型流感预测模型为ARIMA(1,2,1)(1,0,1)₁₂,数据信息提取充分(Q=9.841,P>0.05),但模型预测2019年乙型流感发病率较低,相对误差较高。结论 流感、甲型流感ARIMA乘积季节模型预测效果较好;乙型流感预测模型数据信息提取充分,但相对误差较高,可能与乙型流感发病...     

Impact of Disaster Drills on Patient Flow in a Pediatric Emergency Department

Objectives:  Joint Commission on Accreditation of Healthcare Organizations (JCAHO)-accredited hospitals must conduct disaster drills twice a year, with one incorporating a mass casualty incident to the emergency department (ED). The authors found no studies describing the potential negative impact on the quality of care real patients in the ED receive during these drills. The objective was to determine the impact that mass casualty drills have on the timeliness of care for nondisaster patients in a pediatric ED.
Methods:  Since 2001, nine disaster drills involving mass casualties to the ED were conducted at the authors' institution. The authors studied 5-, 10-, and 24-hour blocks of time surrounding these events and defined quality measures as the timeliness of care in terms of length of stay (LOS) in ED, time-to-triage, and time-to-physician. Drill dates were compared with control dates (the same weekday on the following week). Paired t-tests were used to compare outcomes of interest between drill and control days.
Results:  Nine drill days and nine control days were studied. There was no statistically significant difference between drill dates and control dates in average time-to-triage and time-to-emergency physician and average ED LOS. Admitted patients spent less time in the ED during drill dates.
Conclusions:  Disaster drills at this institution do not appear to significantly affect the timeliness of care to nondisaster drill ED patients. Attention should be paid to the quality of care "real" patients receive to ensure that their care is not jeopardized during an artificial stress to the system during a disaster drill.     

Emergency Department Patient Flow: The Influence of Hospital Census Variables on Emergency Department Length of Stay

Objectives:  The objective was to evaluate the association between hospital census variables and emergency department (ED) length of stay (LOS). This may give insights into future strategies to relieve ED crowding.
Methods:  This multicenter cohort study captured ED LOS and disposition for all ED patients in five hospitals during five 1-week study periods. A stepwise multiple regression analysis was used to examine associations between ED LOS and various hospital census parameters.
Results:  Data were analyzed on 27,325 patients on 161 study days. A significant positive relationship was demonstrated between median ED LOS and intensive care unit (ICU) census, cardiac telemetry census, and the percentage of ED patients admitted each day. There was no relationship in this cohort between ED LOS and ED volume, total hospital occupancy rate, or the number of scheduled cardiac or surgical procedures.
Conclusions:  In multiple hospital settings, ED LOS is correlated with the number of admissions and census of the higher acuity nursing units, more so than the number of ED patients each day, particularly in larger hospitals with busier EDs. Streamlining ED admissions and improving availability of inpatient critical care beds may reduce ED LOS.     

Estimating Observation Unit Profitability with Options Modeling

Background:  Over the past two decades, the use of observation units to treat such common conditions as chest pain, asthma, and others has greatly increased. These units allow patients to be directed out of emergency department (ED) acute care beds while potentially avoiding inpatient admission. Many studies have demonstrated the clinical effectiveness of care delivered in such a setting compared to the ED or inpatient ward. However, there are limited data published about observation unit finance.
Methods:  Using the economic principles of stock options, opportunity costs, and net present value (NPV), a model that captures the value generated by admitting a patient to an observation unit was derived. In addition, an appendix is included showing how this model can be used to calculate the dollar value of an observation unit admission.
Results:  A model is presented that captures more complexity of observation finance than the simple difference between payments and costs. The calculated estimate in the Appendix suggests that the average value of a single observation unit admission was about $2,908, which is about 40% higher than expected.
Conclusion:  Subtraction of costs from payments may significantly underestimate the financial value of an observation unit admission. However, the positive value generated by an observation unit bed must be considered in the context of other projects available to hospital administrators.     

Introduction of a Stat Laboratory Reduces Emergency Department Length of Stay

Objectives:  Emergency department (ED) length of stay (LOS) impacts patient satisfaction and overcrowding. Laboratory turnaround time (TAT) is a major determinant of ED LOS. The authors determined the impact of a Stat laboratory (Stat lab) on ED LOS. The authors hypothesized that a Stat lab would reduce ED LOS for admitted patients by 1 hour.
Methods:  This was a before-and-after study conducted at an academic suburban ED with 75,000 annual patient visits. All patients presenting to the ED during the months of August and October 2006 were considered. A Stat lab located within the central laboratory was introduced in September 2006 to reduce laboratory TAT. The test TATs and ED LOS before (August 2006) and after (October 2006) implementing the Stat lab for all ED patients were the data of interest. ED LOS before and after the Stat lab was introduced was compared with the Mann-Whitney U-test. A sample size of 5,000 patients in each group had 99% power to detect a 1-hour difference in ED LOS.
Results:  There were 5,631 ED visits before and 5,635 visits after implementing the Stat lab. Groups were similar in age (34 years vs. 36 years) and gender (51% males in both). The percentages of patients with laboratory tests before and after Stat lab implementation were 68.7 and 71.3%, respectively. Test TATs for admitted patients were significantly improved after the Stat lab introduction. Implementation of the Stat lab was associated with a significant reduction in the median ED LOS from 466 (interquartile range [IQR] = minutes before to 402 (IQR = 296–553) minutes after implementing the Stat lab. The effects of the Stat lab on ED LOS were less marked for discharged patients.
Conclusions:  Introduction of a Stat lab dedicated to the ED within the central laboratory was associated with shorter laboratory TATs and shorter ED LOS for admitted patients, by approximately 1 hour.     

