a b s t r a c t

Background: Demographic shifts and care delivery system evolution affect the number of Emergency Department (ED) visits and associated costs. Recent aggregate trends in ED visit rates and charges between 2010 and 2016 have not been evaluated.

Methods: Data from the National Emergency Department Sample, comprising approximately 30 million annual patient visits, were used to estimate the ED visit rate and charges per visit from 2010 to 2016. ED visits were grouped into 144 mutually exclusive clinical categories. Visit rates, compound annual growth rates (CAGRs), and per visit charges were estimated.

Results: From 2010 to 2016, the number of ED visits increased from 128.97 million to 144.82 million; the cumu- lative growth was 12.29% and the CAGR was 1.95%, while the population grew at a CAGR of 0.73%. Expressed as a population rate, ED visits per 1000 persons increased from 416.92 in 2010 to 448.19 in 2016 (p value b0.001). The mean charges per visit increased from $2061 (standard deviation $2962) in 2010 to $3516 (standard deviation

$2962) in 2016; the CAGR was 9.31% (p value b0.001). Of 144 clinical categories, 140 categories had a CAGR for mean charges per visit of at least 5%.

Conclusion: The rate of ED visits per 1000 persons and the mean charge per ED visit increased significantly be- tween 2010 and 2016. Mean charges increased for both high- and low-acuity clinical categories. Visits for the 5 most common clinical categories comprise about 30% of ED visits, and may represent focus areas for increasing the value of ED care.

(C) 2019

Introduction

Care in Emergency Departments (EDs) has been estimated to repre- sent five to 6% – approximately $200 billion – of United States (US) health care spending [1]. One recent study indicated that emergency care is growing faster than other areas of healthcare spending, with a 6.4% annualized growth rate [2]. Growth in healthcare spending may be due to demographic factors, care delivery system-related factors, or both. Demographic factors, including population size changes and shifts in the age composition of the population, are not amenable to interven- tion. Care delivery system-related factors, which could possibly be af- fected by public policy, include: the number of ED visits, the intensity of care (testing and therapies) delivered in the ED, and prices charged for the care.

? Source of support: None.Name of organization and date of assembly if the article has been presented: Not applicable.

* Corresponding author at: 231 Albert Sabin Way ML 0769, Cincinnati, OH 45267-0769, USA.

E-mail address: hookerea@ucmail.uc.edu (E. Hooker).

Factors associated with increased volume of ED visits in recent years include the Patient Protection and Affordable Care Act, the expanding role EDs play in hospital admissions, and the development of freestand- ing EDs [3-7]. Although the impact is not clear, the expansion of alterna- tive sites of care, including urgent care centers and retail clinics, are intended to decrease the utilization of EDs [8-10].

The availability of expensive technology, especially Cross-sectional imaging, has led to increased intensity of care and higher cost [11]. This has resulted in efforts, like American College of Emergency Physi- cians’ (ACEP) choosing wisely campaign and others, to decrease the use of unnecessary cross-sectional imaging and low-value or inappro- priate care [11-13]. Evidence suggests that significant variation exists in visit intensity, but the proportion of such variation that is unwar- ranted is unknown [14].

Patients, policy makers and clinicians, have questioned the high cost for ED care and the wide variation in cost between healthcare organiza- tions [15,16]. Especially concerning are increases in the cost of care for low-acuity complaints (where clinical interventions are less likely to have changed) over the last 20-30 years [17,18].

The Centers for Disease Control and Prevention (CDC) estimated that there were 145.5 million ED visits in 2016, which was the highest ever

https://doi.org/10.1016/j.ajem.2019.158423

0735-6757/(C) 2019

reported for a single year [19]. However, associated ED charges and un- derlying trends in common clinical complaints were not reported. The objective of this study was to utilize the National Emergency Depart- ment Sample (NEDS) database in order to better characterize the trends in ED visits and the associated charges.

