Surgical procedures and their cost estimates among women with newly diagnosed endometriosis: a US database study

Abstract

Objective:

This descriptive study assessed the rate and costs of surgical procedures among newly diagnosed endometriosis patients.

Methods:

Utilizing the Medstat MarketScan database, commercially insured women aged 18–45 with endometriosis newly diagnosed during 2006–2007 were identified. Each endometriosis patient was matched to four women without endometriosis (population controls) based on age and region of residence. Surgical procedures received during the 12 months post-diagnosis were assessed. Costs of surgical procedures were the amount paid by the insurance companies.

Results:

This study identified 15,891 women with newly diagnosed endometriosis and 63,564 population controls. More than 65% of endometriosis patients received an endometriosis-related surgical procedure within 1 year of the initial diagnosis. The most common procedure was therapeutic laparoscopy (31.6%), followed by abdominal hysterectomy (22.1%) and vaginal hysterectomy (6.8%). Prevalence and type of surgery performed varied by patient age, including a hysterectomy rate of approximately 16% in patients younger than 35 and 37% among patients aged 35–45 years. Average costs ranged from $4,289 (standard deviation [SD]: $3,313) for diagnostic laparoscopy to $11,397 (SD: $8,749) for abdominal hysterectomy.

Limitations:

Diagnosis of endometriosis cannot be validated against medical records, and information on the severity of endometriosis-related symptoms is not available in administrative claims data.

Conclusions:

Over 65% of patients had endometriosis-related surgical procedures, including hysterectomy, within 1 year of being diagnosed with endometriosis. The cost of surgical procedures related to endometriosis places a significant financial burden on the healthcare system.

Key words::

• Administrative claims
• Cost
• Endometriosis
• Surgery

Introduction

Endometriosis is a chronic estrogen-dependent disease that occurs in approximately 10% of women at reproductive age1,2. The condition often causes pain symptoms, such as dysmenorrhea, chronic non-menstrual pain and dyspareunia, as well as infertility. The primary purposes of treatments for women with endometriosis include the management of pain and the treatment of infertility. Both medical therapies and various forms of surgical interventions are currently used for the management of endometriosis symptoms3,4. Medical therapy of endometriosis includes oral contraceptives, non-steroidal anti-inflammatory drugs, high-dose progestins, gonadotropin releasing hormone agonists (GnRH-a), and danazol. Endometriosis patients can also choose from various surgical treatments based on their age, preference of preserving reproductive potential, and severity of endometriosis. Laparoscopic removal of endometriotic lesions is currently the most common type of surgical intervention in women with endometriosis. Generally, laparoscopic surgical treatment may be divided into ablation using diathermy, laser, and ultrasound energy sources, or excision techniques. More radical surgical approaches include hysterectomy with or without oophorectomy. Regardless of the type of surgical treatment, pain recurrence is high among endometriosis patients5.

Due to the difficulty of diagnosing and treating the disease, endometriosis incurs significant healthcare utilization and costs. Assuming a prevalence rate of 10%, and these rates would be much higher if the entire disease population were diagnosed and treated6, endometriosis was estimated to cost about $22 billion in the US during 2002, $5.9 billion of which were indirect costs due to the loss of productivity; of the remaining $16.1 billion, 90% were attributed to hospital care. Hospitalizations were the most significant cost driver, totaling $12,644 per patient among those who were hospitalized in 20026, with endometriosis being the third leading cause of gynecological hospitalizations in the United States7. High treatment costs for endometriosis also arise as a result of the use of specialists, as patients often visit obstetricians and gynecologists. It is estimated that an individual woman suffering from endometriosis loses about 1 month per year in work-related activity, translating into a $1,595 loss to employers8. Women with endometriosis also have a higher cost for concomitant conditions such as infertility, depression, irritable bowel syndrome, and abdominal pain, among others9.

The goal of this study was to (1) document the demographic and clinical characteristics of patients with newly diagnosed endometriosis and the surgical treatment they received during or following diagnosis; and (2) assess the costs of surgical procedures related to the treatment of endometriosis in both inpatient and outpatient settings. To the authors’ knowledge, this is the first study to assess surgical rates among patients with newly diagnosed endometriosis compared to matched population controls, and to provide cost estimates for endometriosis-related surgical procedures using a large claims database. The most recent studies on costs of surgical procedures for the treatment of endometriosis were published in early 1990s. Given the advances in medical technology and changes in the healthcare market over the past 20 years, the information needs to be updated. Knowledge gained from this study will inform healthcare providers and payers of the current practice of treating endometrioses in the real world, and provide parameters for cost-effectiveness evaluation of various treatment options of endometriosis.

