November 2011  << Back  

  Can't open the PDF? Click here for help.



Patients with obsessive-compulsive disorder vs depression have comparable health care costs: A retrospective claims analysis of Florida Medicaid enrollees

Cheryl S. Hankin, PhD

BioMedEcon, LLC, Moss Beach, CA, USA

Lorrin Koran, MD

Department of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, CA, USA

David V. Sheehan, MD

Department of Psychiatry and Behavioral Medicine, South Florida College of Medicine, Tampa, FL, USA

Eric Hollander, MD

Montefiore Medical Center University Hospital, Albert Einstein College of Medicine, New York, NY, USA

Larry Culpepper, MD

Department of Family Medicine, Boston University School of Medicine, Boston, MA, USA

Donald W. Black, MD

Department of Psychiatry, Carver College of Medicine, Iowa City, IA, USA

John Knispel, MD

Knispel Associates, Singer Island, FL, USA

Jeffrey Dunn, PharmD

SelectHealth, Inc., Salt Lake City, UT, USA

Darin D. Dougherty, MD

Department of Psychiatry, Harvard Medical School, Obsessive-Compulsive Disorder Institute, Massachusetts General Hospital, Boston, MA, USA

Amy Bronstone, PhD

BioMedEcon, LLC, Moss Beach, CA, USA

Zhaohui Wang, MS

BioMedEcon, LLC, Moss Beach, CA, USA

BACKGROUND: The health care burden of obsessive-compulsive disorder (OCD) is relatively unknown.

OBJECTIVE: To compare the health care burden of patients with OCD vs depression.

METHODS: This retrospective claims analysis compared the 2-year median per-patient health care claims and costs for Florida Medicaid adult enrollees (1997 to 2006) newly diagnosed with “pure OCD” (P-OCD; OCD without comorbid major depression, bipolar disorder, psychosis, organic mental disorder, pervasive developmental disorder, nonpsychotic brain damage, developmental delay, or mental retardation) with matched patients newly diagnosed with “pure depression” (P-D; similar to P-OCD but excluding OCD instead of depression).

RESULTS: Eighty-five newly diagnosed P-OCD patients were matched with 14,906 P-D patients. Although median per-patient total health care costs were comparable across groups, patients with P-D incurred significantly higher median outpatient medical costs ($1,928 vs $363, P = .003), while those with P-OCD incurred almost three-fold greater psychiatric costs ($2,028 vs $759, P < .0001). The latter was due primarily to significantly higher costs of psychotropic medications among those with P-OCD ($4,307 vs $2,317, P = .0006) rather than to psychiatric outpatient care.

CONCLUSIONS: Patients with P-D and P-OCD carry a similar burden in overall health care costs. However, the burden of those with P-D was largely attributable to outpatient medical costs while that of those with P-OCD was due to higher costs of psychotropic medications.

KEYWORDS: obsessive-compulsive disorder, major depression, health care resources, retrospective claims analysis



The burden of depression borne by health care systems, in terms of increased resource use and costs, is well documented. Yet, relatively little is known about the health care burden associated with specific anxiety disorders,1-3 including obsessive-compulsive disorder (OCD). Depressed persons use significantly more outpatient, inpatient, emergency, laboratory, diagnostic, and specialty services than non-depressed persons.4-9 Total health care costs are 50% to 100% higher among depressed vs non-depressed outpatients,8,10-12 and are comparable to or exceed those incurred by patients with chronic physical diseases, such as diabetes, heart disease, or hypertension.13 The substantial burden of depression cannot fully be explained by differences in physical illness severity4,8,10,12 or use of mental health services.8

In contrast to depression, the health care burden of anxiety disorders is less well known, despite their importance. Anxiety disorders accounted for approximately 12 million physician office visits annually in the United States in 1997,14 and approximately 15% of general medicine patients satisfy criteria for a current anxiety disorder.2,15,16 Anxiety disorders account for approximately one-third of US mental health expenditures, with an annual estimated cost of $42.3 billion in 1990 ($101.8 billion in 2010).17

Within the group of anxiety disorders, particularly little is known about the health care burden associated with OCD, which has an estimated prevalence of 1% to 2%.18,19 Direct costs associated with OCD were estimated at only $2.1 billion annually in 1990 ($5.0 billion in 2010),20 but it is unclear how per-patient health care use and costs may differ among those with OCD vs those with more common psychiatric disorders, such as depression. Some evidence suggests that persons with OCD, which is characterized by obsessive, distressing, intrusive thoughts (obsessions) and/or compulsions (tasks or rituals),21 may tend to overutilize medical care services because of obsessions about contamination or illness.21 In a retrospective analysis of claims data from a large, pre-paid, private health plan, Koran et al reported that patients with OCD had 63% higher mean annual costs for nonpsychiatric (ie, medical) visits, and 56% higher costs for laboratory and radiology services, than patients with no psychiatric visits.22 Kennedy et al reported that patients with OCD visited dermatologists more frequently than did patients with other anxiety disorders.23 Also, several small studies have found that an unusually high percentage (14% to 25%) of dermatology patients satisfy criteria for current OCD.24-26 Investigators attribute compulsive behaviors (such as excessive washing or picking) and somatic obsessions (such as exaggerated focus on bodily appearance) to high rates of OCD among dermatology patients.24-28