Materials and methods

Data

We performed a retrospective analysis using the Healthcare Cost and Utilization Project’s (HCUP) National Emergency Department Sam- ple (NEDS) datasets for 2010-2016. HCUP is managed by the Agency for Health Care Research and Quality (AHRQ) [20,21]. The NEDS datasets are the largest all-payer publicly available database in the US for ED visits, with approximately 30 million patient visits per year from nearly 1000 community hospitals across 35 participating states [21]. HCUP produces weights for each year and hospital to generate nationally rep- resentative estimates of ED visits. The NEDS data sets used the AHRQ- developed Clinical Classifications Software (CCS) to summarize the pri- mary reason for the visit into 263 mutually-exclusive categories using International Classification of Diseases 9th or 10th Revision (ICD-9/10) Diagnosis codes [22,23]. From a resource planning and utilization per- spective, many of the CCS categories are similar and we used a previ- ously described CCS roll-up (CCS-RU) that condensed the categories into 144 mutually exclusive categories based on clinical similarity from the ED perspective [24]. Data included the number of ED visits, hospital admissions, patient demographics, visit outcomes, and hospital charges.

Data analysis

Survey analysis techniques were used to account for clustering and stratification of visits for all continuous and categorical variables per AHRQ recommendations. National estimates of the visits and charges were created using the sampling weights provided by NEDS [25]. The weighted estimated ED visits per 1000 persons were calculated using US Census Bureau figures for the overall number of visits, diagnosis, dis- position from ED, and visit mortality rate [26]. The American Commu- nity Survey (ACS) and the Centers for Medicare and Medicaid Services enrollment estimates were used to calculate the rate of ED visits for sex, age, and primary insurance cohorts [26,27]. Hospital charges were used as a proxy of healthcare utilization. The charges represent the amount billed by the hospital and do not reflect the actual cost or re- imbursement for the visit. The average, median, standard deviation, and inter-quartile range (IQR) were reported for hospital charge data.

The compound annual growth rate (CAGR) was used to estimate trends over time with the following formula:

1

Ending Value # -1

increased from 416.92 in 2010 to 448.19 in 2016 (p value b0.001), which represents a cumulative growth rate of 7.51% and a CAGR of 1.21% compared to the population cumulative growth rate of 4.46% and CAGR of 0.73% (Fig. 1).

Females visited the ED at a higher rate (per 1000 persons) than males in each year. Female visits grew from 455.09 to 487.02, representing a 1.14% CAGR (p value 0.009). Male visits grew from

379.01 to 406.52, representing a 1.17% CAGR (p value 0.038). The rate of ED visits by age cohort did not increase significantly, with the excep- tion of 45-64 year-olds (355.57 to 401.66, p value 0.007). There was a significant increase in the rate of visits by patients with private insur- ance (240.37 to 261.33; p value 0.013), Medicaid (594.50 to 641.92; p value 0.048), and Other insurance status (594.80 to 658.12; p value 0.025). Routine discharges increased from 335.05 in 2010 to 368.30 in 2016 (CAGR 1.59%; p value 0.015). There was no significant change in the rate of admission to the same hospital from 2010 to 2016 (-1.39% CAGR; p value 0.067).

The overall charges per visit are shown in Table 2 and Fig. 2. Charges increased by 56% between 2010 and 2016 (Fig. 1). The average charge per ED visit increased from $2061 (Std. dev. $2962) in 2010 to $3516 (St. dev. $2962) in 2016, representing a CAGR of 9.31% (p value b0.001). The corresponding median charge per visit increased from

$1118 in 2010 to $1959 in 2016, representing a CAGR of 8.82%. Between 2010 and 2016 the spread in the IQR increased $1183, indicating a greater variance in the charge per visit.

In 2016, hospitals located in urban areas (specifically, metropolitan areas by Census Bureau designation), with a teaching status, or with a trauma center designation experienced a greater proportion of visits than in 2010 compared to hospitals in micropolitan or rural areas, lack- ing teaching status, or lacking a trauma center designation, respectively (Table 3).

The top five reasons for visiting the ED in 2016 were: Abdominal Pain (53.24 per 1000 persons); Upper Respiratory infections and Phar- yngitis (20.39 per 1000 persons); Chest Pain, Myocardial Infarction, and Heart Disease (19.65 per 1000 persons); Mental Health and Sub- stance Abuse (19.56 per 1000 persons); and Superficial Injuries (17.18 per 1000 persons). Superficial Injuries was the only reason for a visit to show a decline in the rate of visits between 2010 and 2016 (19.11 to 17.18 per 1000; -1.76% CAGR; p value b0.001). However, the aver- age charge per visit for Superficial Injuries increased by $1117 to

$2715 in 2016 (CAGR 9.24%, p value b0.001). Chest Pain, Myocardial In- farction, and Heart Disease had the highest charge per visit of $7096 (Std. dev. $9679), and had a CAGR for visits of 2.05% (p value b0.001). The CAGR for mean charges in this group was 7.52% (p value b0.001).