Methods

Data source

The authors analyzed administrative claims data from the Medstat MarketScan commercial claims databases (Thomson Reuters Healthcare Inc., Ann Arbor, MI, USA) collected between 2005 and 2008. The data represent a large, geographically distributed population in the US and provided enrollment records, pharmacy and medical service claims. This Health Insurance Portability and Accountability Act (HIPAA) compliant database contains records for approximately 30 million unique commercially-insured (working adults and their dependants) and 3 million Medicare-insured individuals. These records were linked via encrypted and unique patient-level identifiers allowing for comprehensive temporal tracking of individuals’ healthcare utilization.

Demographic characteristics such as age, gender, US geographic region of residence, health insurance plan type, and plan enrollment status were available in the enrollment files. The medical services claims recorded detailed information about inpatient and outpatient care, including date and place of service, provider type, payments, and up to 15 International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes. Pharmacy claims files included information on National Drug Code (NDC), dispense date, quantity, days supplied, and plan- and patient-paid amounts.

Patient selection

Two study cohorts were constructed: patients with newly diagnosed endometriosis and a matched population control cohort. To construct the endometriosis cohort, all women aged 18–45 years with at least one inpatient or outpatient medical claim for endometriosis (ICD-9-CM 617.xx)10 during 2006–2007 were selected. The date of the first endometriosis claim was denoted as the index date. To ensure that this cohort only consisted of newly diagnosed endometriosis patients, those with endometriosis claims during the 12 months prior to the index date were excluded. Eligible study patients were required to have continuous enrollment from 12 months prior through 12 months following the index date. To assess the clinical and economic burden of endometriosis, each endometriosis patient was matched to four women without endometriosis (population controls) based on age and region of residence. All selected population controls had at least 24 months of continuous health plan enrollment between 2005 and 2008, with the midpoint of enrollment set as the index date. Patients in either cohort were excluded if they had a diagnosis of uterine fibroid (ICD-9-CM 218.xx)10 or malignant neoplasm of female reproductive organs (ICD-9-CM: 179.xx–184.xx)10, or if they received GnRH-a during the 12-month period prior to the index date to ensure that included subjects are newly diagnosed and estimated surgery rates are for endometriosis and not for uterine fibroid.

Characteristics of patients with newly diagnosed endometriosis

Baseline demographic characteristics, including age, type of health insurance plan (e.g., comprehensive, health maintenance organization, and point of service), and geographic region of residence (e.g., northeast, north central, west, and south) were assessed for all patients based on information collected during the 12 months prior to the index date. The presence of endometriosis-related conditions at baseline was also assessed and included abdominal pain, infertility, migraine, depression, interstitial cystitis, irritable bowel syndrome, and chronic fatigue identified based on ICD-9-CM diagnosis codes (see Appendix) recorded in the medical service claims. Using the algorithm developed by Deyo et al.11, the Charlson Comorbidity Index (CCI) was calculated to assess the burden of chronic diseases among endometriosis patients and population controls. Also reported was the proportion of patients with selected gynecological surgical procedures and the number of surgical procedures that were performed on or during the 12 months following the index date. The surgical procedures assessed included abdominal hysterectomy, vaginal hysterectomy, diagnostic laparoscopy, therapeutic laparoscopy, laparotomy, and oophorectomy.

Patient demographics, clinical characteristics, and rates of surgical procedures were reported for each cohort. Mean and standard deviations (SD) were reported for age, CCI and number of surgical procedures. Student’s t-tests were employed to detect statistical differences of age between cohorts, and Wilcoxon rank-sum tests were employed for CCI and number of surgical procedures. Percentages were reported for categorical variables (including age categories, type of insurance plan, region of residence, presence of comorbidities, and receipt of surgical procedures) and chi-square tests were used to detect statistical differences between endometriosis and population control cohorts.

The analysis of rate and number of surgical procedures received following the initial endometriosis diagnosis were further stratified by patients’ age. Cochran–Armitage trend tests were employed to detect the trend of surgical rate across age subgroups, and Cochrane Mantel–Haenszel tests were employed to detect the statistical difference of categories of number of surgical procedures. Differences between cohorts or age subgroups were considered statistically significant if p-values were <0.05.

