Trends in Emergency Department Utilization Among Women With Leiomyomas in the United States

Abstract

Objective:

To describe trends in emergency department (ED) visits in the United States with a primary diagnosis of leiomyomas, subsequent admissions, and associated charges.

Methods:

The Healthcare Cost and Utilization Project Nationwide Emergency Department Sample database was used to retrospectively identify all ED visits from 2006 to 2017 among women aged 18–55 years with a primary diagnosis of leiomyomas as indicated by International Classification of Diseases (ICD) diagnosis codes. Trends in ED visits and subsequent admissions were analyzed and stratified by patient and hospital characteristics. Secondary ICD codes, Current Procedural Terminology codes, and hospital charges were analyzed. A multivariate regression model was used to identify predictors of admission.

Results:

While the number of ED visits for leiomyomas increased from 28,732 in 2006 to 65,685 in 2017, the admission rate decreased, from 23.9% in 2006 to 11.1% in 2017. ED visits for leiomyomas were highest among women who were ages 36–45 (44.5%), in the lowest income quartile (36.1%), privately insured (38.3%), and living in the South (46.2%). Admission was more likely at nonteaching hospitals [OR 1.23 (1.08, 1.39)] or those located in the Northeast [OR 1.39 (1.15, 1.68)]. Patient characteristics associated with admission included older age [26–35: OR 1.42 (1.21, 1.66); 36–45: OR 2.01 (1.72, 2.34); 46–55: OR 2.60 (2.23, 3.03)] and bleeding-related complaints [OR 14.92 (14.00, 15.90)]. Admission was least likely in uninsured patients [Medicare: OR 1.37 (1.21, 1.54); Medicaid: OR 1.26 (1.16, 1.36); Private: OR 1.44 (1.32, 1.56)].

Conclusion:

While ED visits for leiomyomas are increasing, admission rates for these visits are decreasing. The substantial decline in admissions suggests many of these visits could potentially be addressed in a non-acute care setting. However, when women with leiomyomas present with a bleeding-related complaint, the odds of admission increase fifteenfold. There is an apparent disparity in likelihood of admission based on insurance status.

Precis

Leiomyoma-related emergency department visits increased significantly from 2006 to 2017, while subsequent hospitalizations decreased, with significant variation and disparities noted based on hospital and patient characteristics.

Introduction

Uterine leiomyomas are the most common benign gynecologic condition in the United States, with a prevalence of up to 70% by age 50 years.¹ Although the majority of leiomyomas are asymptomatic, approximately 25–50% of patients will experience symptoms, most commonly heavy menstrual bleeding and pelvic pain or pressure.² Because of their associated morbidity, leiomyomas continue to be the leading cause of hysterectomy in the United States³ and a cause of protracted symptoms for which there are limited long-term treatment options.

Although leiomyomas are often a chronic condition, many women with symptomatic leiomyomas will seek evaluation in the emergency department (ED). It has been estimated that one in five will visit the ED in the first year following their diagnosis of leiomyomas.⁴ Furthermore, women with leiomyomas have more ED visits than women without leiomyomas.⁴ Receiving care in the ED is costly, and has been found to be ten times more expensive than receiving care at an urgent care center for the same diagnosis.⁵ The total annual costs of leiomyomas has been estimated to be as high as $5.9–34.4 billion United States dollars.⁶ Both ED visits and hospital admissions contribute substantially to the overall economic burden that leiomyoma disorders pose to both individual patients and society.

Gaining insight into the volume of these visits and drivers of admission could help redirect appropriate patients to alternative care settings, resulting in significant cost savings. However, little is known currently about leiomyoma-related care in the ED or about the women who seek care in this setting. The objective of this study was to describe national trends in leiomyoma-related ED visits and associated hospital charges, as well as to determine factors associated with subsequent hospital admission.

