Healthcare resource utilization and expenditures of women diagnosed with hypoactive sexual desire disorder

Abstract

Objective:

To describe healthcare utilization and costs among commercially insured women with a diagnosis of hypoactive sexual desire disorder (HSDD) in the US and to compare them with an age-matched control cohort.

Methods:

The Thomson Reuters MarketScan Commercial Database was used to identify women aged 18–64 with an ICD-9-CM coded diagnosis of HSDD from 1/1/1998 to 9/30/2006. A control group of women with no diagnosis of any sexual dysfunction was matched 3:1 to cases based on age, health plan, and enrolment period. Healthcare utilization and costs were examined in the year prior to (pre-period) and following (post-period) index. Multivariate analyses were used to determine the adjusted difference in cost between women with and without HSDD in the post-period.

Results:

In both the pre- and post-periods, women with HSDD had more outpatient office visits, radiology services, prescription medication use, and medical visits (e.g., laboratory and outpatient surgeries) relative to controls. In the 12-month post-period, women with HSDD had significantly higher total costs relative to controls ($5,504 ± 11,132 vs. $4,606 ± 12,601, p < 0.001). After adjusting for clinical characteristics, women with HSDD had total healthcare expenditures that were 16.8% higher than controls (p < 0.001).

Limitations:

There is a potential for selection bias among the women who actually received a diagnosis of HSDD from a clinician. Women who received a diagnosis may be different from women without a diagnosis in ways that cannot be measured in this study. Additionally, it is possible that some women in the control group had HSDD but were undiagnosed. To the extent that the control group included women who did have HSDD, the study estimates of differences between the two groups would be underestimated.

Conclusions:

Women diagnosed with HSDD use significantly more healthcare services than women without diagnosed sexual dysfunction. These higher costs are driven by a greater use of outpatient services and prescription medications.

Key words:

• costs
• healthcare utilization
• hypoactive sexual desire disorder
• multivariate analysis

Introduction

Current research provides fairly consistent evidence that roughly 9–10% of women over the age of 18 suffer from symptoms of decreased sexual desire with distress, which could be hypoactive sexual desire disorder (HSDD)1,2. The increasing recognition of HSDD as a health condition among both pre- and post-menopausal women has been accompanied by a growth in the literature on the diagnosis of female sexual dysfunctions and potential treatments3. According to the Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV), HSDD is characterized by a persistent deficiency or absence of sexual fantasies or thoughts, and/or the absence of desire for sexual activity4. To be diagnosed, the symptoms must cause marked distress or interpersonal difficulty and cannot be accounted for by medications or the presence of another psychiatric or medical disorder.

In addition to improved diagnosis and the development of screening tools5,6, research has begun to document the emotional and quality of life burden of HSDD. A recent telephone survey of 1,189 postmenopausal women in stable relationships revealed that women meeting some of the criteria for HSDD (e.g., low sexual desire with concomitant sexually related distress) had significantly lower scores on several dimensions of health-related quality of life relative to women who didn’t meet any of the criteria for HSDD7. Specifically, using the SF-12, women with distressing low desire scored about 5.6 points lower than women without distressing low desire on the mental health dimension, 4.3 points lower on vitality, 3.6 points lower on social function, and 5.3 points lower on bodily pain.

Despite the increasing literature on the impact that low sexual desire can have on a woman's emotional state and quality of life, no data currently describe healthcare resource utilization and associated medical costs for these women. It is not known whether women with a diagnosis of HSDD use more health services and have a greater prevalence of particular comorbid conditions than women without an HSDD diagnosis, or whether undiagnosed HSDD leads to the utilization of potentially unnecessary service and inappropriate medications. Administrative claims data were used to examine the types of healthcare services used before and after the patient received an ICD-9-CM coded diagnosis of HSDD in order to determine whether resource utilization differed for women with HSDD relative to women without an HSDD diagnosis and whether there were overall differences in the costs of healthcare utilization for women with and without an HSDD diagnosis. The resulting costs associated with potentially higher utilization in this population should be of importance to both payers and patients.

