Impact of initiation of asenapine on patterns of utilization and cost of healthcare resources associated with the treatment of bipolar I disorder

Abstract

Objective:

To assess the impact of initiation of asenapine on “real-world” levels of utilization and cost of healthcare services for the treatment of bipolar I disorder (BPD) in the US.

Methods:

Using two large US healthcare claims databases that collectively included commercially insured patients aged < 65 years and Medicare enrollees, this study identified all adults (≥18 years) with evidence of BPD who began therapy with asenapine between 2009–2012. The date of the earliest claim for asenapine during this period was deemed the ‘index date’, and patients without continuous enrollment for the 6-month periods before and after this date were excluded (‘pre-index’ and ‘post-index’, respectively). Healthcare claims with a BPD diagnosis, plus psychiatric medications and the costs thereof (2012 dollars) were deemed ‘BPD-related’. Differences in BPD-related utilization and cost of healthcare services were compared between the pre- and post-index periods.

Results:

A total of 1403 patients met all selection criteria; the mean age was 42.8 years and 70.6% were women. Relative to pre-index, significant decreases were noted in post-index use of BPD-related healthcare services, most notably admissions (from 24.0% to 12.3% during the post-index period) and emergency department visits (from 4.6% to 2.6%) (both p < 0.05). While pharmacy costs increased, mean total post-index BPD-related healthcare costs were $979 lower than pre-index ($5002 vs $5981; p < 0.05), primarily due to the decrease in BPD-related admissions.

Conclusions:

Relative to the 6-month period beforehand, levels of utilization of BPD-related healthcare services and costs decreased during the 6-month period immediately following initiation of asenapine therapy.

Key words::

• Bipolar I disorder
• Asenapine
• Healthcare utilization
• Healthcare cost

Introduction

Bipolar I disorder (BPD), also known as manic depressive disorder, is a chronic disorder characterized by unusual shifts in mood, energy, and activity levels; the median age of symptom onset is 25 years1^,2. In the US adult population, ∼5.7 million individuals (2.6% of the US population) have BPD3. These symptoms often are associated with substantial patient burden and a marked deleterious impact on health-related quality-of-life among persons with this disorder4.

BPD is associated with a high prevalence of comorbidities, including both psychiatric disorders (e.g., substance abuse, anxiety disorders) and medical conditions (e.g., thyroid disease, migraine headaches, heart disease, diabetes, obesity)5^,6. Accordingly, it is not surprising that patients with BPD tend to have relatively high levels of health resource utilization (HRU) and attendant healthcare costs4. In 2002, BPD was classified as the most expensive behavioral healthcare diagnosis7; in 2009, the estimated direct cost of BPD in the US was over $30.7 billion8.

A number of pharmacotherapy and psychosocial interventions have proven effective in managing BPD9. Among the former, atypical antipsychotics (AAPs) have been increasingly used in recent years, mainly due to low rates of extra-pyramidal symptoms (some AAPs have dose dependent relationship), negative symptoms (e.g., lack of emotion, interest, and/or expression)10, and conditions such as tardive dyskinesia, as compared to conventional (typical) antipsychotics1, with evidence of efficacy as well as effectiveness in treating BPD11^,12. AAPs have been reported to lower the rate of hospital admissions and emergency room visits, and also decrease total all-cause and mental health-related medical costs in patients with BPD13–16. For instance, the mean per patient reduction in costs for all-cause and mental health-related services was $1395 and $1038, respectively, for aripiprazole16. Asenapine is a relatively new AAP that was approved by the US Food and Drug Administration in 2009 for the treatment of BPD. While its efficacy and safety have been confirmed by a number of clinical studies17–19, there is a lack of ‘real-world’ evidence on the degree to which initiation of asenapine impacts levels of BPD-related HRU and costs in US clinical practice. Consequently, we describe herein the results of our examination of this issue in two large US healthcare claims databases.