Methods

Information on ED visits for leiomyomas in the United States from 2006 to 2017 was obtained from the Nationwide Emergency Department Sample (NEDS), Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality. The NEDS is the largest all-payer publicly available database of ED visits in the United States.⁷ This database contains information from 33.5 million ED visits at 984 hospitals to create a 20% stratified sample of U.S. hospital based EDs. Weighted, the database describes more than 100 million ED visits annually. Hospital and patient characteristics, as well as information on the nature of the visits, are included in the database. The University of Michigan Institutional Review Board reviewed this study and determined that it was exempt as the data are both deidentified and publicly available.

ED visits were selected for inclusion if they involved female patients ages 18–55 years old with a diagnosis of leiomyomas. International Classification of Diseases (both ninth (ICD-9)⁸ and tenth (ICD-10)⁹ revisions) diagnosis codes were used to identify visits in which the primary diagnosis was leiomyomas (ICD-9 codes: 218.X, 218.0, 218.1, 218.2, and 218.9; ICD-10 codes: D25.X, D25.0, D25.1, D25.2, and D25.9).

The weighted percentage of ED visits amongst this cohort was calculated for each year during 2006 to 2017, as well as overall during the study period. ED visits for leiomyomas were stratified by a number of covariates, including age, income, payment type, and hospital location and type. Age was grouped into four categories: 18–25, 26–35, 36–45, and 46–55 years. Income quartile by zip code was included in the NEDS database. No patient-specific income data was provided. Payment type included five categories: Medicaid, Medicare, private insurance, uninsured, and other. The uninsured category included an insurance status of “self-pay” and “no charge.”¹⁰ “No charge” indicated that the hospital did not charge any fee for the encounter and included charity, non-payment, professional courtesy, etc. The geographic location of the hospital was categorized as Midwest, Northeast, South, or West. Hospital type was divided into teaching and non-teaching. Teaching hospitals were all metropolitan based. Non-teaching hospitals included both metropolitan non-teaching and non-metropolitan hospitals. Non-metropolitan teaching hospitals were not classified separately by NEDS, as these hospitals were rare.

Secondary diagnoses associated with leiomyoma-related ED visits were also identified using ICD-9 and ICD-10 codes (Appendix 1, available online at http://links.lww.com/xxx). Diagnosis codes that represented similar clinical entities were grouped together. For instance, code N939 (abnormal uterine and vaginal bleeding, unspecified) and code N938 (other specified abnormal uterine and vaginal bleeding), were grouped together as “bleeding-related” diagnoses. Common Procedural Technology (CPT) codes associated with leiomyoma-related ED visits were also identified in order to ascertain tests and procedures ordered during these visits (Appendix 2, available online at http://links.lww.com/xxx). Again, clinically similar CPT codes were grouped together, such as CPT code 76856 (non-obstetrical pelvic ultrasound, real time with image documentation; complete) and CPT code 76830 (non-obstetrical transvaginal ultrasound) both being categorized as “pelvic ultrasound.” The frequencies of each group of secondary diagnoses and each group of CPT codes were calculated.

ED charges were obtained from the NEDS database and adjusted for inflation using the Consumer Price Index¹¹ related to the 2017 United States dollar. If ED charges were excessively low or high, the value was set to missing by NEDS. Missing charge values were treated as missing at random and imputed for the calculation of total charges. Age, region, income, and the presence of a leiomyoma diagnosis were included as covariates for the imputation analysis. Average annual percentage changes of charges were estimated by fitting trend data to a log-linear model using Joinpoint software version 4.7.0.0 (National Cancer Institute, Bethesda, Maryland).

Software SAS 9.4 (Cary, NC) was used to perform the statistical analyses. Descriptive statistics were calculated as counts and percentages for categorical variables and means for continuous variables. T-test, Chi Square test, and F-test were used to carry out significant difference tests and p values <0.05 were considered statistically significant.