Patients and methods

Data for this study came from the Thomson Reuters MarketScan Commercial Claims and Encounters (Commercial) Database, which contains information for the population under age 65 covered by private-sector health plans. These data include health insurance claims across the continuum of care (e.g., inpatient, outpatient, outpatient prescription drug, carve-out behavioral health) as well as enrolment data from large employers and health plans across the United States who provide private healthcare to several million employees and their dependents annually. This administrative claims database includes a variety of fee-for-service and managed-care health plans, including exclusive provider organizations, preferred provider organizations (PPOs), point-of-service (POS) plans, indemnity plans, health maintenance organizations (HMOs), and consumer-driven healthcare plans (CDHPs). Patients selected for the study were women aged 18–64 with at least one claim of HSDD (ICD-9-CM code 302.71) between January 1, 1998, and September 30, 2006. The index date was defined as the date of the evaluation and management service or psychiatric (E&M/Psych) claim (CPT codes 99201–99499, 90801–90802, 90804–90899, 96150–96151, 96152–96155) with an HSDD diagnosis or, for non-E&M/Psych claims with an HSDD diagnosis, the date of the E&M/Psych claim with an HSDD diagnosis closest to the date on the non-E&M/Psych HSDD claim within the previous 14 days. For study inclusion, patients were required to show continuous enrolment with pharmacy benefits in the database for the 12 months before and 12 months after the index date.

Excluded from the study were women aged 50 through 64 with evidence of pregnancy or birth, patients with any HSDD diagnosis (as defined above) occurring in the 12 months leading up to the index date, and women aged 18 through 49 with indication of hysterectomy or bilateral oophorectomy or a diagnosis of menopausal symptoms in the 12 months before the index date.

A cohort of women without HSDD was matched to the HSDD cases at a ratio of 3:1. Women in the matched control group were required to have no diagnostic indication of either HSDD or any diagnosis related to female sexual dysfunction at any time, 24 months of continuous enrolment between 1/1/1997 and 9/30/2007, at least one healthcare claim in the study period, and none of the exclusion criteria applied to the HSDD cohort. This direct matching was based on age, health plan type, and number of months of continuous enrolment in the database. These variables were chosen to determine whether women with HSDD commonly had more comorbidities or used more health services than did women in general.

Statistical comparisons between HSDD cases and controls were conducted using chi-squares for categorical variables and t-tests for continuous variables. Given that approximately 50 variables were tested for statistical difference between the groups, a Bonferroni correction with an alpha of 0.05 and n = 50 was applied to set the threshold for statistical significance at p < 0.0018. Thus, only differences where the p-value was less than 0.001 were reported as statistically significant.

Utilization in the pre-period was compared to that in the post-period in the HSDD cases and controls. Paired t-tests were performed to measure statistical significance.

A multivariate analysis of the difference in medical costs between HSDD cases and controls was conducted to control for statistically significant clinical differences between the two groups. A generalized linear model with a gamma distribution and a log link using the SAS GENMOD procedure served to account for the skewed distribution of the total medical costs. The exponent of the parameter estimate can be interpreted as the percentage difference between the variable of interest and its associated referent group.

Demographics and clinical characteristics

Demographic variables were captured from the claims and included patient age, insurance type, geographic region, and physician specialty on the index claim. Clinical events and diagnoses that could lead to decreased sexual desire were identified in the 12-month period prior to the HSDD diagnoses. These events and diagnoses included clinical correlates of HSDD such as gynecological surgeries (for women aged 50–64), pregnancy (for women aged 18–49), birth control prescriptions, and genito-urinary infections; medications that may lead to decreased sexual desire, including anticholinergics, antidepressants, antihypertensives, barbiturates, benzodiazepines, chemotherapy, clomiphene, GnRH analogues, and narcotics; and comorbid conditions associated with HSDD such as diabetes, anxiety, thyroid abnormality, cancer or history of cancer, depression, alcohol or substance abuse, fibromyalgia, irritable bowel syndrome, and migraine9–15. A modified Deyo–Charlson Comorbidity Index (excluding pregnancy) was used as an aggregate measure of comorbidity16.

Healthcare resource utilization and costs

Healthcare resource utilization and costs were assessed for three categories of services: inpatient and emergency room (ER) care, outpatient care, and prescription medications. Inpatient care included hospital admissions and ER visits. Outpatient care was broken into medical office visits, radiology, other medical services (i.e., laboratory and outpatient surgeries or procedures), and behavioral health services. Behavioral services included psychiatric, counselor, or psychologist office visits, as defined by CPT and HCPCS codes.

The total number of unique prescriptions filled was calculated in the pre- and post-periods for potential HSDD treatments as well as for all other medications. Medications identified as potential treatments for HSDD included bupropion, estrogen, testosterone, PDE-5 inhibitors, and DHEA3,6.