时序预测模型对广州市急救需求量的预测价值北大核心CSCD

目的探讨时序预测模型中的差分自回归滑动平均(ARIMA)和自回归(AR)模型在预测广州市急救调度日出车数量方面的价值。方法采用Matlab仿真软件对广州市2021年1月1日至2021年12月31日的急救调度出车记录分析计算日出车数量时间序列,对该序列进行时序预测模型辨识,得到ARIMA(1,1,1)、AR(4)以及AR(7)模型,利用这些模型对日出车数量做出预测拟合。ARIMA(1,1,1)模型将数据分为训练集和测试集,参数运算采用Prony方法,预测拟合未来的出车数量;AR(4)和AR(7)模型采用均匀系数,预测当天出车数量。结果ARIMA(1,1,1)、AR(4)以及AR(7)都可以实现对日出车数量的有效预测,ARIMA(1,1,1)的预测拟合误差随着预测时间的延长下降。两个月内的急救调度日出车量预测拟合平均绝对百分比误差(MAPE)低于6%,结果基本都位于95%置信区间内,利用模型的残差分析验证了模型显著有效。结论ARIMA模型可以对两个月内的急救调度日出车量做长期预测拟合,AR模型可以对急救调度日出车量做短期有效预测。     

Reduced access block causes shorter emergency department waiting times: An historical control observational study

Objective: To study the effect of changes in hospital occupancy and ED occupancy on ED waiting times during a 13‐day period of improved bed access. Methods: A comparative, observational study of 1133 ED attendances in the study period and 2332 attendances in a historical control period. Results: During the study period, mean hospital occupancy decreased from 94.9% to 89.0% (P < 0.001), mean ED occupancy decreased from 19.1 to 14.8 patients (P < 0.001) and the mean ED waiting time decreased from 58.5 to 37.1 min (P < 0.001). There were statistically significant reductions in waiting times for patients in Australasian triage scale (ATS) categories 2–5. Departmental staffing levels, attendances and patient acuity were not significantly different during the study and control periods. Conclusions: Modest decreases in hospital occupancy resulted in highly significant reductions in ED waiting times. Emergency department overcrowding due to large numbers of admitted patients awaiting hospital admission is a major cause of ED dysfunction.     

Clinical Significance of Brush Allodynia in Emergency Patients With Migraine

Background.— Cutaneous brush allodynia may be a practical and readily assessable marker of progression of an acute migraine attack. We determined the relative frequency of this finding in emergency department (ED) patients with acute migraine and tested the hypothesis that the presence of cutaneous brush allodynia prior to initial treatment in the ED could predict poor 2-hour and 24-hour pain intensity outcomes.
Methods.— As part of a multicenter ED-based clinical trial testing the benefit of dexamethasone vs placebo for the adjuvant parenteral treatment of acute migraine, cutaneous brush allodynia was assessed prior to treatment using an established methodology. In addition to dexamethasone or placebo, all patients received intravenous metoclopramide + diphenhydramine as primary treatment for their migraine. Pain intensity outcomes were assessed in the ED 2 hours after medication administration and again by telephone 24 hours after medication administration.
Results.— An assessment of cutaneous brush allodynia was performed in 182 migraineurs from 3 different EDs, of whom 26 (14%, 95% CI: 10-20%) had cutaneous brush allodynia. A pain-free state within 2 hours of medication administration was achieved by 46% of the allodynic patients and by 47% of the nonallodynic patients ( P  = .91). Median headache intensity over the 24 hours after ED discharge, as measured on a pain intensity scale from zero to 10, was 3 in the allodynic patients and 3 in the nonallodynic patients ( P  = .23).
Conclusions.— Cutaneous brush allodynia is an uncommon finding in the ED, occurring in fewer than 1 in 5 migraineurs. It does not seem to have prognostic relevance for the ED-based management of the acute migraine attack.     