All five of the top reasons for an ED visit varied significantly from the overall average charge per visit (Table 4). upper respiratory infections (URI) and Pharyngitis had the lowest average of $1481 (p value b0.001), which was $2035 less than the average visit. As noted previ- ously, Chest Pain, Myocardial Infarction, and Heart Disease had the highest average charge of the top reasons and was $3580 (p value

b0.001) higher than the overall visit.

CAGR 1/4

Beginning Value

of years

Out of 144 CCS-RU clinical categories, 140 had a CAGR for mean charges of at least 5%, and 60 had a CAGR of over 10% (see Supplemen-

Weighted Linear regression models were used to test the trend over time for the overall rate of ED visits and top five reasons for ED visits, as well as for average charges [28,29]. For clarity, only data from 2010 and 2016 were shown. Statistical significance was determined using a two- tailed p-value b0.05. All analyses were conducted using SPSS 24.0 (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp).

Results

The estimated number of total ED visits increased from 128.97 mil- lion in 2010 to 144.82 million in 2016, representing a CAGR of 1.95% (Table 1). Expressed as a population rate, ED visits per 1000 persons

tary Table).

Discussion

We found that the rate of ED visits increased at a CAGR of 1.21% from 2010 to 2016, which was faster than the CAGR for the population (0.73%). During the same periods, charges for ED care grew by 56%, representing a CAGR of 9.31%. Increasing utilization and charges for care are inconsistent with policymakers’ efforts to reduce healthcare spending growth.

Stratification by age showed only a single group, 45-64 year olds, with a statistically significant p-value over the 6-year period (CAGR 2.05%). Broad care delivery system factors rather than shifts unique to

Table 1

Patient characteristics of emergency department visits, 2010-2016.

Weighted No. of Visits, Millions (95% CI) Weighted Estimate of Total Annual Visits, % (95%

CI)

Weighted Estimated ED Visits per 1000 Population (95% CI)