The costs of all claims in each of the above resource categories were summed to determine the total costs of each service type as well as the total medical costs for each patient. Costs in the claims database used for this study represent the actual dollar amounts paid by the insurer and the patient (i.e., deductible, co-payments, and coinsurance) for each service provided. All costs were inflated to 2007 US dollars by multiplying each year's cost by the Medical Care Consumer Price Index.

Results

Overall study sample

Between January 1, 1998, and September 30, 2006, a total of 15,488 women had at least one diagnosis of female sexual dysfunction on a non-diagnostic medical claim. Among them, 4,831 had at least 12 months of continuous enrolment both before and after the first diagnosis. Women with HSDD represented the single largest diagnostic category, making up 59% (n = 2,870) of women in the total female sexual dysfunction sample. These women with HSDD comprised the study sample for this analysis. A total of 8,610 women without HSDD were matched in a ratio of 3:1 to women with HSDD.

Within the HSDD group, 70.5% (n = 2,024) fell within the ages of 18–49, with the remaining 29.5 % (n = 846) aged 50–64 (Table 1). The mean age across all women was 42.8 (SD 10.3) and was the same for both HSDD cases and controls.

Table 1.  Demographics: all hypoactive sexual desire disorder (HSDD) vs. all controls.

	HSDD n = 2,870		Controls n = 8,610
	n/mean	%/SD	n/mean	%/SD
Mean (SD) age	42.8	10.3	42.8	10.3
Geographic region*
Northeast	225	7.8%	947	11.0%
North Central	761	26.5%	2,215	25.7%
South	1,259	43.9%	3,666	42.6%
West	580	20.2%	1,636	19.0%
Unknown	45	1.6%	146	1.7%
Plan type
Indemnity	541	18.9%	1,621	18.8%
HMO	454	15.8%	1,370	15.9%
POS	309	10.8%	931	10.8%
PPO	1,188	41.4%	3,586	41.6%
CDHP	5	0.2%	40	0.5%
Other	351	12.2%	1,002	11.6%
Missing/unknown	22	0.8%	60	0.7%
Diagnosing provider type
PCP	858	29.9%	2,574	29.9%
Specialist
OB/Gyn	1,619	56.4%	4,857	56.4%
Psychiatrist	17	0.6%	51	0.6%
Counseling	117	4.1%	351	4.1%
Other	209	7.3%	627	7.3%
Unknown/missing	50	1.7%	150	1.7%
Relationship*
Employee	1,348	47.0%	4,652	54.0%
Dependent	1,522	53.0%	3,958	46.0%
Index year
1998	59	2.1%	173	2.0%
1999	152	5.3%	457	5.3%
2000	212	7.4%	634	7.4%
2001	212	7.4%	850	9.9%
2002	461	16.1%	1,376	16.0%
2003	528	18.4%	1,590	18.5%
2004	487	17.0%	1,463	17.0%
2005	354	12.3%	1,066	12.4%
2006	334	11.6%	1,001	11.6%

*Chi-square test for differences between HSDD cases and controls is significant at p < 0.001.

SD, standard deviation; HMO, health maintenance organization; POS, point of service; PPO, preferred provider organization; CDHP, consumer driven health plan; PCP, primary care provider.

Clinical characteristics

Although there was no statistical difference between HSDD cases and controls using the Deyo–Charlson Comorbidity Index as a measure of general health status (Table 2), women with HSDD had a significantly higher prevalence of medical conditions known to be correlates to female sexual dysfunction, comorbid conditions known to increase the likelihood of sexual dysfunction, and prescription medications known to induce sexual dysfunction. Women with HSDD were significantly more likely to have undergone a gynecological surgery (3.0 vs. 1.1%; p < 0.001), been pregnant in the year prior (6.0 vs. 0.1%; p < 0.001), had a prescription fill for birth control (19.2 vs. 9.8%; p = <0.001), and had a genito-urinary infection (18.9 vs. 11.5%; p < 0.001). Women with HSDD were also more likely to have been diagnosed with anxiety, depression, and migraine relative to women without HSDD (p < 0.001) (Table 2). Consistent with the greater diagnosis of depression and anxiety, women with HSDD had higher use of SSRIs and benzodiazepines (p < 0.001). On the other hand, women with HSDD were less likely to have diabetes (2.8 vs. 4.5%, p < 0.001) and use an antihypertensive (12.8 vs. 15.3%, p < 0.001) than were women without HSDD.