This study aimed to compare the per-patient health care burden associated with adult-onset OCD vs depression. While the same drug class (selective serotonin reuptake inhibitors [SSRIs]) is used as first-line pharmacotherapy for both disorders, the overall health care burden has been well documented for depression but not OCD.21,29 This study addresses 2 major questions:

  1. How do adult patients newly diagnosed with OCD differ from those newly diagnosed with depression in terms of demographic and illness characteristics?

  2. How do health care resource use and costs in the 24 months after diagnosis differ in an adult with OCD who seeks treatment as compared with a similar individual who seeks treatment for major depression?


Florida Medicaid dataset

Florida Medicaid provides medical coverage to >2 million underserved enrollees, approximately half of whom are adults. All paid claims are captured on computerized records, which contain the following information: patient demographics, diagnoses (per International Classification of Diseases, 9th Revision [ICD-9]), procedures (current procedural terminology), medications (national drug codes [NDCs], drug names, doses, quantities, and numbers of days supplied), and payment sources. Information is patient de-identified and fully compliant with the Health Insurance Portability and Accountability Act privacy rule. Accordingly, this study was exempt from review by an institutional review board.

Definition of terms

Mental disorder diagnoses. The presence of OCD was identified by ICD-9 300.3; depression by ICD-9 296.2, 296.3, 296.9, 300.4, 309.0, 309.1, or 311; bipolar disorder by ICD-9 296 or 297; psychosis by ICD-9 295 or 298; organic mental disorder by ICD-9 290, 291, 292, 293, or 294; pervasive developmental disorder (PDD) by ICD-9 299; nonpsychotic brain damage by ICD-9 310; development delay by ICD-9 315; and mental retardation by ICD-9 317, 318, or 319.

Index claim. This is the first claim in which a diagnosis of interest was filed.

Newly diagnosed OCD or depression. A patient was considered newly diagnosed if his or her index claim for OCD or depression was preceded by 2 years in which no such diagnosis was documented.

Pure OCD. Because our goal was to understand the comparative costs of patients with a primary illness OCD or depression, we excluded patients with select comorbid psychiatric diagnoses. We characterized patients as having “pure OCD” (P-OCD) if no claim was filed for depression, bipolar disorder, psychosis, organic mental disorder, PDD, nonpsychotic brain damage, developmental delay, or mental retardation during the 2 years preceding and following their index OCD diagnosis. We excluded depression because it is the targeted comparator. Other diagnoses (see above) were excluded by consensus opinion of the co-authors because: 1) their diagnostic, phenomenologic, and/or neurobiologic overlap with OCD30-36 may have resulted in a misdiagnosis of OCD in some patients; and 2) a preliminary examination of the data revealed that a much larger proportion of OCD cases had comorbid schizophrenia than would be expected from epidemiological studies. This indicated that patients with diagnosed OCD in this Medicaid population are not typical of either OCD subjects in the community, or those diagnosed in private insurance settings.

Pure depression. “Pure depression” (P-D) was defined similarly to P-OCD, except that the index diagnosis was depression and the exclusion diagnoses included OCD rather than depression.

Mental disorder or medical illness resource use. Mental disorder resource use was identified by ICD-9 290–319. All other ICD-9 diagnoses were attributed to medical illness resource use.

Psychotropic or nonpsychotropic medications. The NDC for each pharmacy claim was examined and classified according to drug class. Amphetamines, antidepressants, antimanics, antipsychotics, anxiolytics, hypnotics, mood stabilizers, and stimulants were assumed to be prescribed for the treatment of mental disorders and were therefore classified as psychotropics. All other medications were assumed to be prescribed for treatment of medical illness and were classified as nonpsychotropics.

Overall health care costs. These include costs for inpatient, outpatient, and pharmacy services.

Sample Identification

Florida Medicaid-enrolled adults (age ≥18) who had ≥1 paid claim for OCD or depression from July 1997 through June 2006 were identified. Practice patterns for the treatment of OCD were likely highly variable prior to dissemination of the first national OCD clinical guidelines in 199737; thus, 1997 was selected as the first year of study. From this pool of patients, we selected those who had at least 2 years of data prior to and following their index OCD or depression claim. The 2-year time frame preceding the index diagnosis was set a priori to ensure that 1) only newly diagnosed patients were captured, and 2) patients with serious comorbid psychiatric conditions were excluded. Among this subset, patients with P-OCD or P-D were included in the analyses.