Methods

Data sources

All data were extracted from the Truven MarketScan Commercial Claims and Encounters Database and the Truven Medicare Supplemental Coordination of Benefits Database. The former contains administrative claims and eligibility records for ∼30 million enrollees in distinct sets of files for commercially-insured individuals (i.e., working age adults and their dependents) across the US; the latter contains similar information for ∼3 million Medicare enrollees who purchase supplemental commercial insurance. Both databases include information on age, gender, geographical area of residence, health insurance payer type, employment status, monthly enrollment status, and an annual indication of whether the patient’s health plan contains a mental health ‘carve out’ (i.e., does not cover mental health-related services). Both databases also contain detailed inpatient and outpatient medical claims, including diagnoses (in International Classification of Diseases, 9th Version, Clinical Modification [ICD-9-CM] format) and procedures (in ICD-9-CM, Current Procedural Terminology 4th Edition [CPT-4], and Healthcare Common Procedure Coding System [HCPCS] format), dates of service, provider types, places of service, and total reimbursed amounts (i.e., plan payment plus patient liability, e.g., co-pay, deductible). Pharmacy claims include medication dispensed (in National Drug Codes [NDC] format), therapeutic class, dispense date, quantity and days supplied, and total reimbursed amount. Rigorous validation methods are utilized to ensure that claims and enrollment data are complete, accurate, and reliable. Patient identifiers in both databases have been encrypted, and each database is fully de-identified and compliant with the Health Insurance Portability and Accountability Act of 1996 (HIPAA) and, thus, exempt from Institutional Review Board (IRB) approvals. Both databases spanned the period January 1, 2009 to December 31, 2012.

Sample selection

We identified all adults (aged ≥18 years) with ≥1 prescription dispensed for asenapine between August 1, 2009 and December 31, 2012. The date of the earliest claim of asenapine during this period was denoted as the ‘index date’, and patients not continuously enrolled in the database for the 6-month periods immediately before and after this date (‘pre-index’ and ‘post-index’, respectively) were excluded to ensure complete medical and pharmacy data were available. We also excluded patients: (1) without ≥1 medical (i.e., inpatient or outpatient) claim with a diagnosis of BPD (ICD-9-CM diagnosis codes 296.0x, 296.1x, 296.4x, 296.6x, 296.7x, 296.80, 296.81, 296.89) during the 12-month period prior to the index date to identify only patients with BPD receiving asenapine; (2) with evidence of any use of asenapine during the 6-month pre-index period to identify only new users; (3) with evidence of any use of a depot formulation of an AAP during pre- or post-index as it may indicate difficulty complying with oral regimens; (4) with evidence of a mental health insurance carve-out during pre- or post-index as complete utilization and cost data were unavailable; and/or (5) cyclothymic disorder (ICD-9-CM code: 301.13), or BPD induced by substance/medication or another medical condition (ICD-9-CM codes: 291.89, 292.84, 293.83), as some may not respond well to medication as compared to other BPD patients.

We compiled all healthcare claims during the 6-month pre- and post-index periods for all patients who satisfied all selection criteria described above.

Study measures

Demographics and clinical characteristics

Demographic characteristics such as age, gender, insurance type, and geographic region were ascertained as of the index date. We estimated the prevalence of selected psychiatric comorbidities such as schizophrenia, anxiety, depression, and alcohol/substance abuse, as well as the Charlson Comorbidity Index (CCI)20^,21, which is a composite measure of comorbidity. All comorbidities and the CCI were estimated using primary and secondary ICD-9-CM diagnosis codes identified on all pre-index healthcare claims.

Utilization and cost of healthcare services

We examined utilization of BPD-related healthcare services, including; (1) hospitalizations; (2) hospital outpatient visits; (3) physician office visits (including but not limited to psychiatrist visits); and (4) emergency department (ED) visits. Utilization was assessed in terms of the proportions of patients with ≥1 claim for each service, as well as the corresponding number of claims for that service; we also calculated mean length of stay associated with BPD-related admissions.

We examined BPD-related costs, including inpatient care, ED care, outpatient care, and prescription pharmacotherapy; we also examined aggregate BPD-related costs. With the exception of pharmacy, all claims with a BPD diagnosis code (primary or secondary) were designated BPD-related; for pharmacy, claim for any psychiatric medication was considered BPD-related. The costs reported in our study represent the total amount paid for each service/prescription, including the amount paid by the third-party insurer as well as any out-of-pocket payments made by the patient (e.g., deductible, copayment, and coinsurance). All costs were adjusted to 2012 dollars using the annual medical-care component of the Consumer Price Index (CPI) to reflect inflation between 2009–201222.

All measures of healthcare utilization and cost were constituted during the 6-month pre- and post-index periods, respectively.