Table 2.  Clinical characteristics of women with hypoactive sexual desire disorder (HSDD) and controls in the pre-period.

	HSDD n = 2,870		Controls n = 8,610
	n	%	n	%
Correlates of female sexual dysfunction
Gynecological surgeries*	86	3.0%	93	1.1%
Pregnancy*	173	6.0%	6	0.1%
Evidence of birth control prescription*	550	19.2%	841	9.8%
Genitourinary infections*	542	18.9%	988	11.5%
Common comorbid conditions
Diabetes*	80	2.8%	388	4.5%
Anxiety*	137	4.8%	140	1.6%
Depression*	296	10.3%	379	4.4%
Thyroid abnormality	237	8.3%	620	7.2%
Cancer or history of cancer	102	3.6%	302	3.5%
Alcohol or substance abuse	14	0.5%	35	0.4%
Fibromyalgia	72	2.5%	151	1.8%
Migraine*	104	3.6%	175	2.0%
Modified Deyo–Charlson Comorbidity Index (mean, SD)	0.22	0.65	0.25	0.70
Prescription medications
Antidepressants
Selective serotonin reuptake inhibitors (SSRIs)*	548	19.1%	965	11.2%
Tricyclic antidepressants (TCAs)	97	3.4%	229	2.7%
Monoamine oxidase inhibitors (MAOIs)	0	0.0%	1	0.0%
Other antidepressants (other than bupropion)*	236	8.2%	335	3.9%
Antihypertensives*	368	12.8%	1,317	15.3%
Benzodiazopines*	341	11.9%	752	8.7%
Chemotherapy	23	0.8%	47	0.5%
Clomid	4	0.1%	18	0.2%
GnRH analogues	8	0.3%	10	0.1%
Narcotics (opiates)*	587	20.5%	1,356	15.7%
Standard of care for HSDD treatment
CNS mediated agents
Buproprion*	244	8.5%	239	2.8%
Dopamine agonists	8	0.3%	17	0.2%
Hormones
Testosterone*	244	8.5%	19	0.2%
Estrogen*	727	25.3%	1,287	14.9%
DHEA	1	0.0%	0	0.0%
Vasoactive agents	0	0.0%	0	0.0%
PDE-5 inhibitors	4	0.1%	2	0.0%

*Chi-square test for differences between HSDD cases and controls is significant at p < 0.001.

SD, standard deviation; GnRH, gonadotropin-releasing hormone; CNS, central nervous system; DHEA, dehydroepiandrosterone; PDE-5, phosphodiesterase type 5.

Resource utilization

Prior to diagnosis of HSDD, women with HSDD used more healthcare services than their matched controls (Table 4). In particular, these women had more outpatient office visits, radiology, and other medical visits such as laboratory and outpatient surgeries. Although the amount of use was not substantial, women with HSDD also had more inpatient admissions and behavioral health office visits and filled more prescriptions compared with matched controls. Details on the types of medications used by women with and without HSDD are shown in Table 2 for the pre-period and are discussed above.

These trends toward greater utilization of primarily office-based care and prescription medications remained and increased somewhat in the post-diagnosis period (Table 4). Women with HSDD had fewer inpatient admissions than their controls did after diagnosis, which was not true prior to diagnosis. No differences were noted in ER use between the two groups of women in either time period.

The study also measured the difference in pre- and post-utilization for women with HSDD and their matched controls (Table 5). In the women with HSDD, small but significant increases from pre-diagnosis to post-diagnosis were noted for outpatient medical office visits, outpatient radiology, other outpatient medical visits such as laboratory and surgeries, outpatient behavioral health and both HSDD- and non-HSDD related prescriptions. Similar trends were seen in the controls for all but HSDD-related prescriptions but with even smaller increases over time. Inpatient admissions, though not substantial, were fewer after diagnosis for the women with HSDD and greater for the controls.

Costs

Consistent with the greater utilization of outpatient-based medical and behavioral healthcare, women with HSDD also had higher overall medical expenses relative to women without HSDD. Details on unadjusted medical expenditures by type of service in the pre- and post-periods are shown in Table 3. On average, women with HSDD incurred medical expenses higher by $1,146 (or about 27.9%) than those of controls in the year prior to diagnosis. In the year following diagnosis, expenditures for both groups of women increased slightly, but a difference of $897 (19.5% higher) remained between women with HSDD and controls (Table 3). The largest drivers of medical costs for women with HSDD were outpatient medical visits such as laboratory and outpatient surgeries and prescription medications.