A sensitivity analysis of surgery rates was conducted among a subgroup of patients with at least one inpatient claim or two outpatient claims for endometriosis during the screening period. The more stringent inclusion criterion was applied to outpatient visits to rule out cases with diagnostic tests of endometriosis, as we can be reasonably certain that an individual with more than one outpatient visit with claims for endometriosis has the disease. These stringent criteria may reduce sample size, but were most likely to give true endometriosis patients for the sensitivity analysis. Surgery rates were reported for the post-index period and by patient age.

Characteristics of endometriosis-related surgical procedures

Characteristics of selected gynecological surgical procedures were assessed. The authors identified 6,642 inpatient stays and 5,908 outpatient visits with these surgical procedures, from which 1,948 inpatient stays or 3,108 outpatient visits were further selected with diagnosis codes of endometriosis in the associated claims. Some patients may have received multiple surgical procedures during one inpatient stay (n = 221) or a single outpatient visit (n = 426). Such cases were excluded from the analytical sample because of the difficulty to attribute costs to individual surgical procedures. Additionally, nine inpatient stays with diagnostic laparoscopy were excluded. While diagnostic laparoscopies are generally performed in outpatient settings, the nine inpatient stays with diagnostic laparoscopy were associated with high costs (mean: $17,743, SD: $11,877) and long length of stay (mean: 3.6 days, SD: 4.7 days). Further examination found that these inpatient stays had other diagnoses, such as missed abortion or appendicitis, which increased their costs. Of the remaining cases (n = 4,400), the characteristics of select surgical procedures were reported by setting (i.e., inpatient and outpatient). Costs of surgical procedures performed in inpatient settings consisted of the paid amount of the inpatient stay during which the surgical procedure was performed. Costs of surgical procedures performed in the outpatient settings consisted of paid amount of all services that the patient received during the date of the surgery. The place of service for surgical procedures performed in the outpatient settings was also reported. Costs were adjusted to 2009 US dollars using the medical care component of the consumer price index12. Analyses were conducted using SAS version 9.2 (SAS Institute, Inc., Cary, NC, USA).

Results

Baseline demographics and comorbidities

This study identified 15,891 endometriosis patients and 63,564 matched population controls. The distributions of patient age, region, and insurance plan were similar across cohorts (Table 1). Endometriosis patients had a higher prevalence of abdominal pain (22.0 vs. 1.9%), depression (6.5 vs. 1.1%), infertility (5.5 vs. 0.2%), and migraine (4.7 vs. 0.6%) than population controls during the pre-index period (all p < 0.01). Endometriosis patients also had a higher mean CCI compared to the population control cohort (0.14 vs. 0.03, p < 0.01).

Table 1.  Demographics and comorbidities at baseline, by study cohort.

	Study cohorts
	Endometriosis cohort	Population control cohort	p-value*
	(n = 15,891)	(n = 63,564)
Age (%)			0.99
18–24	5.2	5.2
25–29	9.8	9.8
30–34	19.4	19.4
35–39	28.0	28.0
40–45	37.7	37.7
Age, mean (SD)	36.4 (6.32)	36.4 (6.47)	0.37
Region (%)			0.99
Northeast	6.7	6.7
North Central	23.5	23.5
South	52.9	52.9
West	15.9	15.9
Unknown	1.0	1.0
Type of insurance plan enrolled, %			<0.01
Missing	2.5	4.7
Comprehensive	7.7	5.2
Exclusive provider organization	0.7	0.9
Health maintenance organization	19.8	19.4
Point of service	12.0	11.4
Preferred provider organization	53.2	54.0
Point of service with capitation	4.0	3.2
Consumer driven health plan	0.1	1.2
Endometriosis-related comorbidities, %
Abdominal pain	22.0	1.9	<0.01
Infertility	5.5	0.2	<0.01
Migraine	4.7	0.6	<0.01
Depression	6.5	1.1	<0.01
Interstitial cystitis	0.7	0.02	<0.01
Irritable bowel syndrome	1.9	0.2	<0.01
Chronic fatigue	0.3	0.1	<0.01
Charlson Comorbidity Index, mean (SD)	0.14 (0.53)	0.03 (0.25)	<0.01
Percent			<0.01
0	88.2	97.7
1	6.8	1.3
2	3.9	0.8
>2	1.2	0.2

*χ² test was used for the comparison of age categories, region of residence, type of insurance plan enrolled, and the presence of comorbidities, Student’s t-test was used for the comparison of age, and Wilcoxon rank-sum test was used for the comparison of Charlson Comorbidity Index.