Statistical analysis

Means (SDs) were reported for continuous variables; numbers (%) were reported for categorical variables. The statistical significance of differences in levels of HRU and cost between the pre- and post-index periods were assessed using paired t-tests for continuous variables and McNemar’s tests for categorical variables. For each assessment, differences associated with a p-value of <0.05 were considered statistically significant. All analyses were undertaken on the aggregate sample, as well as by sub-groups of patients stratified by receipt of other AAPs during pre-index. The purpose of stratification was to assess the impact of initiating asenapine on HRU and costs separately among prevalent and new users of AAPs.

All analyses were performed using SAS software version 9.2 (SAS Institute, Cary, NC).

Results

Sample selection

We identified a total of 1403 patients who were newly started on asenapine for BPD and who met all other selection criteria. Among all such patients, a total of 471 (33.5%) had no evidence of use of AAPs during pre-index, whereas 932 (66.4%) did.

Demographic and clinical characteristics

The mean (SD) age of patients initiating asenapine was 42.8 (14.2) years, and 70.6% were women (Table 1). The mean (SD) CCI score was 0.46 (1.03). Co-morbid mental health disorders were relatively common, including depression (44.1% of the study sample), alcohol/substance abuse (15.3%), anxiety (14.5%), and schizophrenia (10.4%); 10.6% had diabetes.

Table 1. Demographic and clinical characteristics of asenapine patients, by pre-index use of atypical antipsychotics.

	All study subjects	Pre-index use of atypical antipsychotics
		Yes	No
Number of patients	1403	932	471
Age, years, Mean (SD)	42.8 (14.2)	43.12 (14.11)	42.23 (14.23)
Age group, years, %
18–24	16.5	15.7	18.3
25–34	12.5	13.0	11.5
35–44	21.4	21.0	22.1
45–54	27.2	27.6	26.3
55–64	18.5	18.5	18.7
≥65	3.9	4.3	3.2
Female, %	70.6	72.2	67.5
Plan type, %
Health maintenance organization	15.3	16.1	13.6
Point-of-Service	12.1	12.6	11.3
Preferred provider organization	53.7	53.3	54.4
Other	17.9	16.7	20.2
Missing/unknown	1.1	1.3	0.6
Region, %
Northeast	14.3	13.8	15.1
Midwest	28.3	28.5	27.8
South	38.8	37.9	40.6
West	18.4	19.6	15.9
Unknown	0.3	0.1	0.6
Charlson Comorbidity Index Score, Mean (SD)	0.46 (1.03)	0.48 (1.06)	0.41 (0.97)
Co-morbid conditions, %
Schizophrenia	10.4	12.6	6.2
Anxiety	14.5	14.4	14.7
Depression	44.1	45.9	40.6
Alcohol/substance abuse	15.3	15.8	14.4
Diabetes	10.6	10.9	10.0

SD, Standard deviation.

In stratified analyses, the mean age of patients in each sub-group was comparable to that of the aggregate sample. However, those with evidence of pre-index use of AAPs were nominally more likely to be women (72.2% vs 67.5% for those without evidence of use of AAPs during pre-index). They also were nominally more likely to have schizophrenia (12.6% vs 6.2% for those without evidence of use of AAPs during pre-index) or depression (45.9% vs 40.6%); their mean (SD) CCI score also was nominally higher (0.48 [1.06] vs 0.41 [0.97]).

Utilization and cost of BPD-related healthcare services during the pre- and post- index periods

As compared with pre-index, the proportions of patients accessing various BPD-related healthcare services decreased in the 6-month period following the initiation of asenapine, including hospital admissions (from 24.0% during pre-index to 12.3% during post-index), ED visits (4.6% to 2.6%), outpatient hospital visits (18.0% to 15.8%), and physicians’ office visits (77.0% to 73.7%) (all p < 0.05) (Table 2). There were also attendant reductions observed in the mean (SD) number of hospital admissions (from 0.33 [0.70] during pre-index to 0.17 [0.55] during post-index), length of stay in hospital (from 2.69 [7.81] days to 1.29 [5.23] days), and number of ED visits (from 0.05 [0.24] to 0.03 [0.21]) (all p < 0.05).

Table 2. Utilization of BPD-related healthcare services among asenapine patients during the pre- and post-index periods, by pre-index use of atypical antipsychotics.