Table 3.  Unadjusted healthcare costs for women with and without hypoactive sexual desire disorder (HSDD) in the pre- and post-periods.

	Women with HSDD		Matched Cohort		Delta	p-value
	n = 2,870		n = 8,610
	Mean	SD	Mean	SD
Pre-period expenditures ($)
Inpatient admissions	1,157.05	6,740.99	915.33	5,746.47	241.72	0.062
Emergency room visits	125.49	556.37	109.63	770.58	15.86	0.309
Outpatient medical, office visits	395.34	514.99	270.31	395.93	125.03	<0.001
Outpatient radiology	378.20	1,344.76	270.34	1,119.58	107.86	<0.001
Outpatient medical, other*	1,913.40	4,317.15	1,521.45	5,118.24	391.95	<0.001
Outpatient behavioral health, office visits	107.50	487.49	47.32	349.60	60.18	<0.001
Outpatient behavioral health, other†	6.96	118.27	4.67	114.40	2.29	0.357
Standard of care for HSDD treatment	101.98	259.51	57.12	191.89	44.86	<0.001
All other pharmaceutical	1,051.97	2,047.65	898.21	2,211.88	153.76	0.001
Total pre-period expenditures	5,240.60	10,486.09	4,094.38	10,466.45	1,146.22	<0.001
Post-period expenditures ($)
Inpatient admissions	863.07	6,754.74	968.31	6,695.99	(105.24)	0.467
Emergency room visits	146.82	690.84	115.93	869.69	30.89	0.084
Outpatient medical, office visits	469.76	535.52	288.56	409.73	181.20	<0.001
Outpatient radiology	400.94	1,394.78	317.18	1,161.79	83.76	0.002
Outpatient medical, other*	2,063.24	4,626.76	1,765.24	6,195.58	298.00	0.018
Outpatient behavioral health, office visits	158.17	608.14	51.14	346.13	107.03	<0.001
Outpatient behavioral health, other†	6.41	82.55	3.45	92.26	2.96	0.127
Standard of care for HSDD treatment	151.97	357.83	61.87	216.90	90.10	<0.001
All other pharmaceutical	1,243.24	2,350.34	1,034.56	2,563.81	208.68	<0.001
Total post-period expenditures	5,503.62	11,132.15	4,606.24	12,600.62	897.38	<0.001

*Includes other professional visits, outpatient surgeries, laboratory, and all other outpatient services.

†Includes any service, other than office visit, provided by psychiatrist, psychologist, or counselor.

Table 4.  Comparison of utilization for the hypoactive sexual desire disorder (HSDD) cohort and matched controls during the pre-index and post-index periods.

Utilization measure	Pre-index				Post-index
	HSDD cohort	Matched controls	Difference	p-value*	HSDD cohort	Matched controls	Difference	p-value*
Number of patients	2,870	8,610	n/a	n/a	2,870	8,610	n/a	n/a
Inpatient admissions	0.10	0.07	0.03	0.0003	0.06	0.08	(0.02)	0.0081
Emergency room visits	0.22	0.21	0.01	0.7049	0.22	0.20	0.02	0.4110
Outpatient medical office visits	4.72	3.25	1.47	<0.0001	5.70	3.47	2.23	<0.0001
Outpatient radiology services	2.12	1.73	0.39	<0.0001	2.36	1.96	0.40	<0.0001
Other outpatient medical services	15.70	11.64	4.05	<0.0001	17.67	12.59	5.07	<0.0001
Outpatient behavioral health office visits	1.00	0.45	0.56	<0.0001	1.50	0.51	0.99	<0.0001
Other outpatient behavioral health services	0.07	0.04	0.03	0.1201	0.05	0.05	0.00	0.9666
Standard of care for HSDD treatment	1.63	0.93	0.70	<0.0001	2.10	0.91	1.19	<0.0001
All other pharmaceutical	14.38	11.65	2.73	<0.0001	15.96	12.91	3.05	<0.0001

*Two-sample t-test.

Table 5.  Comparison of utilization in the pre-index and post-index periods for the hypoactive sexual desire disorder (HSDD) cohort and matched controls.