A weighted multivariable logistic regression model was constructed to determine factors associated with hospital admission for leiomyoma-related visits. The patient and hospital characteristics were used as covariates in the model. The dependent variable was an indicator of ED visit admission status. The independent variables included in logistic model were: age group, region, payment type, income quartile by zip-code, hospital teaching status, concomitant diagnosis-bleeding and concomitant diagnosis-pain. The logistic model estimation incorporated different weights based on national hospital stratum. We obtained maximum likelihood estimates of regression parameters and standard errors as well as the corresponding odds ratios and confidence intervals for each independent predictor.

Results

During the study period, there were a total of 533,963 ED visits for leiomyomas among women ages 18 to 55 years. There were 487,688,338 total ED visits by women in this age group for all diagnoses during the study period. The number of ED visits for leiomyomas within this population increased 129% from 2006 (28,732 visits) to 2017 (65,685 visits) (Figure 1). The proportion of all ED visits that were for a primary diagnosis of leiomyomas also increased over the study period, from 0.079% in 2006 to 0.154% in 2017 (Table 1), suggesting that the increase was not simply from an increase in overall population volume. The overall admission rate for women ages 18 to 55 years presenting to the ED for leiomyomas during the 12-year study period was 16.3%, and decreased by over half, from 23.9% in 2006 to 11.1% in 2017. In comparison, the rate of admission for all other non-leiomyoma diagnoses presenting to the ED was relatively stable throughout the study period (9.2% in 2006 and 8.1% in 2017). The overall admission rate for patients presenting to the ED with all other diagnoses during the study period was only 8.5%.

Figure 1:

Yearly emergency department (ED) visits for leiomyomas among women aged 18–55 years by admitted status, 2006–2017.

Table 1.

Proportion of total ED visits for fibroids among women ages 18–55 years from 2006 to 2017.

Year	Total ED Visits (All Diagnoses)	ED Visits for Fibroids	Proportion of Total ED Visits for Fibroids
2006	36,241,653	28,732	0.079%
2007	37,165,635	31,933	0.086%
2008	38,522,166	33,850	0.088%
2009	39,542,196	37,728	0.095%
2010	40,788,108	40,304	0.099%
2011	40,477,498	42,918	0.106%
2012	41,607,380	44,985	0.108%
2013	41,574,457	45,461	0.109%
2014	42,761,683	53,033	0.124%
2015	42,755,919	51,843	0.121%
2016	43,477,273	57,491	0.132%
2017	42,774,371	65,685	0.154%

ED, emergency department

Table 2 depicts the descriptive characteristics of ED visits and admissions for leiomyomas during the entire study period. Almost half (44.5%) of women were ages 36–45 years, whereas only 3.4% were ages 18–25 years. Women presenting to the ED were most likely to be in the lowest income quartile (36.1%) and least likely to be in the highest income quartile (16.4%). The majority of ED visits for leiomyomas during the study period were paid by private insurance (38.3%), Medicaid (27.1%), or self-pay (24.4%). Regarding geographic location, women presenting to the ED with leiomyomas during the study period were most likely to be from the South (46.2%) and least likely to be from the Midwest (12.2%).

Table 2.

Characteristics of ED visits and admissions for fibroids among women ages 18–55 years from 2006 to 2017 based and associated logistic regression^*.