CAGR P Value for Trend

1578

B.H. Lane et al. / American Journal of Emergency Medicine 38 (2020) 1576–1581

	2010		2016			2010		2016			2010		2016
Total Visits	128.97	(123.57-134.37)	144.82	(137.34-152.32)		100		100			416.92	(399.47-434.37)	448.19	(425.01-471.36)		1.21%		b 0.001
Sex (1)
Male	57.53	(55.21-59.84)	64.45	(61.19-67.72)		44.61	(44.37-44.85)	44.51	(44.55-44.46)		379.01	(363.75-394.27)	406.52	(385.92-427.11)		1.17%		0.038
Female	71.43	(68.32-74.54)	80.37	(76.11-84.63)		55.39	(55.15-55.63)	55.49	(55.42-55.56)		455.09	(435.25-474.92)	487.02	(461.19-512.85)		1.14%		0.009
Age, years (1)
b5	11.44	(10.49-12.39)	12.38	(10.75-14.02)		8.87	(8.14-9.61)	8.55	(7.42-9.68)		566.72	(519.73-613.72)	620.08	(538.34-701.82)		1.51%		0.843
5-17	14.07	(13.21-14.92)	16.41	(14.58-18.25)		10.91	(10.25-11.57)	11.33	(10.62-11.98)		260.86	(245.01-276.71)	264.18	(234.69-293.67)		0.21%		0.289
18-44	52.30	(49.82-54.78)	55.06	(51.10-59.02)		40.56	(39.83-41.30)	38.01	(37.21-38.75)		462.97	(440.97-484.97)	512.05	(475.24-548.85)		1.69%		0.382
45-64	29.09	(27.86-30.31)	33.84	(31.58-36.10)		22.55	(22.14-22.97)	23.36	(22.99-23.70)		355.67	(340.73-370.61)	401.66	(374.78-428.54)		2.05%		0.007
65 +	22.06	(21.02-23.10)	27.13	(25.30-28.97)		17.11	(16.55-17.69)	18.73	(18.42-19.02)		545.07	(519.38-570.76)	551.01	(513.76-588.25)		0.18%		0.412
Primary Insurance (2)
Medicare	26.93	(25.75-28.11)	33.40	(31.58-35.22)	20.96		(20.40-21.52)	23.09	(23.02-23.14)	564.91		(540.15-589.68)	588.06	(556.09-620.04)	0.67%		0.398
Medicaid	32.48	(30.59-34.34)	46.73	(43.70-49.80)	25.27		(24.40-26.16)	32.30	(21.89-32.70)	594.50		(560.17-628.83)	641.92	(600.33-683.51)	1.29%		0.048
Private	40.17	(37.99-42.35)	41.38	(38.92-43.94)	31.27		(30.30-32.26)	28.60	(28.37-28.81)	240.37		(227.35-253.39)	261.33	(245.78-276.88)	1.40%		0.013
Other	28.90	(26.65-31.17)	23.16	(20.95-25.38)	22.50		(21.19-24.01)	16.01	(15.27-16.68)	594.80		(548.27-641.33)	658.12	(595.18-721.05)	1.70%		0.025
Disposition from ED (3)
Routine	103.64	(99.14-108.15)	119.01	(112.79-125.22)	80.36		(79.81-80.90)	82.16	(77.87-86.45)	335.05		(320.48-349.62)	368.30	(349.08-387.52)	1.59%		0.015
Transfer to short-term hospitals	3.23	(2.96-3.49)	4.15	(3.82-4.47)	1.51		(1.40-1.63)	1.63	(1.51-1.75)	6.28		(5.82-6.74)	7.31	(6.77-7.86)	2.56%		0.048
Admitted to same hospital	19.73	(18.77-20.70)	18.96	(17.75-20.16)	15.30		(14.87-15.74)	13.09	(12.93-13.23)	63.79		(60.67-66.91)	58.66	(54.94-62.38)	-1.39%		0.067
Left against medical advice	1.64	(1.50-1.79)	2.02	(1.81-2.33)	1.28		(1.17-1.39)	1.40	(1.25-1.54)	5.33		(4.86-5.81)	6.26	(5.62-6.91)	2.72%		0.363
Other (4)	0.72	(0.30-1.14)	0.71	(0.56-0.86)	1.55		(1.25-2.71)	1.72	(1.59-1.83)	6.47		(4.72-8.22)	7.71	(6.79-8.64)	2.97%		0.109
Visit Mortality (3)
Died during visit	0.73	(0.69-0.77)	0.70	(0.65-0.74)	0.56		(0.55-0.59)	0.48	(0.47-0.49)	2.35		(2.23-2.48)	2.15	(2.01-2.29)	-1.47%		0.466

CAGR: compound annual growth rate.

1: Visits rates reflect estimated ED visits per 1000 persons per group; data from US Census Bureau.

2: Visits rates reflected estimated ED visits per 1000 enrollees; data from CMS for Medicare and Medicaid; data from US Census Bureau American Community Survey for Private and Other. 3: Visits rates reflect estimated ED visits per 1000 persons; data from US Census Bureau.

4: Other includes uninsured, other, and missing.

Fig. 1. Cumulative Percentage change in Emergency Department Visits per 1000 Population, Average Charge per Visit, and United States Population Growth, 2010-2016.

population demographics are likely the predominant driver of increased ED visits.

Stratification of ED visit population rates by insurance status re- vealed significant positive trends for Medicaid, Private insurance, and Other (including self-pay). These results suggest broad-based factors underlying increased ED visits, and notably contrast with prior trends in 1997-2007 (when only Medicaid exhibited a significant positive trend) [30]. Lin and co-authors, also using NEDS data but examining a subset of visits with dispositions that could have resulted in admission, examined trends from 2006 to 2014 and found an unadjusted decline in proportion of visits for Private insurance and Self-pay [31]. The present analysis agrees with these unadjusted findings but also finds that when re-weighted by population totals, statistically significant growth in ED

Fig. 2. Emergency Department Charge per Visit, 2010-2016.

visits per covered population is occurring for both Private insurance and Other (self-pay). The finding of a growth in ED visits per covered population for Medicaid and Private insurance contrasts with prior anal- ysis of NEDS data covering the period of 2010-2014 (when Private in- surance exhibited a significant negative trend while Medicaid did not exhibit a significant trend), suggesting that the more recent years in- cluded in this data set have resumed the decades-long pattern of growth in per-population ED visits in these populations [24].