Data analyses

Data from July 1997 through June 2006 were provided by the Florida Medicaid Program in 27 compressed text format files. These were decompressed and imported for analysis using SAS/STAT statistical software version 7 (2006; SAS Institute Inc, Cary, NC).

Because the distributions of sex and race in the overall Medicaid dataset were unbalanced, we adjusted for these imbalances by weighting the distribution of sex and race/ethnicity in our samples. To address Question 1, we conducted χ2 tests to determine whether P-OCD and P-D patients significantly differed on sex, race/ethnicity, age at first diagnosis, and medical illness severity (Charlson Comorbidity Index).38 When a χ2 test was statistically significant (P < .05, 2-tailed), we used logistic regression to compare the likelihood of being diagnosed with P-OCD or P-D for each subgroup.

To address Question 2, each patient with P-OCD was matched to at least 1 patient with P-D on observed covariates (and year of index diagnosis to control for potential differences in treatment approaches over time). If there was no P-D match available, the P-OCD patient was excluded from further analysis. Propensity score-matching methods were not necessary due to the relatively small number of variables selected for matching.

As is commonly seen when analyzing data on health care utilization and cost, data were not normally distributed, but highly skewed to the right (ie, a small number of subjects had extremely high resource use and costs)39; therefore, we conducted Wilcoxon signed-rank tests to compare the median health resource use and costs between P-OCD and P-D groups during the 2 years following each patient’s index diagnosis. We compared total, inpatient, outpatient, and pharmacy health care claims and costs between the 2 groups. In addition, based on the patient’s primary diagnosis for each claim, we further categorized claims and costs as either medical or psychiatric, and examined group differences within these subcategories. Arithmetic means for health care utilization and costs are valuable in health care policy decision making and, thus, were presented. Health care costs for each year were not adjusted for inflation.


Sample characteristics

As shown in the FIGURE, 99 adults were identified as newly diagnosed patients with P-OCD and 14,906 as newly diagnosed patients with P-D. TABLE 1 presents the demographic characteristics of patients with P-OCD and P-D. Adjusting for the distribution of male and female adults in the overall Medicaid dataset, women were 73% less likely to be newly diagnosed with P-OCD (odds ratio [OR], 0.27; 95% confidence interval [CI], 0.19 to 0.38; P < .0001) and 3.7 times more likely to be newly diagnosed with P-D (OR, 3.73; 95% CI, 2.60 to 5.34; P < .0001), than men. Adjusting for the distribution of race/ethnicity in the overall Medicaid dataset, African American patients were approximately half as likely to be newly diagnosed with P-OCD than white patients (OR, 0.49; 95% CI, 0.27 to 0.90; P = .02), and twice as likely to receive a new diagnosis of P-D than whites (OR, 2.03; 95% CI, 1.10 to 3.72; P = .02) Patients age ≥45 were 2 times more likely to receive a new diagnosis of P-D than P-OCD (OR, 2.08; 95% CI, 1.27 to 3.39; P = .004).

FIGURE : Sample selection

aExclusionary diagnoses: OCD = ICD-9 300.3; depression = ICD-9 296.2, 296.3, 296.9, 300.4, 309.0, 309.1, 311; psychoses = ICD-9 295, 298; bipolar disorder = ICD-9 296 297; organic mental disorders = ICD-9 290, 291, 292, 293, 294; pervasive developmental disorder = ICD-9 299; nonpsychotic brain damage = ICD-9 310; development delays = ICD-9 315; mental retardation = ICD-9 317, 318, 319.
ICD: International Classification of Diseases, 9th revision; OCD: obsessive-compulsive disorder.


Demographic characteristics of adults with pure OCD and pure depression

Characteristic P-OCD (N = 99) P-D (N = 14,906) P value
Sex, N (%) <.0001
Male 41 (41.4%) 2,356 (15.8%)
Female 58 (58.6%) 12,550 (84.2%)
Weighted sex, N (%) <.0001
Male 63 (63.6%) 4,755 (31.9%)
Female 36 (36.4%) 10,151 (68.1%)
Race/ethnicity, N (%) .08
White 66 (66.7%) 8,097 (54.3%)
African American 14 (14.1%) 3,486 (23.4%)
Hispanic 8 (8.1%) 1,448 (9.7%)
Other 11 (11.1%) 1,875 (12.6%)
Weighted race/ethnicity, N (%) .08
White 56 (56.8%) 6,603 (44.3%)
African American 14 (13.7%) 3,220 (21.6%)
Hispanic 20 (20.0%) 3,473 (23.3%)
Other 9 (9.5%) 1,610 (10.8%)
Age at index diagnosis, N (%) .004
18 to 29 years 32 (32.3%) 4,850 (32.5%)
30 to 44 years 47 (47.5%) 4,920 (33.0%)
45 to 64 years 17 (17.2%) 3,629 (24.4%)
Mean (SD) 37.1 (13.5) 40.5 (17.0) .01
Medical illness severity, N (%) <.0001
0 88 (88.9%) 10,107 (67.8%)
≥1 11 (11.1%) 4,799 (32.2%)
Mean (SD) 0.2 (0.9) 0.6 (1.2) <.0001
OCD: obsessive-compulsive disorder; P-D: pure depression; P-OCD: pure OCD; SD: standard deviation.