Variables	Total ED Visits (%)†	Admitted (%)†	Not admitted (%)†	P value	Admission OR	95% CI	Admission Adjusted OR	95% CI
Total	533963(100)	87025(16.3)	446938(83.7)
Age (years)
18–25‡	18182(3.4)	1252(1.4)	16930(3.8)	P<0.001	1.0
26–35	108986(20.4)	11214(12.9)	97772(21.9)		1.551	(1.35 – 1.79)	1.418	(1.21 – 1.66)
36–45	237801(44.5)	38604(44.4)	199197(44.6)		2.62	(2.28 – 3.01)	2.005	(1.72 – 2.34)
46–55	168994(31.6)	35956(41.3)	133038(29.8)		3.654	(3.18 – 4.2)	2.602	(2.23 – 3.03)
Region
Northeast	120721(22.6)	21516(24.7)	99205(22.2)	P<0.001	1.299	(1.12 – 1.51)	1.389	(1.15 – 1.68)
Midwest	65290(12.2)	10487(12.1)	54802(12.3)		1.146	(0.95 – 1.39)	1.126	(0.96 – 1.32)
South‡	246436(46.2)	35254(40.5)	211182(47.3)		1.0
West	101517(19)	19768(22.7)	81749(18.3)		1.449	(1.3 – 1.62)	1.114	(0.99 – 1.26)
Insurance
Medicare	21089(3.9)	3575(4.1)	17514(3.9)	P<0.001	1.323	(1.18 – 1.48)	1.366	(1.21 – 1.54)
Medicaid	144884(27.1)	24589(28.3)	120295(26.9)		1.324	(1.2 – 1.46)	1.258	(1.16 – 1.36)
Private	204299(38.3)	35886(41.2)	168414(37.7)		1.38	(1.27 – 1.5)	1.435	(1.32 – 1.56)
Uninsured‡	143111(26.8)	19134(22.0)	123978(27.8)		1.0
Other	19603(3.7)	3741(4.3)	15863(3.5)		1.528	(1.29 – 1.81)	1.535	(1.31 – 1.79)
Income Quartile
Lowest	192734(36.1)	29415(33.8)	163319(36.5)	P<0.001	0.931	(0.87 – 1)	1.019	(0.94 – 1.1)
Second	129503(24.3)	20461(23.5)	109041(24.4)		0.97	(0.91 – 1.04)	1.019	(0.95 – 1.1)
Third‡	112878(21.1)	18303(21)	94574(21.2)		1.0
Highest	87679(16.4)	15231(17.5)	72448(16.2)		1.086	(1.01 – 1.17)	1.01	(0.93 – 1.09)
Teaching Status
Teaching‡	313247(58.7)	50221(57.7)	263025(58.9)	P<0.001	1.0
Nonteaching§	220716(41.3)	36804(42.3)	183913(41.1)		1.048	(0.94 – 1.17)	1.225	(1.08 – 1.39)
Metropolitan Status
Metro, 1mil+	375185(70.3)	63565(73)	311620(69.7)	p=0.003	1.316	(1.18 – 1.47)	1.287	(1.13 – 1.47)
Metro,50k-<1mil	120149(22.5)	18081(20.8)	102068(22.8)		1.143	(0.97 – 1.35)	1.19	(0.98 – 1.44)
Non Metro‡	36403(6.8)	4886(5.6)	31517(7.1)		1.0
Concomitant Diagnoses
Pain related -No‡	377952(70.8)	18346(21.1)	359606(80.5)	p<0.001	1.0
Pain related -Yes	156011(29.2)	68679(78.9)	87332(19.5)		0.22	(0.2 – 0.25)	0.294	(0.27 – 0.33)
Bleeding related - No‡	427162(80)	81749(93.9)	345413(77.3)	p<0.001	1.0
Bleeding related -Yes	106801(20)	5276(6.1)	101525(22.7)		15.415	(14.48 – 16.41)	14.919	(14 – 15.9)

^*Variables included in the final logistic regression model included age group, region, insurance, teaching status, metropolitan status, and concomitant diagnoses.

^†Count (n) estimates are unweighted; percentage estimates are weighted using Healthcare Cost and Utilization Project discharge weights which are representative of the reported total of ED visits in the United States. The percentages within each covariate are column percentages. Some counts do not sum to totals because of missing data for factor; percentages calculated from all non-missing data. Less than 0.5% of the data were missing.

^‡Reference group

^§Non-teaching hospitals included both metropolitan non-teaching and non-metropolitan hospitals.