Possible reasons for the increasing rate of ED visits include increased prevalence of chronic illness, lack of accessible alternative sites of care appropriate for the illness (especially Mental illness), and the evolving role for Emergency Departments (e.g., acting as a “hospital’s front door” with a decrease in direct admissions). While our data do not ex- amine the volume for emerging sites of care such as retail clinics, urgent care centers, and telemedicine, the positive trend in the rate of ED visits

Table 2

Hospital visits and charges, 2010-2016.

2010		2016	Absolute Change CAGR P Value for Trend
Overall
ED Visits per 1000 Population	416.92	448.25	31.33	1.21%	b 0.001
Average (SD)	$2061 ($2962)	$3516 ($5284)	$1455	9.31%	b 0.001
Median	$1180	$1959	$779	8.82%
IQR	$619-$2297	$1016-$3877
Top 5 Reasons for ED Visit, 2016
Abdominal pain
ED Visits per 1000 Population	47.52	53.24	5.72	1.91	b 0.001
Average (SD)	$3086 ($3744)	$5111 ($6394)	$2025	8.77%	b 0.001
Median	$1844	$3119	$1275	9.15%
IQR	$951-$3772	$1573-$6237
Upper respiratory infections and pharyngitis ED Visits per 1000 Population	17.88	20.39	2.51	2.21%	b 0.001
Average (SD)	$896 ($1058)	$1481 ($1797)	$739	8.75	b 0.001
Median	$623	$1034	$411	8.81%
IQR	$372-$1011	$581-$1695
Chest pain, myocardial infarction, and heart disease
ED Visits per 1000 Population	17.40	19.65	2.25	2.05%	b 0.001
Average (SD)	$4665 ($6122)	$7096 ($9679)	$2431	7.24%	b 0.001
Median	$2542	$3928	$1386	7.52%
IQR	$1455-$5143	$2298-$7713
Mental health and substance abuse
ED Visits per 1000 Population	16.33	19.56	3.23	3.05%	b 0.001
Average (SD)	$1917 ($2015)	$3166 ($3578)	$1249	8.71%	b 0.001
Median	$1380	$2251	$871	8.50%
IQR	$774-$2343	$1260-$3849
Superficial injuries, contusion ED Visits per 1000 Population	19.11	17.18	-1.93	-1.76%	b 0.001
Average (SD)	$1598 ($2246)	$2715 ($4205)	$1117	9.24%	b 0.001
Median	$922	$1451	$529	7.85%
IQR	$534-$1678	$801-$2777

CAGR: compound annual growth rate.

1580 B.H. Lane et al. / American Journal of Emergency Medicine 38 (2020) 1576–1581

Table 3

hospital characteristics of emergency department visits, 2010-2016.

Weighted Estimate of Total Annual Visits, % (95% CI) Absolute Change, % CAGR P Value for Trend

	2010		2016
Urban influence status (1)
Metropolitan	81.57	(76.77-86.66)	84.58	(79.35-89.99)	3.69	0.61	b0.001
Micropolitan	10.38	(9.45-11.39)	8.28	(7.33-9.34)	-20.23	-3.70	b0.001
Rural	8.05	(7.04-9.24)	7.14	(6.29-8.10)	-11.30	-1.98	b0.001
Teaching status (2)
Yes	39.54	(37.35-41.77)	55.85	(53.40-58.28)	41.25	5.92	b0.001
No	60.46	(57.26-63.76)	44.15	(40.87-47.60)	-26.98	-5.10	b0.001
Trauma level designation (3)
Non-Trauma Center	63.76	(61.65-65.83)	55.05	(52.22-57.97)	-13.66	-2.42	b0.001
Trauma Center	36.23	(31.11-42.14)	44.93	(40.34-50.39)	24.01	3.65	b0.001

CAGR: compound annual growth rate.

1: Metropolitan contains a core urban county with N50,000 people and surrounding counties relying on the core county. Micropolitan counties a core urban county with N10,000 people and b50,000 and surrounding counties relying on the core county.

2: Teaching hospital status was obtained from the American Hospital Association Annual Survey of Hospitals.

3: Hospitals were identified in the data as a trauma center if they met level I,II, or III by the American College of Surgeon’s Committee on Trauma.

suggests that the effect of alternative sites in diverting low-acuity pa- tient encounters seen in other data sources is being outweighed by other factors [10].