Health care resource use and costs. Patients in the 2 groups were matched by sex, race/ethnicity, exact age at index diagnosis, year of index diagnosis, and medical illness severity (Charlson Comorbidity Index). Of the 99 patients newly diagnosed with P-OCD, 85 were matched to 14,906 newly diagnosed patients with P-D (14 patients in the P-OCD group could not be matched). The number of P-D patients matched to each P-OCD patient ranged from 1 to 75. Two-year, median, per-patient health care use and costs for the matched groups are shown in TABLE 2.


Two-year, per-patient health care utilization and costs for newly diagnosed patients with pure OCD vs pure depression

Type of Claim Patients with P-OCD Patients with P-D P valuea
N 2-year per-patient claims, mean (SD) 2-year median per-patient claims N 2-year per-patient claims, mean (SD) 2-year median per-patient claims
Total number inpatient stays 5 1.0 (0) 1.0 18 1.5 (0.5) 1.5 .12
Medical 5 1.0 (0) 1.0 18 1.5 (0.5) 1.5 .12
Psychiatric 0 0 0
Total number outpatient visits 85 15.9 (20.5) 7.0 963 19.3 (11.7) 17.3 .004
Medical 65 12.0 (15.6) 5.0 711 18.0 (12.1) 16.4 .0002
Psychiatric 79 7.3 (16.1) 2.0 822 3.5 (2.6) 2.9 .39
Total number pharmacy fills 73 75.3 (64.3) 61.0 765 53.9 (38.9) 42.3 .01
Medical 71 45.9 (48.1) 32.0 723 41.1 (31.5) 32.0 .93
Psychiatric 61 35.1 (26.8) 29.0 581 17.8 (14.3) 14.1 <.0001
Cost category Patients with P-OCD Patients with P-D P valuea
N 2-year per-patient costs, mean (SD) 2-year median per-patient costs N 2-year per-patient costs, mean (SD) 2-year median per-patient costs
Total health care costsb 85 $22,031 ($71,486) $6,588 963 $9,472 ($14,839) $5,347 .27
Medical 74 $20,397 ($73,955) $3,652 793 $8,963 ($15,077) $4,573 .12
Psychiatric 82 $4,415 ($6,444) $2,028 892 $1,127 ($1,052) $759 <.0001
Total inpatient costs 5 $5,453 ($2,489) $4,942 18 $5,351 ($4,765) $3,632 1.00
Medical 5 $5,453 ($2,489) $4,942 18 $5,351 ($4,765) $3,632 1.00
Psychiatric 0 0
Total outpatient costs 85 $1,815 ($3,183) $495 963 $2,315 ($1,793) $1,888 .003
Medical 65 $1,414 ($1,975) $363 711 $2,245 ($1,800) $1,928 .003
Psychiatric 79 $770 ($2,866) $139 822 $278 ($341) $219 .76
Total pharmacy costs 73 $7,604 ($10,132) $4,307 765 $3,132 ($2,820) $2,317 .0006
Medical 71 $3,622 ($7,677) $2,059 723 $2,326 ($2,401) $1,623 .67
Psychiatric 61 $4,716 ($6,555) $2,921 581 $1,090 ($1,133) $682 <.0001
aP values compare 2-year median claims using Wilcoxon signed-rank test or 2-year log-transformed mean costs using paired t-tests.
bExcludes costs in the home and community service category.
OCD: obsessive-compulsive disorder; P-D: pure depression; P-OCD: pure OCD; SD: standard deviation.

Overall health care costs. Although there were no significant differences between groups with regard to 2-year, median, per-patient overall health care costs, patients with P-OCD incurred nearly 3-times greater total costs for treatment of psychiatric disorders than did patients with P-D ($2,028 vs $759, P < .0001).

Inpatient use and costs. The 2-year, median number of medical inpatient stays was 1.0 for the P-OCD group compared with 1.5 for the P-D group (P = .12). Inpatient costs were similar between groups. Because no patient with P-OCD had an inpatient stay for treatment of mental disorder, we did not conduct subanalyses of utilization and costs between groups.

Outpatient use and costs. Compared with patients with P-OCD, those with P-D had significantly greater 2-year, median, per-patient total number of outpatient visits (17.3 vs 7.0, P = .004) and costs ($1,888 vs $495, P = .003). Patients with P-D incurred significantly greater number of outpatient visits for treatment of medical illness (16.4 vs 5.0, P = .0002) at significantly higher costs ($1,928 vs $363, P = .003) than did their P-OCD counterparts. There were no significant differences between groups with respect to number of outpatient visits or costs for treatment of mental disorders.