	All study subjects (n = 1403)		Pre-index use of atypical antipsychotics
			Yes (n = 932)		No (n = 471)
	Pre-index	Post-index	Pre-index	Post-index	Pre-index	Post-index
Hospitalizations, %	24.0	12.3*	25.9	13.5*	20.4	10.0*
Length of stay, mean (SD)	2.69 (7.81)	1.29 (5.23)*	3.08 (8.41)	1.44 (5.71)*	1.93 (6.43)	1.01 (4.10)*
Number of admissions, mean (SD)	0.33 (0.70)	0.17 (0.55)*	0.38 (0.77)	0.20 (0.59)*	0.24 (0.54)	0.13 (0.44)*
Outpatient hospital visits (%)	18.0	15.8*	19.9	17.4	14.4	12.5
Number of outpatient visits, mean (SD)	0.88 (3.40)	0.97 (4.38)	1.07 (3.86)	1.14 (4.89)	0.50 (2.17)	0.64 (3.11)
Physician office visits (%)	77.0	73.7*	78.1	74.7*	74.7	71.8
Number of physician office visits, mean (SD)	5.41 (6.82)	5.55 (7.06)	6.08 (7.28)	6.1 (7.61)	4.08 (5.58)	4.51 (5.68)
Emergency department visits (%)	4.6	2.6*	5.5	2.7*	3.0	2.3
Number of emergency department visits, mean (SD)	0.05 (0.24)	0.03 (0.21)*	0.06 (0.26)	0.04 (0.24)	0.03 (0.20)	0.02 (0.15)*

SD, Standard deviation.

*p-values < 0.05.

Findings were similar among the sub-groups of asenapine patients with and without evidence of use of AAPs during pre-index, albeit not always statistically significant. Both sub-groups experienced significant decreases in BPD-related admissions (from 25.9% to 13.5% among patients with pre-index use of AAPs and from 20.4% to 10.0% among patients without pre-index use of AAPs), with corresponding reductions in the mean (SD) number of admissions (from 0.38 [0.77] to 0.20 [0.59] and from 0.24 [0.54] to 0.13 [0.44]) and length of stay in hospital (from 3.08 [8.41] days to 1.44 [5.71] days and from 1.93 [6.43] days to 1.01 [4.10] days) (all p < 0.05). Patients with pre-index use of AAPs were less likely to go to the ED following the initiation of asenapine (from 5.5% with ≥1 such visits during pre-index to 2.7% during post-index); they also were less likely to visit physicians’ offices (from 78.1% to 74.7%) (all p < 0.05). Patients without pre-index use of AAPs exhibited a modest, albeit statistically significant, decrease in the mean (SD) number of ED visits (from 0.03 [0.20] to 0.02 [0.15]; p < 0.05).

Mean (SD) BPD-related prescription pharmacotherapy costs increased by $839 (from $1334 [$1878] during pre-index to $2173 [$1984] during post-index; p < 0.05) (Table 3). This increase, however, was offset by a corresponding reduction of $1806 in the cost of inpatient care (from $3766 [$11,140] to $1960 [$7988]), leading to a decrease in overall total BPD-related costs of $979 (from $5981 [$11,798] to $5002 [$8851]) (both p < 0.05).

Table 3. Cost of BPD-related healthcare services for asenapine patients during the pre- and post-index periods, by pre-index use of atypical antipsychotics.

	Overall (n = 1403)		Prior use (n = 932)		No prior use (n = 471)
	Pre-index $	Post-index $	Pre-index $	Post-index $	Pre-index $	Post-index $
Mean (SD)
Total	5981 (11,798)	5002 (8851)*	7316 (12,102)	5783 (9402)*	3338 (10,704)	3457 (7413)
Inpatient	3766 (11,140)	1960 (7988)*	4264 (11,335)	2208 (8534)*	2780 (10,686)	1470 (6762)*
Outpatient	858 (1961)	848 (2234)	1023 (2273)	989 (2598)	533 (1042)	570 (1184)
ED	23 (216)	20 (399)	22 (209)	25 (480)	25 (230)	12 (138)
Pharmacy	1334 (1878)	2173 (1984)*	2007 (1989)	2562 (2139)*	0 (0)	1406 (1340)*
Median (IQR)
Total	2125 (587–5789)	2519 (1,210–4729)	3066 (1,475–7157)	3078 (1670–5555)	384 (153–1744)	1563 (671–3171)
Inpatient	0 (0–0)	0 (0–0)	0 (0–1214)	0 (0–0)	0 (0–0)	0 (0–0)
Outpatient	337 (99–819)	311 (53–821)	397 (120–984)	363 (66–910)	244 (56–552)	219 (15–658)
ED	0 (0–0)	0 (0–0)	0 (0–0)	0 (0–0)	0 (0–0)	0 (0–0)
Pharmacy	556 (0–2014)	1692 (708–3024)	1422 (563–2868)	2040 (1059–3486)	0 (0–0)	1013 (539–1973)

SD, Standard deviation; ED, Emergency department; IQR, Interquartile range.