ED, emergency department; OR, odds ratio; CI, confidence interval

The most common secondary diagnoses amongst patients with a primary diagnosis of leiomyomas were those related to blood loss—i.e. anemia or abnormal menstrual bleeding (40.9%). The next most common secondary diagnoses were those related to abdominal or pelvic pain (21.4%). Table 3 shows the most common procedures and tests ordered during ED visits for leiomyomas. The most commonly ordered tests were chemistry studies (61.9%), hematologic studies (58.0%) and pelvic or abdominal ultrasounds (57.3%). We found that 14.2% of patients with a primary diagnosis of leiomyomas received intravenous hydration. This was significantly higher than the 8.6% of age-matched women presenting to the ED with all other diagnoses during the study period (p <0.001). Furthermore, 1.8% of women with a primary diagnosis of leiomyomas received a blood transfusion. This was also significantly higher than the 0.1% of age-matched women with all other diagnoses who received a blood transfusion (p<0.001).

Table 3.

Most common CPT codes associated with ED visits for fibroids amongst women ages 18–55 from 2006 to 2017.

Procedure/Test	Frequency
Chemistry studies	61.9%
Hematologic studies	58.0%
Pelvic/abdominal ultrasound	57.3%
Urinalysis/Urine culture	47.9%
STI/Vaginitis screen	39.6%
Injections & Infusions*	38.5%
Pregnancy test	36.1%
Computerized tomography abdomen/pelvis	18.9%
Blood typing	16.7%
Intravenous hydration	14.2%

CPT, common procedural technology; ED, emergency department; STI, sexually transmitted infection.

^*Defined as “Therapeutic, Prophylactic, & Diagnostic Injections & Infusions” per the Centers for Medicare & Medicaid.³⁴

From 2006 to 2017, the median ED visit charges for leiomyomas more than doubled, from $2,586 to $6,193. Average ED charges for leiomyomas were consistently higher than charges for non-leiomyoma related visits (p<0.001) (Figure 2). Furthermore, the total ED visit charges for leiomyomas increased by 445% during the study period, totaling nearly $500 million in 2017. The annual average percentage change for leiomyoma-related ED visit charges was significantly higher than that of all other age matched diagnoses during the overall study period (8.7% versus 7.3%; P< 0.01).

Figure 2:

National estimates of average emergency department charges in 2017 U.S. dollars for women aged 18–55 years, 2006–2017. IQR, interquartile range.

Table 2 also shows the multivariate analysis of predictors of hospital admission. The probability of hospital admission for leiomyomas increased with age, with women aged 18–25 years being significantly less likely to be admitted than older women [26–35: OR 1.42 (1.21, 1.66); 36–45: OR 2.01 (1.72, 2.34); 46–55: OR 2.60 (2.23, 3.03)]. ED visits in the Northeast were significantly more likely to result in admission than those in other regions [OR 1.39 (1.15, 1.68)]. Admission was least likely to occur in the South. Admission was least likely in uninsured patients [Medicare: OR 1.37 (1.21, 1.54); Medicaid: OR 1.26 (1.16, 1.36); Private: OR 1.44 (1.32, 1.56)]. Unadjusted chi-square analyses demonstrated that women in the lowest income quartile were less likely to be admitted. Income quartile by zip code did not correlate significantly with admission in our logistic regression model. Patients seen at nonteaching hospitals were more likely to be admitted than those seen at teaching hospitals [OR 1.23 (1.08, 1.39)]. Finally, women with leiomyomas who presented to the ED with bleeding were significantly more likely to be admitted [OR 14.92 (14.00, 15.90)]. Patients who presented with pain were significantly less likely to be admitted [OR 0.29 (0.27, 0.33))].