Pharmacy claims and costs. Patients with P-OCD had significantly greater 2-year, median, per-patient total number of pharmacy claims (61.0 vs 42.3, P = .01) and costs ($4,307 vs $2,317, P = .0006) than those with P-D. There were no significant differences between groups with respect to number or costs for nonpsychotropics. However, patients with P-OCD filled significantly more claims for psychotropics (29.0 vs 14.1, P < .0001) at significantly higher costs ($2,921 vs $682, P < .0001) than patients with P-D. TABLE 3 shows the 2-year, median, per-patient fills and costs for classes of psychotropics and specific drugs by group. P-OCD patients had significantly higher 2-year, median, per-patient fills and costs of antidepressants, SSRIs, second-generation antipsychotics, and anxiolytics/sedatives/hypnotics compared with P-D patients.


Two-year median psychotropic medication fills and costs

Psychiatric medication Patients with P-OCD Patients with P-D
N 2-year median per-patient fills 2-year median per-patient cost N 2-year median per-patient fills 2-year median per-patient cost
Antidepressants 55 16.0 $1,355 511 4.0a $234a
  SSRIs 48 10.0 $1,098 419 3.0a $219a
    Citalopram 7 3.0 $193 74 2.0 $123
    Escitalopram 1 3.0 $191 36 1.0 $70
    Fluoxetine 9 20.0 $1,607 89 3.0b $160b
    Fluvoxamine 14 16.5 $1,978 2 6.0 $636
    Paroxetine 17 316.0 $316 167 220.0 $220
    Sertraline 12 4.5 $316 166 1.5 $110c
  Other antidepressants 24 3.0 $191 258 2.0 $110
    Amitriptyline 4 6.5 $33 55 2.0 $13
    Bupropion 5 12.0 $1,289 104 2.0 $167
    Mirtazapine 3 1.5 $111 34 2.0 $146
    Nefazodone 3 2.0 $152 30 1.0 $84
    Nortriptyline 1 1.0 $12 9 2.0 $17
    Trazodone 8 3.0 $39 55 2.5 $13
    Venlafaxine 6 3.5 $480 59 2.0 $160
Antipsychotics 27 19.0 $1,514 62 3.0b $363c
  SGAs 22 18.5 $2,147 59 3.0c $367c
    Aripiprazole 1 12.0 $3,330 3 4.0 $1,148
    Olanzapine 2 24.0 $3,670 27 2.0b $322
    Quetiapine 9 5.0 $556 22 2.5 $159
    Risperidone 13 8.0 $761 16 5.5 $481
    Ziprasidone 2 11.0 $1,817 1 3.0 $701
  FGAs 5 13.0 $131 4 2.0 $48
    Prochlorperazine 1 3.0 $56 3 1.0 $25
    Thioridazine 5 13.0 $131 1 7.0 $70
Anxiolytics, sedatives, or hypnotics 40 8.0 $167 386 4.0c $53b
    Alprazolam 11 17.0 $140 168 3.5c $27b
    Buspirone 1 6.0 $529 15 1.8 $78
    Chlordiazepoxide 1 2.0 $12 3 1.0 $13
    Clonazepam 11 8.0 $100 59 4.0 $54
    Clorazepate 2 1.0 $50 6 1.0 $32
    Diazepam 1 1.0 $5 52 3.0 $16
    Hydroxyzine 3 3.0 $31 48 1.0 $10
    Lorazepam 10 6.5 $177 51 2.0c $53
    Phenobarbital 1 29.0 $172 7 3.0 $15
    Promethazine 5 1.0 $8 114 1.0 $16
    Temazepam 4 7.0 $58 28 2.0 $16
    Zolpidem 8 3.0 $153 58 3.0 $119
Antimanic agents 2 19.5 $330 2 10.5 $169
Mood stabilizers 6 6.5 $573 27 8.0 $584
Amphetamines 3 18.0 $1,176 7 13.0 $504
aP < .001.
bP < .01.
cP < .05.
FGA: first-generation antipsychotic; P-D: pure depression; P-OCD: pure obsessive-compulsive disorder; SGA: second-generation antipsychotic; SSRI: selective serotonin reuptake inhibitor.