*p-values < 0.05.

In sub-group analyses, initiation of asenapine was associated with a decrease in mean total BPD-related healthcare costs among those with pre-index use of AAPs (from $7316 [$12,102] to $5783 [$9402]; p < 0.05), and nearly no change in such costs among those without evidence of pre-index use of AAPs (from $3338 [$10,704] to $3457 [$7413]; p = 0.83). As in the aggregate sample, mean costs of BPD-related inpatient care decreased significantly in both sub-groups (from $4264 [$11,335] to $2208 [$8534] among patients with pre-index use of AAPs and from $2780 [$10,686] to $1470 [$6762] among patients without pre-index use of AAPs); mean costs of BPD-related prescription pharmacotherapy increased in both sub-groups (from $2007 [$1989] to $2562 [$2139] and from $0 [$0] to $1406 [$1340], respectively) (all p < 0.05).

Discussion

To the best of our knowledge, this is the first study to assess the impact of initiation of asenapine on patterns of utilization and cost of BPD-related care for patients in US clinical practice. We found that patients experienced reductions in levels of utilization and cost of BPD-related healthcare services in the period immediately following initiation of asenapine therapy as compared to the period immediately prior. Observed decreases in healthcare costs were primarily due to reduced costs of BPD-related inpatient care. Moreover, decreases in levels and cost of inpatient care were enough to completely offset increased pharmacotherapy costs related to the initiation of asenapine among patients with evidence of prior use of AAPs, and of sufficient magnitude to render asenapine ‘cost-neutral’ among those without evidence of prior use of AAPs. Our findings, therefore, suggest that asenapine may be deemed a cost-saving option in the treatment of BPD.

The median age at time of initiation of asenapine was ∼44 years, vs the median age of BPD symptom onset, which is 25 years1. First, it may take up to 10 years from the time of symptom onset to a definitive diagnosis of BPD23. Second, AAPs may be reserved until other therapies, such as mood stabilizers, have been found to be ineffective or intolerable. With respect to the latter, the age and gender distribution of patients initiating AAPs in our study approximates those reported in prior studies that were also based on ‘real-world’ data, including Rascati et al.24 (mean age 37 years; 74% women) and Chen et al.25 (mean age = 43 years; 67.7% women). Literature shows that inpatient care accounts for the maximum of total BPD-related cost (50–70%)15^,26. This study found that the initiation of asenapine was associated with a significant decrease in BPD-related inpatient services and associated costs. Similar significant reductions were found among asenapine patients with and without prior use of AAP prescriptions, consequently reducing the total BPD-related cost burden. However, future studies with long-term follow-up are needed to confirm the association between asenapine and reduced inpatient services and costs.

While ours is the first study to assess the ‘real-world’ impact of initiation of asenapine for BPD; our findings are consistent with those of studies that have evaluated the impact of other AAPs on levels of utilization and cost of healthcare services14^,15. Namjoshi et al.27 assessed patients with a diagnosis of BPD disorder with manic and mixed episodes enrolled in a randomized clinical trial who were treated with olanzapine. The study reported savings of almost $900 per month during the 49 weeks of olanzapine therapy. In another study, Pelletier et al.16 observed decreases in mean mental health-related costs of hospitalization ($1038) and length of stay (0.5 days) following the initiation of aripiprazole, but they observed an increase in the cost of pharmacotherapy associated with aripiprazole ($1451); this led to an increase in the overall mean mental health-related costs of $302. These results are as opposed to those of our study, which found the costs of asenapine to be either ‘neutralized’ or totally offset by reductions in the cost of BPD-related healthcare services. However, the differences observed in the magnitude may be due to differences in patient population, definition of BPD, case-selection algorithm, and inconsistencies in study measures such as assessment of ‘mental health-related’ outcomes vs BPD-related outcomes16.