Utilization measure	HSDD cohort				Matched control cohort
	Pre-index	Post-index	Difference	p-value*	Pre-index	Post-index	Difference	p-value*
Number of patients	2,870	2,870	n/a	n/a	8,610	8,610	n/a	n/a
Inpatient admissions	0.10	0.06	(0.04)	<0.0001	0.07	0.08	<0.01	0.5826
Emergency room visits	0.22	0.22	0.00	1.0000	0.21	0.20	(0.01)	0.5380
Outpatient medical office visits	4.72	5.70	0.98	<0.0001	3.25	3.47	0.22	<0.0001
Outpatient radiology services	2.12	2.36	0.24	0.0094	1.73	1.96	0.23	<0.0001
Other outpatient medical services	15.70	17.67	1.97	<0.0001	11.64	12.59	0.95	<0.0001
Outpatient behavioral health office visits	1.00	1.50	0.50	<0.0001	0.45	0.51	0.06	0.0222
Other outpatient behavioral health services	0.07	0.05	(0.01)	0.6191	0.04	0.05	0.01	0.4162
Standard of care for HSDD treatment	1.63	2.10	0.47	<0.0001	0.93	0.91	(0.01)	0.5078
All other pharmaceutical	14.38	15.96	1.58	<0.0001	11.65	12.91	1.25	<0.0001

*Paired t-test.

Given the clinical differences between women with HSDD and the controls (Table 2), multivariate analyses controlling for correlates and potential confounders of HSDD that might also contribute to higher healthcare expenditures were conducted for each time period. The results of the multivariate model for the pre-period expenditures appear in Table 6. Although many of the comorbid conditions have large and significant effects on total healthcare expenditures as noted by the magnitude of the parameter estimate and the p-value, the parameter estimate for HSDD was statistically significant and indicated that women with HSDD had pre-period expenditures 12.6% higher than those of the controls. Similar results were found for post-period expenditures (Table 7), where the difference in total costs after adjusting for the clinical differences was just slightly higher at about 16.8%.

Table 6.  Regression results for pre-period costs.

Explanatory items (measured at index)	Relative cost compared to absence of item	Lower 95% CI	Upper 95% CI	p-value
HSDD vs. controls	1.13	1.07	1.19	<0.0001
Diabetes	1.97	1.78	2.19	<0.0001
Thyroid abnormality	1.34	1.24	1.44	<0.0001
Chemotherapy	3.15	2.43	4.08	<0.0001
Cancer diagnosis	2.18	1.95	2.43	<0.0001
Fibromyalgia	1.41	1.22	1.63	<0.0001
Migraine	1.10	0.96	1.25	0.1596
Anxiety diagnosis	1.03	0.90	1.17	0.711
Depression diagnosis	1.29	1.18	1.42	<0.0001
Substance abuse	1.32	0.97	1.80	0.0767
SSRI	1.35	1.26	1.44	<0.0001
Other antidepressant	1.51	1.39	1.63	<0.0001
Benzodiazepine	1.27	1.18	1.37	<0.0001
Narcotics	2.11	1.99	2.23	<0.0001
GnRH analogues	1.83	1.11	3.03	0.0185
Clomid	1.56	0.99	2.45	0.0564
Antihypertensives	1.70	1.60	1.81	<0.0001
Gyn surgery	3.05	2.60	3.59	<0.0001
Genitourinary infection	1.40	1.32	1.49	<0.0001

Generalized linear model with gamma-distributed error and log link.

Goodness of fit assessed by comparing deviance and degrees of freedom to asymptotic chi-square (p = 0.24).

SSRI, selective serotonin reuptake inhibitors; GnRH, gonadotropin-releasing hormone; Gyn, gynecological.

Table 7.  Regression results for post-period costs.

Explanatory items (measured at index)	Relative cost compared to absence of item	Lower 95% CI	Upper 95% CI	p-value
HSDD vs. controls	1.17	1.11	1.24	<0.0001
Diabetes	1.83	1.64	2.05	<0.0001
Thyroid abnormality	1.27	1.17	1.38	<0.0001
Chemotherapy	3.16	2.39	4.17	<0.0001
Cancer diagnosis	2.25	2.00	2.54	<0.0001
Fibromyalgia	1.45	1.24	1.70	<0.0001
Migraine	1.25	1.08	1.44	0.0023
Anxiety diagnosis	0.97	0.84	1.12	0.6637
Depression diagnosis	1.25	1.13	1.38	<0.0001
Substance abuse	1.12	0.81	1.55	0.5002
SSRI	1.20	1.12	1.29	<0.0001
Other antidepressant	1.39	1.27	1.52	<0.0001
Benzodiazepine	1.42	1.31	1.54	<0.0001
Narcotics	1.30	1.22	1.38	<0.0001
GnRH analogues	1.80	1.04	3.09	0.0346
Clomid	2.16	1.33	3.53	0.002
Antihypertensives	1.54	1.45	1.65	<0.0001
Gyn surgery	0.95	0.80	1.14	0.5994
Genitourinary infection	1.24	1.16	1.32	<0.0001

Generalized linear model with gamma-distributed error and log link.