To our knowledge, this is the first large-scale, retrospective claims analysis that compares the relative per-patient health care burden of OCD to that of depression. Our findings suggest that a typical patient with adult-onset OCD incurs similar total health care expenditures to a typical patient with adult-onset depression during the 2 years following initial diagnosis. Closer examination revealed that the types of services and costs that contribute to the overall health care burden of depression and OCD differ. Whereas a typical depressed patient had significantly greater use and costs of medical outpatient services, the average patient with adult-onset OCD had significantly greater use and higher costs of psychotropic medications. In addition, it is likely that the discrepancy in psychiatric health care costs between an average patient with adult-onset OCD vs depression will be even larger in the long term (decades), because OCD is a chronic disorder whereas major depression is episodic in about 90% of cases.21 Individuals with depression have a 2- to 3-times greater likelihood of lifetime hospitalization for their condition than those with OCD,40,41 which may bring these long-term costs closer together on average. However, the vast majority of depressed persons will never be hospitalized for their depression, and so the impact of this differential utilization on health care expenditures may be relatively small.

Similar to an earlier study of a large health maintenance organization,42 we observed a far lower prevalence rate for clinically recognized OCD (about 0.01% over 9 years) than the 1-month (0.6%)43 and 1-year (1.6%)18 prevalence rates reported from large community prevalence studies. This finding confirms that OCD is associated with low treatment-seeking and/or recognition rates. In addition, the small proportion of patients diagnosed with OCD after age 45 is consistent with the earlier mean age of onset of OCD and the rarity of OCD onset after age 40,44 in contrast to the pattern in depression.

There are several possible explanations for the significantly higher use and costs of psychotropics among patients with adult-onset OCD compared with those with adult-onset depression. First, patients diagnosed with OCD may require a higher dose and longer duration of SSRI treatment to determine effectiveness than patients with depression.45 Second, remission following OCD treatment is rare29; substantial improvement, defined as a decrease in symptom severity of 25% to 35%, occurs much less often in OCD (40% to 55%)29 than in major depression (70%),46 where response is usually defined as a ≥50% decrease in symptom severity.47

There are relatively few differences between the prescribing practices of primary care physicians (PCPs) and psychiatrists who treat patients with anxiety disorders and depression. The Primary Care Anxiety Project, a longitudinal study of individuals with anxiety disorders in primary care settings, reported no differences in rates or mean doses of SSRIs prescribed by PCPs and psychiatrists.48 A survey of US psychiatrists found that only 39% of patients with OCD received clomipramine and SSRIs at a clinically effective dose,49 a rate similar to that reported in general medical settings.50 An analysis of prescribing practices reported by 180 psychiatrists and 813 PCPs revealed that specialists tended to prescribe slightly higher doses of antidepressants for patients with depression than generalists, although doses prescribed by both PCPs and psychiatrists were within the recommended range.51

As expected, patients were diagnosed with adult-onset OCD at a significantly younger age than those who were diagnosed with adult-onset depression; the age of onset of OCD symptoms is typically 14.5 years, although a correct diagnosis is not received until an average age of 30.4 years.52 Approximately 40% of OCD-diagnosed patients were male, consistent with the proportion found in the 5-site Epidemiological Catchment Area (ECA) study,18 and the significantly greater proportion of females vs males diagnosed with depression (P < .0001) reflects similar findings in population-based studies.53 African Americans were half as likely to be newly diagnosed with P-OCD than white patients, which is consistent with data from the earlier ECA study.18 However, this statistic contrasts with the results of the 2001 to 2003 National Survey of American Life, a population-based mental health survey of US African Americans and non-Hispanic whites, which found that African Americans had a 12-month and lifetime prevalence of OCD similar to that found in the overall sample of the National Comorbidity Study Replication.54 The lower rate of OCD found among African Americans in the current study compared with the National Survey of American Life may reflect the use of different diagnostic methodology (diagnosis based upon judgment of a specialist or generalist physician vs a lay-administered structured psychiatric interview), the different settings patients were selected from (general population vs Medicaid enrollees), and other factors.

There are several study limitations that merit discussion. First, the total health care burden associated with P-OCD and P-D is likely to be substantially larger than our figures suggest, given that this analysis was limited to patients with clinically recognized mental disorders. Second, due to the lack of a control group, we cannot estimate the excess health care burden of OCD compared with persons without a mental disorder. Third, study patients were Medicaid enrollees, who may have had limited access to treatment; therefore, findings cannot be generalized to patient populations with other types of health insurance. Fourth, excluding patients with childhood-onset OCD, which appears to be associated with greater symptom severity and a poorer response to treatment than adult-onset OCD,55 may have underestimated costs compared with the broader OCD population. Fifth, because patients with OCD and select comorbidities were excluded to ensure that outcomes were based on patients with a primary diagnosis of OCD, our sample is not representative of all OCD patients, and results likely underestimated the true costs for OCD patients. Persons with OCD and comorbid major depression, representing about one-third of OCD patients,56 are less responsive to treatment than those without depression,57,58 and are therefore likely to incur higher health care costs. Among the approximately 12% of OCD patients with comorbid psychosis,18 OCD symptoms are associated with a worsening psychosis prognosis, and these patients are more difficult to treat than those free from psychosis,59 increasing the costs of care. Finally, this study was not intended to determine the cost offsets of effective treatment for OCD, and therefore the extent to which more effective treatment would reduce “unnecessary” or “excess” health care costs remains to be determined.