BPD is the most expensive mental healthcare diagnosis. Direct costs, i.e., healthcare costs, of BPD are only about one quarter of indirect costs28. The estimated indirect costs for BPD were ∼$120 billion that include lost productivity and opportunity costs associated with time lost seeking treatment and medical services, and caregiver burden8^,29. The goal of pharmacotherapy along with psychosocial therapies is to offer BPD patients an improved quality-of-life, with maximal maintenance of therapeutic gains over the long-term30. This study suggests that initiation of asenapine reduces inpatient services and ED visits, thus enabling patients to function well enough to benefit further from other psychosocial therapies. However, this study assessed only direct costs, based on healthcare utilization, which are only a small portion as compared to indirect costs. As the reductions in indirect costs were not captured, the study may have under-estimated the benefits of asenapine in BPD. Further research is needed to confirm the findings of this study and to better understand the degree to which these reductions are sustained with continued use of asenapine.

The study has several limitations to be noted. First, no causal inference could be drawn due to the observational nature of the database. Lower observed costs in the post-index period may be due to unmeasured factors such as a BPD-related episode that necessitated an admission and, therefore, increased the cost in the pre-index period followed by change of therapy not specific to asenapine. Following patients for a longer period of time would help to better understand the association between asenapine and HRU and costs. Second, we identified health service claims to be BPD-related by the presence of BPD diagnosis on the claim. Also, for pharmacy we assumed the psychiatric medications were related to management of BPD. Third, our criteria for identifying BPD patients, which required ≥1 medical (inpatient or outpatient) claim 12 months before initiation of asenapine, was less stringent and may have led to selecting patients with only new BPD diagnosis or those with lower BPD severity than studies that required ≥2 claims; or ≥1 inpatient or ≥2 outpatient claims for sample selection16^,31. Further, the claims data did not allow for confirmation of whether the medications were prescribed for BPD. Asenapine is also indicated for schizophrenia, which was prevalent in ∼11% of the study population. However, our study only assessed the outcomes with a diagnosis of BPD; therefore, it is unlikely that less stringent sample selection criteria would impact the results. The presence of schizophrenia diagnosis may be because of the difficulty in distinguishing between BPD and schizophrenia or misclassifying schizophrenia as BPD. Unfortunately, without access to patients’ medical records, the degree to which this occurred is unknowable. Our study used an intent-to treat (ITT) approach, and did not examine adherence to asenapine while assessing the outcomes. Literature shows that overall adherence to AAPs in BPD is very low (8.3%) and non-adherence to treatment in BPD has been associated with increased utilization and expenditures on mental health25. Our use of an ITT approach may have under-estimated the impact of initiation of asenapine on healthcare utilization and costs (i.e., our results may be somewhat conservative). As it was a retrospective observational study using a claims database, there may be errors in recording claims as with any claims database. The database did not capture information on the severity of BPD, and additional research is needed as this may influence costs following initiation of asenapine. Finally, the results presented are most generalizable to similar commercially-insured, Medicare supplemental insurance patients, and may not be generalizable to aggregate population of patients with BPD.

Conclusions

Relative to the period immediately prior, initiation of asenapine was associated with significant reductions in levels of utilization and cost of BPD-related healthcare services in US clinical practice. Depending on patient history of use of atypical antipsychotics, the magnitude of reductions observed in utilization and cost of inpatient care and ED visits totally offset the acquisition costs of asenapine or rendered it cost-neutral. Our findings, therefore, suggest that asenapine may be deemed a cost-saving treatment for the management of BPD. Further studies with longer follow-up are needed to confirm sustained association between asenapine and levels of utilization and costs.

Transparency

Declaration of funding

This research was sponsored by Forest Research Institute, an affiliate of Actavis, Inc. Forest Research Institute had no role in the collection, analysis, interpretation, or reporting of these data.

Declaration of financial/other relationships

SS and SD are currently employees of Forest Research Institute, an affiliate of Actavis, Inc. AC, RW, and AT are current employees, and LB is a former employee of Evidera, which provides consulting and other research services to pharmaceutical, device, government, and non-government organizations. In their salaried positions, they work with a variety of companies and organizations and are precluded from receiving payment or honoraria directly from these organizations for services rendered. Evidera received consultancy fees from Forest Research Institute, an affiliate of Actavis, Inc.

Acknowledgments

Tarun Bhagnani of Evidera is acknowledged for his assistance in manuscript preparation.

Previous presentations

A synopsis of the current research was presented in poster format at the American Psychiatric Association meeting (Toronto, May 16–20, 2015), and the ISPOR 20th Annual International Meeting (May 16–20, 2015, Philadelphia, PA).