SD, standard deviation.

Surgery rates

During the 12-month post-index period, a higher proportion of endometriosis patients had endometriosis-related gynecological surgical procedures compared to population controls (65.5 vs. 1.5%, p < 0.01) (Table 2). In all, 49% of endometriosis patients had surgical procedures on the index date. The most common procedure was therapeutic laparoscopy (31.6%), followed by abdominal hysterectomy (22.1%) and vaginal hysterectomy (6.8%). Approximately 20% of endometriosis patients had at least one surgical procedure in the 12 months following the index date. During this period, 10.2% received therapeutic laparoscopy, 5.8% abdominal hysterectomy, 2.4% diagnostic laparoscopy, and 1.3% vaginal hysterectomy. Among patients who had surgery, 3.0% had surgeries both on index date and during the 12-month post index period.

Table 2.  Proportion of patients with endometriosis-related surgical procedures in 1-year post-index period.

	Endometriosis cohort (n = 15891)			Population control cohort (n = 63564)
	All	On index date	After index date	All	On index date	After index date
Any of the selected surgery, %*	65.5	49.0	19.5	1.5	0.0	1.5
Any hysterectomy, %	28.8	22.0	7.0	0.6	0.0	0.6
Abdominal hysterectomy, %	22.1	16.4	5.8	0.4	0.0	0.4
Vaginal hysterectomy, %	6.8	5.7	1.3	0.2	0.0	0.2
Any diagnostic or therapeutic laparoscopy, %	35.6	25.1	11.9	0.7	0.0	0.7
Diagnostic laparoscopy, %	6.2	3.9	2.4	0.1	0.0	0.1
Therapeutic laparoscopy, %	31.6	22.5	10.2	0.6	0.0	0.6
Laparotomy, %	1.0	0.6	0.5	0.0	0.0	0.0
Oophorectomy, %	2.7	1.8	0.9	0.1	0.0	0.1

Note: Proportion of patients receiving surgeries on and after index date may add up to more than the proportion of patients receiving surgeries during the 1-year post-index period because patients may have received repeated surgeries during the year.

χ² tests were used to detect the difference between endometriosis and population control cohorts. All proportions were significantly different between cohorts.

Among endometriosis patients, the rate and type of surgical procedures received during the 12-month post-index period varied by age (Table 3). A lower proportion of endometriosis patients aged 18–24 years received select gynecological surgical procedures compared to those aged 25–34 and 35–45 years (49.3 vs. 59.6% and 69.4%, respectively, p-value for trend <0.01). Only 1% of endometriosis patients aged 18–24 years received abdominal or vaginal hysterectomy, and the proportion increased to 15.6% among patients aged 25–34 years and 36.9% among those aged 35–45 years (p-value for trend <0.01)). However, the proportion of patients receiving diagnostic or therapeutic laparoscopy decreased from 47.0% among patients aged 18–24 years to 31.3% among those aged 35–45 years (p-value for trend <0.01).

Table 3.  Surgery rates by age group during 1-year post-index period (include index date) among endometriosis patients.

	Age group (years)			p-value
	18–24	25–34	35–45
	(n = 820)	(n = 4,637)	(n = 10,434)
Any of the selected surgery, %	49.3	59.6	69.4	<0.01
Any hysterectomy, %	1.1	15.6	36.9	<0.01
Abdominal hysterectomy, %	1.0	11.5	28.5	<0.01
Vaginal hysterectomy, %	0.1	4.1	8.6	<0.01
Any diagnostic or therapeutic laparoscopy, %	47.0	43.2	31.3	<0.01
Diagnostic laparoscopy, %	13.3	8.7	4.5	<0.01
Therapeutic laparoscopy, %	37.9	37.9	28.3	<0.01
Laparotomy, %	0.5	0.9	1.2	0.03
Oophorectomy, %	1.2	2.1	3.0	<0.01
Number of surgeries during 1 year post
Mean (SD)	0.6 (0.7)	0.7 (0.7)	0.8 (0.6)
Percent
0	50.7	40.4	30.6	<0.01
1	41.7	49.2	60.3
2	6.8	9.3	8.0
3+	0.7	1.2	1.1

Cochran–Armitage trend tests were used to detect trend of surgery rates across age subgroups, and Cochrane Mantel–Haenszel tests were used to detect the significant difference of number of surgeries across the age sub-groups. A p-value <0.05 was considered statistically significant.