Although the health care burden associated with anxiety disorders has received scant attention compared with depression,1-3 a greater awareness of the functional impairment, disability, and health care utilization associated with anxiety disorders is changing this trend.1,3 Among anxiety disorders, OCD often is overlooked because of its relatively low prevalence compared with generalized anxiety disorder, social anxiety disorder, and simple phobia.60 Indeed, given that the 1-year prevalence of depression is about 7 times higher than that of OCD (6.7% vs 1%),19 we would expect annual health care costs associated with depression to far exceed those of OCD. Nonetheless, the burden of OCD to the health care system and society is of great concern considering the typical early onset of the disorder, its chronic nature, and its significantly heightened risk for suicide,61 unemployment,62 poor interpersonal relationships,63 and decreased quality of life.64


In summary, we found that the overall per-patient health care burden of adult-onset OCD is comparable to that of adult-onset depression, but that patient groups differed with regard to the components of burden. Whereas patients with P-D incurred significantly greater outpatient medical treatment at higher costs, those with P-OCD filled significantly more claims for psychotropics at significantly higher costs. Some evidence indicates that cognitive-behavioral therapy yields longer lasting benefits than pharmacotherapy alone in the treatment of OCD,65,66 and may therefore allow OCD patients to successfully discontinue chronic psychotropics and reduce medication costs in the long term. Given that individual patients with OCD are likely to experience high treatment costs (particularly for medications) for this chronic disorder, the health care burden associated with OCD may take a large toll over time. Greater awareness of the clinical and health care systems burden associated with OCD may increase efforts of PCPs to more effectively diagnose and treat this relatively rare but often debilitating and costly mental illness. It will also compel payers to appropriately reimburse PCPs for the time required to properly monitor patient response to treatment, and hopefully improve the lives of those affected by this disorder.

ACKNOWLEDGEMENTS: Funding for this study was provided by Jazz Pharmaceuticals, Inc.

DISCLOSURES: Drs. Hankin and Bronstone, and Ms. Wang have served as consultants to Jazz Pharmaceuticals, Inc. Dr. Koran has received grant/research support from Eli Lilly and Company and Jazz Pharmaceutical; and has served on the speaker’s bureau for Forest Laboratories. Dr. Sheehan has served as a consultant to Abbott Laboratories, Alexa Pharmaceuticals, Alza Pharmaceuticals, American Medical Association, American Psychiatric Association, Angelini Labopharm, Anxiety Disorders Resource Center, AstraZeneca, Avera Pharmaceuticals, Bristol-Myers Squibb, Cephalon, Connecticut and Ohio Academies of Family Physicians, Cortex Pharmaceuticals, Council on Anxiety Disorders, CPC Coliseum Medical Center, Cypress Bioscience, Eisai, Inc., Faxmed, Inc., Forest Laboratories, GlaxoSmithKline, INC Research, International Society for CNS Drug Development, Janssen Pharmaceuticals, Jazz Pharmaceuticals, Layton Bioscience, Lundbeck, MediciNova, National Anxiety Awareness Program, National Anxiety Foundation, National Depressive and Manic Depressive Association, National Institutes of Health, Neuronetics, NovaDel, Organon, Orion Pharma, Parexel International Corporation, Pfizer, Pharmacia, Pharmacia and Upjohn, PharmaNeuroBoost, Philadelphia College of Pharmacy and Science, Pierre Fabre, Roche, Sagene, sanofi-aventis, Sanofi-Synthelabo Recherche, Sepracor, Shire Laboratories, Inc., Solvay Pharmaceuticals, Takeda Pharmaceuticals, Tampa General Hospital, Targacept, Therapeutics, The Upjohn Company, Tikvah Titan Pharmaceuticals, U.S. Congress, U.S. Food and Drug Administration, University of South Florida Friends of Research in Psychiatry, World Health Organization, Wyeth-Ayerst, XCENDA, and ZARS; has received grant/research support from Abbott Laboratories, American Medical Association, Anclote Foundation, Angelini Labopharm, AstraZeneca, Avera Pharmaceuticals, Bristol-Myers Squibb, Burroughs Wellcome, Eisai, Inc., Eli Lilly and Company, Forest Laboratories, Glaxo-Wellcome, GlaxoSmithKline, International Clinical Research, Janssen Pharmaceuticals, Jazz Pharmaceuticals, Kali-Duphar, Mead Johnson, MediciNova, Merck Sharp and Dohme, National Institutes of Health, Novartis Pharmaceuticals Corporation, Parke-Davis, Pfizer, Quintiles, Sandoz Pharmaceuticals, sanofi-aventis, Sanofi-Synthelabo Recherche, SmithKlineBeecham, The Upjohn Company, TAP Pharmaceuticals, University of South Florida College of Medicine , United Bioscience, Warner Chilcot, Worldwide Clinical Trials, Wyeth-Ayerst, and Zeneca Pharmaceuticals; has served as a speaker/lecturer for Abbott Laboratories, Angelini Labopharm, AstraZeneca, Boehringer Ingelheim, Boots Pharmaceuticals, Bristol-Myers Squibb, Burroughs Wellcome, Charter Hospitals, Ciba Geigy, Dista Products Company, Eli Lilly and Company, Excerpta Medica Asia, Glaxo Pharmaceuticals, GlaxoSmithKline, Hikma Pharmaceuticals, Hospital Corporation of America, Humana, ICI, INC Research, Janssen Pharmaceuticals, Kali-Duphar, Marion Merrill Dow, McNeil Pharmaceuticals, Mead Johnson, Merck Sharp and Dohme, Novo Nordisk, Organon, Parke-Davis, Pfizer, Pharmacia and Upjohn, PharmaNeuroBoost, Rhone Laboratories, Rhone-Poulenc Rorer Pharmaceuticals, Roerig, Sandoz Pharmaceuticals, sanofi-aventis, Schering Corporation, SmithKlineBeecham, Solvay Pharmaceuticals, The Upjohn Company, TAP Pharmaceuticals, United Bioscience, University of South Florida College of Medicine, Warner Chilcott, Wyeth-Ayerst; and owns stock in Medical Outcomes Systems. Dr. Hollander has served as a consultant to Transcept Pharmaceuticals. Dr. Culpepper has served as a consultant to AstraZeneca, Eli Lilly and Company, Labopharm, Merck, Pfizer Inc., sanofi-aventis, Takeda, Trovis, and Wyeth Pharmaceuticals; and has served on the speaker’s bureau for Merck, Pfizer Inc., and Wyeth Pharmaceuticals. Dr. Black has received grant/research support from AstraZeneca and Psyadon Pharmaceuticals, Inc. Dr. Knispel has served as a consultant to Hologic, Health Strategies, Jazz Pharmaceuticals, PharmaStrat, and Transcept Pharmaceuticals. Dr. Dunn has served as a consultant to AstraZeneca, Endo Pharmaceuticals, Jazz Pharmaceuticals, J & J Pharmaceuticals, Takeda, and Teva Pharmaceuticals. Dr. Dougherty has served as a consultant to and received honoraria from Brand Ideas, Eli Lilly and Company, Medtronic, Ortho-McNeil Inc., and Reed Elsevier; and has received grant/research support from Cyberonics.