We also found that the mean charge per ED visit, which is a function of both the intensity of care provided and pricing, significantly in- creased. While our outcome variable of charges cannot distinguish the relative contribution of care intensity and pricing, we favor the relative importance of pricing based on several suggestive findings. First, the common categories of Superficial Injuries as well as Upper Respiratory Infections and Pharyngitis, which are unlikely to have incorporated new high-intensity diagnostic or therapeutic modalities over the exam- ined six year-period, had greater or comparable CAGRs to high-acuity categories of Abdominal Pain and Chest Pain, Myocardial Infarction, and Heart Disease. Second, the positive CAGRs for increased mean ED visit charges were broadly consistent across categories with 140 of 144 categories with a CAGR of at least 5%. Third, we found that patients discharged from the ED significantly increased over the study period. This result makes less likely any argument that patients presenting to the ED were meaningfully “sicker” at the end of the study period

Table 4

Hospital charges of top 5 reasons for visit compared to overall, 2016.

2016 P Value

Overall

ED Visits per 1000 Population 448.25

Average (SD) $3516 ($5284) Reference

Top 5 Reasons for ED Visit, 2016 Abdominal pain

ED Visits per 1000 Population 53.24

Percent of overall visits 11.88

Average (SD) $5111 ($6394) b 0.001

Upper respiratory infections and pharyngitis

ED Visits per 1000 Population 20.39

Percent of overall visits 4.54

Average (SD) $1481 ($1797) b 0.001

Chest pain, myocardial infarction, and heart disease

ED Visits per 1000 Population 19.65

Percent of overall visits 4.38

Average (SD) $7096 ($9679) b 0.001

Mental health and substance abuse
ED Visits per 1000 Population	19.56
Percent of overall visits	4.36
Average (SD) Superficial injuries	$3166 ($3578)	b 0.001
ED Visits per 1000 Population	17.18
Percent of overall visits	3.83
Average (SD)	$2715 ($4205)	b 0.001

(2016) compared to the beginning (2010) and therefore required greater intensity of care.

For a subset of conditions, however, ED care intensity may have in- creased, in part due to an evolving role for the ED; for example, many EDs continue to expand the use of ED Observation Units for chest pain. Further evaluation with alternative data sources is needed to distin- guish pricing and care intensity, as well as the warranted and unwar- ranted variation in intensity across Emergency Departments.

Using a clinically relevant classification system with 144 categories, we found the 5 most frequent clinical categories together accounted for approximately 30% of ED visits, and that the third most common di- agnostic category (chest pain, myocardial infarction, and heart disease) was the most expensive on a per-visit basis. These findings suggest in- terventions to target ED care delivery for a small number of common di- agnoses could have meaningful impact on overall ED resource utilization and spending.

Comparing our results with prior literature, the observed trend of in- creased rates of ED is consistent with persistence of previously reported trends [30]. Using a 19-year period of Medical Expenditure Panel Survey (MEPS) data from 1996 to 2015, Yun and colleagues reported annual rates of increase in charges somewhat lower than what these NEDS data demonstrate [32]. However, direct comparisons with our data and study period are challenging because Yun and colleagues excluded all ED visits resulting in hospital admissions and MEPS adjusted its sur- vey methodology to address systematic under-reporting of ED visits by respondents in the latter half of 2013. Lastly, the estimate of ED visits per 1000 population (458 +- 28) in the 2016 National Hospital Ambula- tory Medical Care Survey (NHAMCS) is consistent with the NEDS data estimate for 2016 described here (448.19) [19]. Notably, the NHAMCS estimate was widely publicized as indicating ED visits had reached an “all-time high” [33].

Although nationally representative, our results do have limitations, in particular those common with de-identified Claims databases. First, we report charges as a surrogate for health care spending. To mitigate the limitations of charge-based data, we focused our evaluation on trend analysis rather than absolute values, with the assumption that biases in charge data will have remained relatively constant across the study period. MEPS data also suggest trend analysis is similar between expenditures and charges [32]. We investigated the possibility of using Current Procedural Terminology (CPT) coding available in the NEDS dataset as a proxy for clinical intensity, but considered the rate of missing data as too extensive for inclusion. Second, the use of admin- istrative claims data may be prone to coding error and ultimately classi- fication error. The adjudication process with the payer should minimize any error and there is no reason to suspect any errors were non-random in nature. Third, patients cannot be followed over time which precludes

any examination of the effect of multiple visits in temporal trends and charges. Fourth, NEDS does not report the hospital cost or reimburse- ment, so we cannot specifically report on changes in health care spend- ing, hospital costs, or hospital margin. Finally, the NEDS dataset only included visits at non-federal community hospitals, and may be undercounting the actual number of ED visits, though to a lesser degree than other sources like MEPS [1].