Sensitivity analysis on surgical rate

Patients in the subgroup population were slightly older than the endometriosis patients in the primary analysis (71.8% were 35–45 compared to 65.7% in primary analysis; mean age: 37.4 (SD: 5.9) compared to 36.4 (SD: = 6.3)). As noted in Table 4, 77% of patients received an endometriosis-related surgery on the index date or within 1 year of the endometriosis diagnosis (46.6% received hysterectomy and 30.3% received laparoscopy). The same age-related pattern was observed in the subgroup population, as older patients were more likely to receive a hysterectomy and younger patients were more likely to receive a laparoscopy. Surgical rates were generally higher among the subgroup population compared to the endometriosis cohort from the primary analysis. The surgery rates among patients aged 18–24 years was 6.4 percentage points higher in the subgroup population than in the primary endometriosis cohort (55.7 vs. 49.3%), primarily due to the difference in laparoscopy rates (50.3 vs. 47.0%). The difference in surgery rates increased among older patients. Compared to the endometriosis cohort in the primary analysis, the surgery rate was 8.8 percentage points higher for patients aged 25–34 years and 11.8 percentage points higher for those aged 35–45 years. Specifically, the rates of hysterectomy were 14–18 percentage points higher and the rate of laparoscopy was 5 points lower.

Table 4.  Sensitivity analysis: surgery rates by period and age group.

(a) Rate of surgery	Surgery rate in during and 1 year following diagnosis
	All (n = 8,830)	On index date		After index date
Any of the selected surgery, %	77.1	58.8		22.0
Any hysterectomy, %	46.6	36.7		10.1
Abdominal hysterectomy, %	37.1	28.3		9.0
Vaginal hysterectomy, %	9.7	8.5		1.2
Any diagnostic or therapeutic laparoscopy, %	30.3	20.1		11.5
Diagnostic laparoscopy, %	4.9	2.8		2.2
Therapeutic laparoscopy, %	26.9	18.0		10.0
Laparotomy, %	1.6	0.9		0.7
Oophorectomy, %	4.2	2.9		1.3
(b) Rate of surgery by age group (years)	Surgery rate by age at first endometriosis diagnosis					p-value
	18–24 (n = 298)		25–34 (n = 2,188)		35–45 (n = 6,344)
Any of the selected surgery, %	55.7		68.4		81.2	<0.01
Any hysterectomy, %	3.0		29.8		54.5	<0.01
Abdominal hysterectomy, %	2.7		22.9		43.6	<0.01
Vaginal hysterectomy, %	0.3		6.9		11.1	<0.01
Any diagnostic or therapeutic laparoscopy, %	50.3		38.5		26.5	<0.01
Diagnostic laparoscopy, %	14.4		7.8		3.4	<0.01
Therapeutic laparoscopy, %	42.0		33.7		23.9	<0.01
Laparotomy, %	1.0		1.4		1.7	0.16
Oophorectomy, %	2.4		4.0		4.3	0.17
Number of surgeries during 1 year post, mean (SD)	0.7 (0.7)		0.8 (0.7)		0.9 (0.6)
0	44.3		31.6		18.8	<0.01
1	43.3		56.8		72.6
2	10.7		9.8		7.4
3+	1.7		1.7		1.2

Cochran–Armitage trend tests were used to detect trend of surgery rates across age subgroups, and Cochrane Mantel–Haenszel tests were used to detect the significant difference of number of surgeries across the age sub-groups. A p-value <0.05 was considered statistically significant.

Characteristics and costs of selected gynecological surgery

Of the selected surgical procedures, all abdominal hysterectomies and a majority of vaginal hysterectomies (70.5%), laparotomies (72.0%), and oophorectomies (82.7%) took place in an inpatient hospital (Table 5). Length of hospital stay with endometriosis-related surgical procedures averaged 1.5–2.8 days, and most patients were discharged to home self-care after hospitalization. All selected diagnostic and 84.2% of therapeutic laparoscopies took place in an outpatient setting, primarily in outpatient hospitals, physician offices, or ambulatory surgical centers. Average costs of endometriosis-related surgical procedures varied by type, ranging from $4,289 (diagnostic laparoscopy) to $11,397 (abdominal hysterectomy).