Performance of time-series methods in forecasting the demand for red blood cell transfusion

BACKGROUND: Planning the future blood collection efforts must be based on adequate forecasts of transfusion demand. In this study, univariate time-series methods were investigated for their performance in forecasting the monthly demand for RBCs at one tertiary-care, university hospital. STUDY DESIGN AND METHODS: Three time-series methods were investigated: autoregressive integrated moving average (ARIMA), the Holt-Winters family of exponential smoothing models, and one neural-network-based method. The time series consisted of the monthly demand for RBCs from January 1988 to December 2002 and was divided into two segments: the older one was used to fit or train the models, and the younger to test for the accuracy of predictions. Performance was compared across forecasting methods by calculating goodness-of-fit statistics, the percentage of months in which forecast-based supply would have met the RBC demand (coverage rate), and the outdate rate. RESULTS: The RBC transfusion series was best fitted by a seasonal ARIMA(0,1,1)(0,1,1)(12) model. Over 1-year time horizons, forecasts generated by ARIMA or exponential smoothing laid within the +/- 10 percent interval of the real RBC demand in 79 percent of months (62% in the case of neural networks). The coverage rate for the three methods was 89, 91, and 86 percent, respectively. Over 2-year time horizons, exponential smoothing largely outperformed the other methods. Predictions by exponential smoothing laid within the +/- 10 percent interval of real values in 75 percent of the 24 forecasted months, and the coverage rate was 87 percent. CONCLUSION: Over 1-year time horizons, predictions of RBC demand generated by ARIMA or exponential smoothing are accurate enough to be of help in the planning of blood collection efforts. For longer time horizons, exponential smoothing outperforms the other forecasting methods.     

Short-term cloud coverage prediction using the ARIMA time series model

In view of the important role of cloud coverage on the solar (energy) irradiance, the total cloud coverage prediction based on ground-based cloud images is studied in this paper. In traditional prediction techniques, the correlation between cloud coverage over continue time is always neglected. Thus, an autoregressive integrated moving average (ARIMA) time series model is used to predict the short-term cloud coverage. Experimental results on a collected time series database of cloud coverage computed from ground-based cloud images show that the correlation information of time series is useful for cloud coverage prediction. Additionally, the ARIMA model gains a superior prediction performance for forecasts of one minute or longer 20 and 30 minutes. We are able to predict the cloud coverage with an approximate error of 5%, 7%, and 9% for 1, 5, and 20 and 30 minute forecasts, respectively. Furthermore, we found that there are different error rates of predictions for different cloud coverage intervals. High cloud coverage always suffers from a higher error rate.     

RVU Ready? Preparing Emergency Medicine Resident Physicians in Documentation for an Incentive‐based Work Environment

Objectives:  The emergency medicine (EM) job market is increasingly focused on incentive-based reimbursement, which is largely based on relative value units (RVUs) and is directly related to documentation of patient care. Previous studies have shown a need to improve resident education in documentation. The authors created a focused educational intervention on billing and documentation practices to meet this identified need. The hypothesis of this study was that this educational intervention would result in an increase in RVUs generated by EM resident physicians and the average amount billed per patient.
Methods:  The authors used a quasi-experimental study design. An educational intervention included a 1-hour lecture on documentation and billing, biweekly newsletters, and case-specific feedback from the billing department for EM resident physicians. RVUs and charges generated per patient were recorded for all second- and third-year resident physicians for a 3-month period prior to the educational intervention and for a 3-month period following the intervention. Pre- and postintervention data were compared using Student's t-test and repeated-measures analysis of variance, as appropriate.
Results:  The evaluation and management (E/M) chart levels billed during each phase of the study were significantly different (p < 0.0001). The total number of RVUs generated per hour increased from 3.17 in the first phase to 3.71 in the second phase (p = 0.0001). During the initial 3-month phase, the average amount billed per patient seen by a second- or third-year resident was $282.82, which increased to $301.94 in the second phase (p = 0.0004).
Conclusions:  The educational intervention positively affected resident documentation resulting in greater RVUs/hour and greater billing performance in the study emergency department (ED).     

ARIMA乘积季节模型在盐城市手足口病疫情预测中的应用

目的 探讨自回归移动平均（autoregressive integrated moving average model,ARIMA）乘积季节模型在盐城市手足口病发病趋势预测的可行性。方法 利用盐城市2009年1月至2015年12月的手足口病月发病率建立ARIMA乘积季节模型,并对2016年手足口病发病趋势进行预测。结果 盐城市手足口病预测模型为ARIMA（1,0,1）（1,1,0）12,该模型的参数估计具有统计学意义,拟合优度检验统计量最小Normalized BIC=2.997,残差序列检验统计量Ljung-Box=20.692（P0.05）,残差为白噪声,模型能够拟合出手足口病的发病趋势,且实际值都在95%可信区间内,但模型拟合的平均误差率为41.296%,检验模型预测效果的平均误差率为23.998%,模型预测精度高于拟合精度。结论 运用ARIMA乘积季节模型能够对盐城市手足口病发病趋势进行预测和动态分析,对手足口病预防控制产生积极的指导作用。