Conclusion

These data show the rate for ED visits is growing at a CAGR of 1.21% and mean charges are growing even faster at a CAGR of 9.31%, both of which are much faster than the population growth CAGR of 0.73%. The top five reasons for visiting an ED in 2016 were abdominal pain, URI (in- cluding pharyngitis), chest pain (including myocardial Infarction and heart disease), mental health (including substance abuse), and superfi- cial Injuries. These visits may represent focus areas for increasing the value of ED care.

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2019.158423.

References

1. Lee MH, Schuur JD, Zink BJ. Owning the cost of emergency medicine: beyond 2%. Ann Emerg Med. 2013;62:498-505.e3. https://doi.org/10.1016/j.annemergmed. 2013.03.029.
2. Dieleman JL, Squires E, Bui AL, Campbell M, Chapin A, Hamavid H, et al. Factors as- sociated with increases in US health care spending, 1996-2013. JAMA. 2017;318: 1668. https://doi.org/10.1001/jama.2017.15927.
3. Medford-Davis LN, Eswaran V, Shah RM, Dark C. The patient protection and afford- able care act’s effect on emergency medicine: a synthesis of the data. Ann Emerg Med. 2015;66:496-506. https://doi.org/10.1016/j.annemergmed.2015.04.007.
4. Galarraga JE, Pines JM. Anticipated changes in reimbursements for US outpatient emergency department encounters after health reform. Ann Emerg Med. 2014;63: 412-417.e2. https://doi.org/10.1016/j.annemergmed.2013.09.005.
5. Downey L, Burke T. Who pays? How reimburesment impacts the emergency depart- ment. J Health Hum Serv Adm. 2014;36:400-16.
6. Schuur JD, Venkatesh AK. The growing role of emergency departments in hospital admissions. N Engl J Med. 2012;367:391-3. https://doi.org/10.1056/NEJMp1204431.
7. Pines JM, Zocchi MS, Black BS. A comparison of care delivered in hospital-based and freestanding emergency departments. Acad Emerg Med. 2018;25:538-50. https:// doi.org/10.1111/acem.13381.
8. Weinick RM, Burns RM, Mehrotra A. Many emergency department visits could be managed at urgent care centers and retail clinics. Health Aff (Millwood). 2010;29: 1630-6. https://doi.org/10.1377/hlthaff.2009.0748.
9. Ho V, Metcalfe L, Dark C, Vu L, Weber E, Shelton G, et al. Comparing utilization and Costs of care in freestanding emergency departments, hospital emergency depart- ments, and urgent care centers. Ann Emerg Med. 2017;70:846-857.e3. https://doi. org/10.1016/j.annemergmed.2016.12.006.
10. Poon SJ, Schuur JD, Mehrotra A. Trends in visits to acute care venues for treatment of low-acuity conditions in the United States from 2008 to 2015. JAMA Intern Med. 2018;178:1342. https://doi.org/10.1001/jamainternmed.2018.3205.
11. Morganti KG, Bauhoff S, Blanchard JC, Abir M, Iyer N, Smith AC, et al. The evolving role of emergency departments in the United States. Santa Monica, California: RAND Corporation; 2013.
12. Lin MP, Nguyen T, Probst MA, Richardson LD, Schuur JD. emergency physician knowledge, attitudes, and behavior regarding ACEP’s choosing wisely

    recommendations: a survey study. Acad Emerg Med. 2017;24:668-75. https://doi. org/10.1111/acem.13167.

    Schuur JD, Carney DP, Lyn ET, Raja AS, Michael JA, Ross NG, et al. A top-five list for emergency medicine: a pilot project to improve the value of emergency care. JAMA Intern Med. 2014;174:509. https://doi.org/10.1001/jamainternmed.2013. 12688.