Table 5.  Characteristics and cost of surgery.

	Abdominal hysterectomy		Vaginal hysterectomy		Diagnostic laparoscopy		Therapeutic laparoscopy		Laparotomy (n = 25)		Oophorectomy (n = 127)
	(n = 1,010)		(n = 217)		(n = 283)		(n = 2738)
Settings
Inpatient settings, n (%)	1,010	(100)	153	(70.5)	–	–	432	(15.8)	18	(72.0)	105	(82.7)
Outpatient settings, n (%)	–	–	64	(29.5)	283	(96.9)	2,306	(84.2)	7	(28.0)	22	(17.3)
Characteristics of hospital admissions with selected surgery
Length of stay in days, mean (SD)	2.59	(2.31)	1.53	(0.61)	–	–	1.68	(1.40)	2.33	(1.33)	2.81	(1.64)
Discharge status
Home self care, n (%)	1,000	(99)	152	(99)	–	–	430	(100)	18	(100)	105	(100)
Other, n (%)	10	(1)	1	(1)	–	–	2	(0)	0	(0)	0	(0)
Characteristics of surgery performed in outpatient settings
Place of service
Outpatient hospitals, n (%)	–	–	58	(90.6)	247	(87.3)	1,959	(85.0)	6	(85.7)	21	(95.5)
Physician’s office, n (%)	–	–	2	(3.1)	4	(1.4)	45	(2.0)	0	(0.0)	0	(0.0)
Ambulatory surgical centers, n (%)	–	–	1	(1.6)	29	(10.3)	259	(11.2)	1	(14.3)	0	(0.0)
Other, n (%)	–	–	3	(4.7)	3	(1.1)	43	(1.9)	0	(0.0)	1	(4.6)
Total cost of surgical procedures, mean (SD)	$11,397	($8,749)	$8,568	($4,678)	$4,289	($3,313)	$6,856	($5,801)	$7,806	($5,236)	$10,084	($7,443)

All costs were adjusted to 2009 US dollars.

SD, standard deviation.

Discussion

This study reported the rate and costs of surgical procedures received among women with newly diagnosed endometriosis. Approximately 65% of endometriosis patients had select gynecological surgical procedures within 12 months following the initial diagnosis, and the majority of the surgical procedures were performed on the date of diagnosis. The economic burden of these surgical procedures was high, with average costs ranging from $4,289 to $11,397. Surgery rates among patients included in the sensitivity analysis were generally higher than those in the primary analysis.

The rate of undergoing endometriosis-related surgical procedures within 1 year of initial diagnosis varied by patient age, as did the specific surgical procedure performed. Overall, older patients were more likely to receive surgical treatment. A higher proportion of younger patients received conservative surgical procedures (e.g., laparoscopy), whereas the preserve was found for semi-conservative and radical surgeries (e.g., hysterectomy and/or oophorectomy). Such variation reflects patient treatment preference and desire to reserve fertility at a younger age. This study also found that about half of the incidence endometriosis patients received one or more surgical procedures on the index date, suggesting that many patients were confirmed with the diagnosis while receiving surgical treatment. Similar to the findings of this study, Mirkin et al. reported an annual surgery rate of 64% as well as a higher rate among older patients9. However, Mirkin et al. did not assess surgery rates by the type of surgical procedures.

The proportion of patients who had hysterectomy within 1 year of initial endometriosis diagnosis was surprisingly high (28.8%). As the analysis was further stratified by patients’ age, hysterectomy was mainly concentrated among endometriosis patients aged 35–45 years. However, the proportion was still quite high among endometriosis patients aged 25–34 years (15.6%) as compared to the rate of hysterectomy in US civilian residents, which were 2.6 per 1000 females aged 25–29 years, 5.4 aged 30–34 years, 8.9 aged 35–49 years, and 11.7 aged 40–44 years in 199913. The impact of hysterectomy on patient quality-of-life and other clinical outcomes among younger patients needs to be further studied.

According to the medical claims data, 6.2% of patients with endometriosis received diagnosed laparoscopy within 1 year of the initial diagnosis. For ethical reasons, patients with endometriosis should not receive laparoscopy just for diagnosing purposes; lesion ablation or resection should also be performed. Probably the ‘diagnostic laparoscopy’ were coded to reflect the initial purposes of the surgery instead of the actual treatments being performed during the surgery. Additional studies are needed to investigate whether the use of diagnostic laparoscopy procedure codes in the claims data reflects the practice pattern in the real world or a coding problem in the claims data, and if it is the latter, what are the common treatments being conducted during the laparoscopy procedures.