Discussion

In this analysis of a large nationally representative database, ED visits with a primary diagnosis of leiomyomas progressively increased while admission rates decreased over the 12-year study period. Approximately one in ten ED visits for leiomyomas resulted in admission in 2017. The growing disparity between the number of ED visits for leiomyomas and the number of those visits resulting in admission warrants exploration. It is possible that women are increasingly using the ED for non-urgent leiomyoma-related issues. Alternatively, this finding could potentially reflect changes in ED care patterns, such as managing pain control and administering blood transfusions in the ED as opposed to admitting. Increasing use of effective medical therapies in the ED, such as tranexamic acid, to manage heavy bleeding, could also contribute to the decreasing admission rate. Since admission rates varied significantly by region and payer, this pattern may also result from local policy or health care professional biases. The increasing number of leiomyoma-related ED visits is particularly interesting given the recent finding of a decreasing trend of new leiomyoma diagnoses among women aged 18–65 years between 2005–2015. ¹² Future research exploring patients’ motivations for seeking care in the ED may provide clarification.

The overall economic burden of symptomatic leiomyomas on society has been estimated to be $5.9–34.4 billion United States dollars annually in both direct costs and indirect costs.⁶ We found an average ED charge of over $6,000 per visit and $500 million in total charges in 2017. Furthermore, average ED charges for leiomyoma-related visits were consistently twice as high as ED charges for other diagnoses. The high proportion of patients who received costly ED-based imaging studies likely contributed significantly to these costs. It is likely that many of these patients were appropriate candidates for outpatient imaging, which potentially could have saved significant money and resources. Prior studies have similarly found that women with leiomyomas incur total health care costs over twice that of their counterparts without leiomyomas.¹³ This cohort of women should be targeted for intervention to improve access to outpatient care, thereby mitigating unnecessary, costly ED utilization.

Based on the results of our logistic regression, we further characterized patients who are more likely to be admitted after presenting to the ED for leiomyomas. Women who presented with a bleeding-related complaint were fifteen-times more likely to be admitted than those who did not present with a bleeding-related complaint. On the contrary, if the primary complaint was pain-related, the odds of admission were significantly lower. These associations are logical, as bleeding-related issues can quickly lead to hemodynamic instability, a scenario which almost exclusively warrants admission. Pain-related leiomyoma symptoms are poorly understood, are often multifactorial, and may not be directly related to leiomyomas when present. Pain is also less likely to be emergent in nature and can often be managed as an outpatient.

The probability of admission was significantly lower amongst women aged 18–25 years, and nearly nine out of every ten women admitted for leiomyomas were ages 36–55 years old. This finding is consistent with results from prior studies. Cox et al. analyzed reproductive health-related ED visits by women in Maryland from 1999 to 2005 and reported an increased odds of admission with increasing age.¹⁴ Several other studies have looked specifically at women with leiomyomas, and have all concluded that the hospitalization rate for leiomyomas increases with age, peaking amongst women ages 40 to 54 years.^15–17 Admission rates are lowest amongst younger women, possibly due to different perceptions of what constitutes a medical emergency, leading to higher care-seeking behavior. Indeed, prior research has demonstrated that young adults receive a greater proportion of their care in the ED compared with older adults¹⁸. Furthermore, younger adults are least likely to report the seriousness of their medical problem as the reason for seeking care in the ED compared with older age groups¹⁹. Additionally, older women may have more comorbidities contributing to their leiomyoma-related ED visit, increasing their rate of admission. Finally, it is also likely that younger women have less leiomyoma burden.

Women in the South had the lowest probability of hospital admission, though the number of leiomyoma-related ED visits in this region was highest. While the high volume of ED visits in the South may be in part because it is the most populous region in the country as defined by NEDS,²⁰ it is also known to be the region with the highest prevalence of leiomyomas.²¹ Regional differences in leiomyoma prevalence could largely be explained by differences in racial makeup. Black women, who are disproportionately affected by leiomyomas, comprise the greater proportion of the total population in the South.²² Prior NEDS studies have also demonstrated significant variation in hospital admission rates across United States regions for other medical conditions.^23–26 Potential factors contributing to this variation include regional differences in healthcare access, systemic racism, disease severity, physician density, and environmental factors.