13. Kharbanda AB, Hall M, Shah SS, Freedman SB, Mistry RD, Macias CG, et al. Variation in resource utilization across a national sample of pediatric emergency departments. J Pediatr. 2013;163:230-6. https://doi.org/10.1016/j.jpeds.2012.12.013.
14. Hsia RY, MacIsaac D, Palm E, Baker LC. Trends in charges and payments for nonhos- pitalized emergency department pediatric visits, 1996-2003. Acad Emerg Med. 2008;15:347-54. https://doi.org/10.1111/j.1553-2712.2008.00075.x.
15. Caldwell N, Srebotnjak T, Wang T, Hsia R. “How much will I get charged for this?” patient charges for top ten diagnoses in the emergency department. PLoS ONE. 2013;8:e55491. https://doi.org/10.1371/journal.pone.0055491.
16. Williams RM. The costs of visits to emergency departments. N Engl J Med. 1996;334:

    5.

    Galarraga JE, Pines JM. Costs of ED episodes of care in the United States. Am J Emerg Med. 2016;34:357-65. https://doi.org/10.1016/j.ajem.2015.06.001.

17. Rui P, Kang K, Ashman J. National Hospital Ambulatory Medical Care Survey: 2016 emergency department summary tables. . https://www.cdc.gov/nchs/data/ahcd/ nhamcs_emergency/2016_ed_web_tables.pdf; 2016. Accessed date: 23 April 2019.
18. Healthcare Cost and Utilization Project. HCUP-US Overview. . https://www.hcup-us. ahrq.gov/overview; 2018. Accessed date: 1 April 2019.
19. Healthcare Cost and Utilization Project. NEDS database documentation. . https:// www.hcup-us.ahrq.gov/db/nation/neds/nedsdbdocumentation.jsp; 2018. Accessed

    date: 1 April 2019.

    Healthcare Cost and Utilization Project. HCUP-US Tools & Software Page. . https:// www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs.jsp; 2017. Accessed date: 1 April

    2019.

    Healthcare Cost and Utilization Project. Beta clinical classifications software (CCS) for ICD-10-CM/PCS. . https://www.hcup-us.ahrq.gov/toolssoftware/ccs10/ccs10.jsp; 2018. Accessed date: 1 April 2019.

20. Hooker EA, Mallow PJ, Oglesby MM. Characteristics and trends of emergency depart- ment visits in the United States (2010-2014). J Emerg Med. 2019. https://doi.org/10. 1016/j.jemermed.2018.12.025.
21. Houchens R, Ross D, Elixhauser A. Using the HCUP National Inpatient Sample to es- timate trends: HCUP methods series report #2006-05 ONLINE. . https://www.hcup- us.ahrq.gov/reports/methods/2006_05_NISTrendsReport_1988-2004.pdf; 2015.

    Accessed date: 20 May 2019.

    United States Census Bureau. National Population Totals: 2010-2018. . https://www.

    census.gov/data/tables/time-series/demo/popest/2010s-national-total.html; 2019.

    Accessed date: 1 April 2019.

    United States Census Bureau. American community survey tables for health insur- ance coverage. . https://www.census.gov/data/tables/time-series/demo/health- insurance/acs-hi.html; 2018. Accessed date: 1 April 2019.

22. Morris JK, Tan J, Fryers P, Bestwick J. Evaluation of stability of directly standardized rates for sparse data using simulation methods. Popul Health Metr. 2018;16. https://doi.org/10.1186/s12963-018-0177-1.
23. Dobson AJ, Kuulasmaa K, Eberle E, Scherer J. Confidence intervals for weighted sums of Poisson parameters. Stat Med. 1991;10:457-62.
24. Tang N, Stein J, Hsia RY, Maselli JH, Gonzales R. Trends and characteristics of US emergency department visits, 1997-2007. JAMA. 2010;304:664. https://doi.org/10. 1001/jama.2010.1112.
25. Lin MP, Baker O, Richardson LD, Schuur JD. Trends in emergency department visits and admission rates among US acute care hospitals. JAMA Intern Med. 2018;178 (12):1708-10. https://doi.org/10.1001/jamainternmed.2018.4725.
26. Yun J, Oehlman K, Johansen M. Per visit emergency department expenditures by in- surance type, 1996-2015. Health Aff (Millwood). 2018;37:1109-14. https://doi.org/ 10.1377/hlthaff.2018.0083.
27. American College of Emergency Physicians. ACEP: emergency visits reach all-time high. . http://newsroom.acep.org/2019-04-05-ACEP-Emergency-Visits-Reach-All- Time-High; 2019. Accessed date: 23 April 2019.