The sensitivity analysis was conducted on a subgroup of patients who had at least one inpatient or two outpatient claims with endometriosis diagnoses. The primary purpose was to exclude patients who had ‘rule-out’ diagnostic tests. Using the criteria, we might have also selected patients with more severe endometriosis. Compared to the whole study sample, patients included in the sensitivity analysis were slightly older and had higher rates of surgical procedures. Furthermore, a higher proportion of patients in the subgroup received abdominal or vaginal hysterectomy. After we stratified the analysis by age, the proportion of patients with any surgical procedure and with hysterectomy was still higher in the subgroup than the whole study sample.

Compared to previous studies, these findings suggest lower average costs of endometriosis-related surgical procedures and shorter average length of hospital stays14,15. Zhao et al. estimated that the length of hospital stays for the majority of endometriosis patients was 3–4 days, with average total hospital charges estimated at $6597 ($13,999 in 2009 dollars) for 1991 and $7450 ($14,719 in 2009 dollars) for 1992 (14). Luciano et al. estimated that inpatient costs were $5014 ($10,640 in 2009 dollars) for laparoscopy and $9533 ($20,229 in 2009 dollars) for laparotomy14.

Limitations

There are certain limitations that should be taken into account with respect to this administrative claims database analysis. Medical conditions documented in the claims were not validated with patient medical records. Sensitivity analysis with more stringent inclusion criteria was conducted and results were similar to those from the main analysis. Severity of endometriosis was not captured in the database, and neither were other clinical presentations such as patient-reported pain severity. Further prospective studies with primary data collection are needed to study the treatments and costs associated with endometriosis of different severity. Furthermore, only 12-month data prior to the index date were assessed to select incidence patients. Some selected patients may have had recurrent endometriosis. Such patients may be more likely to receive surgical treatment than newly diagnosed patients and, thus, the surgery rate among newly diagnosed patients may have been overestimated by categorizing recurrent patients as new patient.

Implications

Surgical treatment was common in endometriosis patients within 1 year of initial diagnosis. More than 65% of patients received some type of surgical procedure. The proportion of patients receiving hysterectomy was surprisingly high, especially in patients who were 25–34 years old. Selection of radical treatment procedures may reflect the refractory characteristic of the disease, the lack of or patient unawareness of other treatment options, and the frustration experienced by patients and healthcare providers. Costs of surgical procedures treating endometriosis were high, ranging from $4,289 (diagnostic laparoscopy) to $11,397 (abdominal hysterectomy). Such costs put significant financial burden on the healthcare system. Given that the recurrence rate of surgical and medical interventions are both high among endometriosis patients and that these interventions have different impacts on patient quality-of-life (both short-term and long-term) and fertility, further studies evaluating the cost effectiveness of various treatment options of endometriosis are needed.

Conclusion

Employing a large claims database, this retrospective study described the real-world practice of the treatment of endometrioses among newly diagnosed patients. More than 65% of patients received some type of surgical procedure within 1 year of the initial diagnosis. Older patients were more likely to receive hysterectomy and less likely to receive diagnostic or therapeutic laparoscopy than younger ones. Costs of endometriosis-related surgical procedures were high.

Transparency

Declaration of funding

The funding of this project was provided by Abbott. Drs Fuldeore, Chwalisz and Marx are employees of Abbott, the manufacturer of Lupron. Dr Wu, Mr. Boulanger, Dr Ma, and Ms Lamothe are employees of United BioSource Corporation, an independent health outcomes research firm contracted by Abbott to conduct this analysis. The statements contained in this paper are solely those of the authors and no endorsement by Abbott should be inferred or implied. There are no other relationships to be declared for any of the authors.

Acknowledgments

There are no formal acknowledgements to be made.

Appendix: List of ICD-9-CM codes for identifying endometriosis-related conditions

Table

Conditions	ICD-9-CM code10
Abdominal pain	789.0
Infertility	628.xx
Migraine	346.x
Depression	296.2x, 296.3x, 300.4, 309.1, 311.xx
Interstitial cystitis	595.1
Irritable bowel syndrome	564.1x
Chronic fatigue syndrome	780.71