Income was not found to be a significant predictor of hospital admission. There were, however, notable differences in ED use across income quartiles. ED visits for leiomyomas were highest amongst women in the lowest income bracket. Overall healthcare utilization has previously been shown to be influenced by socioeconomic status.²⁷ In a large national cross-sectional study, Nicholson et al. found a significant inverse relationship between ED usage for gynecologic conditions and median household income.²⁸ There are likely numerous reasons to explain this finding. For instance, many low-income women do not have a primary care physician and report limited access to primary care services.²⁹ Additionally, studies have found that patients of lower socioeconomic status prefer to receive medical care in the ED over ambulatory care centers due to affordability and trust in the technical quality of care.³⁰ As a result, many of these women use the ED for their regular source of medical care.²⁹

Given the observed association between number of ED visits and lower socioeconomic status, we had anticipated that the most frequently billed payor would be Medicaid. Indeed, Nicholson et al. found in their national cross-sectional study that ED visits for gynecologic conditions were three times more common amongst women with Medicaid compared with private insurance.²⁸ The results of our analysis unexpectedly showed that the most frequently billed payor for leiomyoma-related ED visits was private insurance. Potential explanations could include higher primary care physician referrals to the ED amongst the privately insured or the expansion of private coverage under the Affordable Care Act. Akosa Antwi et al. looked at the impact of the Affordable Care Act on patterns of ED use amongst adults aged 19–29 years. From 2007 to 2011, the fraction of privately insured patients increased, whereas the fraction of those insured through Medicaid decreased.³¹ Our results also showed that women who were uninsured were least likely to be admitted for their leiomyomas. Prior studies have demonstrated that admission rates for several other conditions are higher in privately insured women and lower in the uninsured.^32,33

As the NEDS database does not provide detailed clinical data or associated outcomes data, we were unable to elucidate whether this apparent disparity is being driven by unnecessary admissions of the privately insured or inappropriate discharges of the uninsured. The former would suggest current practices are contributing to unnecessary healthcare expenditures, while the latter means that a vulnerable cohort of patients are being placed at risk for suboptimal care. Other possible explanations to explain this disparity include differences in disease severity or patient choice regarding admission. Reasons for the observed variation in admission rates based on insurance coverage warrants further investigation.

Our study has some limitations. First, patients with several ED visits during the study period were potentially represented more than once, as analyses were visit-based, not patient-based. Second, there are limitations associated with the use of ICD codes, which are intended for billing purposes, not disease surveillance, and do not perfectly capture the clinical picture of the visit. It is also possible that the transition from ICD-9 to ICD-10 that occurred during the study period resulted in coding errors and disruptions in observed rates. We did not, however, notice any sharp or unexplained trend changes during this period. Finally, the NEDS database does not contain information on race or ethnicity or patient-specific income levels or other socioeconomic variables, which limited our ability to assess the impact of social determinants of health.

Despite these limitations, this work has many strengths. It is based on the largest, publicly available all-payer ED database in the United States, with information on 33.5 million ED visits derived from 984 hospitals. The large number of ED records in the NEDS database decreases the risk of sampling error. Furthermore, estimates from the NEDS database are nationally representative, increasing the external validity of our findings. Finally, our data were stratified by several key variables, providing a comprehensive understanding of ED visits for leiomyomas in the United States.

This nationally representative analysis provides a foundation for understanding ED visits and hospitalizations where uterine leiomyomas are the primary diagnosis. These encounters impart a significant and increasing economic burden on patients, our health care system, and society, and highlight an opportunity to change the trajectory of ED utilization for this often chronic and typically non-emergent condition. Improving our understanding of patients with leiomyomas who present to the ED but are not admitted will help target interventions to better care for these patients in the outpatient setting and ultimately reduce the burden of leiomyomas for both patients and society. Finally, there is an apparent disparity in the likelihood of admission based on insurance status. This finding warrants further study to ensure that all women presenting with leiomyomas receive equitable care.