Goodness of fit assessed by comparing deviance and degrees of freedom to asymptotic chi-square (p = 0.26).

SSRI, selective serotonin reuptake inhibitors; GnRH, gonadotropin-releasing hormone; Gyn, gynecological.

Discussion

A growing body of research describes the emotional and quality-of-life implications of decreased sexual desire with distress for women and their spouses1,2,7. No research to date, however, has examined the healthcare resource utilization and costs for women with HSDD, a subgroup of women with distressing low desire, relative to women without HSDD. This study examined the healthcare experiences of 2,870 women aged 18–64 with commercial health insurance and compared them with a matched group of women with no diagnostic indicators of any sexual dysfunction diagnosis.

The results show that women with HSDD used a significantly higher number of outpatient services and prescription medications relative to controls – specifically, greater utilization of medical office visits and other services such as laboratory and outpatient services. In addition, women with HSDD filled more prescriptions for medications in the 1 year following diagnosis than did the matched controls.

Although women diagnosed with HSDD did use more behavioral health services than women without HSDD, the overall use was low. Perhaps when individuals seek mental health-related services, they choose to pay for such services out of pocket to maintain privacy17. Thus the use of behavioral health services detected in this study may be a low estimate of true utilization rates.

Although the two groups of women studied had similar scores on the Deyo–Charlson Comorbidity Index indicating that their general health was similar to that of women without HSDD, women with HSDD had significantly more comorbid conditions that were potential correlates and confounders of HSDD9–14. Women with HSDD were significantly more likely to have undergone gynecological surgery, been pregnant in the year prior, and had a genito-urinary infection than were controls. They were also more likely to have been diagnosed with anxiety, depression, and migraines in the year prior to diagnosis. Even after controlling for these differences in clinical characteristics, the study showed that women with HSDD still had total healthcare costs that were 12.6% higher in the pre-period and 16.8% higher in the post-period than those of the controls.

In interpreting the results of this study, several limitations should be noted. First is the potential for selection bias among the women who actually received a diagnosis of HSDD from a clinician. Women who received a diagnosis may differ from women without a diagnosis in ways that could not be measured in this study, such as overall propensity to use healthcare services and their attitudes towards sex and their sexual relationships. It is possible that some women in the control group had HSDD but were undiagnosed. To the extent that the control group included women who did have HSDD, the differences between the two groups would be underestimated. Furthermore, not all conditions diagnosed by a physician appear on the medical billing claim, possibly leading to an underestimation of the prevalence of comorbid conditions as well as the presence of HSDD. The potential underestimation of the use of behavioral health services must also be acknowledged, and the estimates provided here should not be interpreted as the actual prevalence of use of such services. However, the utilization and costs presented here do accurately reflect the services for which a third-party payer would be responsible and can therefore be used in future studies examining the costs of HSDD from a payer perspective.

Conclusion

This study provides some of the first data thus far reported on the use and costs of healthcare services among women with HSDD compared with women without HSDD. These results fill a gap in the current literature and will provide payers with a better understanding of utilization patterns and healthcare costs in women with HSDD. The findings indicate that women diagnosed with HSDD use significantly more healthcare services than do women without diagnosed sexual dysfunction. These higher costs are driven by a greater use of outpatient services and prescription medications.

Transparency

Declaration of funding

This work was funded by Boehringer-Ingelheim Pharmaceuticals, Inc (BIPI).

Declaration of financial/other relationships

K.F. has disclosed that she was an employee of Thomson Reuters at the time of this study, and B.J. has disclosed that she is a current employee of Thomson Reuters, a company that received funding from BIPI to conduct this study. D.F. has disclosed that he is an employee of BIPI.

Reviewers on this manuscript have disclosed that they have no relevant financial relationships.

Notes

*MarketScan is a registered trademark of Thomson Reuters (Healthcare) Inc.