  1. Stein MB, Roy-Byrne PP, Craske MG, et al. Functional impact and health utility of anxiety disorders in primary care outpatients. Med Care. 2005;43:1164–1170.
  2. Kroenke K, Spitzer RL, Williams JB, et al. Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection. Ann Intern Med. 2007; 146:317–325.
  3. Katon W, Roy-Byrne P. Anxiety disorders: efficient screening is the first step in improving outcomes. Ann Intern Med. 2007;146:390–392.
  4. Luber MP, Meyers BS, Williams-Russo PG, et al. Depression and service utilization in elderly primary care patients. Am J Geriatr Psychiatry. 2001;9:169–176.
  5. Pearson SD, Katzelnick DJ, Simon GE, et al. Depression among high utilizers of medical care. J Gen Intern Med. 1999;14:461–468.
  6. Regier DA, Hirschfeld RM, Goodwin FK, et al. The NIMH Depression Awareness, Recognition, and Treatment Program: structure, aims, and scientific basis. Am J Psychiatry. 1988;145:1351–1357.
  7. Katon W, Berg AO, Robins AJ, et al. Depression—medical utilization and somatization. West J Med. 1986; 144:564–568.
  8. Unutzer J, Patrick DL, Simon G, et al. Depressive symptoms and the cost of health services in HMO patients aged 65 years and older. A 4-year prospective study.  JAMA. 1997;277:1618–1623.
  9. Callahan CM, Hui SL, Nienaber NA, et al. Longitudinal study of depression and health services use among elderly primary care patients. J Am Geriatr Soc. 1994;42:833–838.
  10. Katon WJ, Lin E, Russo J, et al. Increased medical costs of a population-based sample of depressed elderly patients. Arch Gen Psychiatry. 2003;60:897–903.
  11. Henk HJ, Katzelnick DJ, Kobak KA, et al. Medical costs attributed to depression among patients with a history of high medical expenses in a health maintenance organization. Arch Gen Psychiatry. 1996;53:899–904.
  12. Simon GE, VonKorff M, Barlow W. Health care costs of primary care patients with recognized depression. Arch Gen Psychiatry. 1995;52:850–856.
  13. Druss BG, Rosenheck RA, Sledge WH. Health and disability costs of depressive illness in a major U.S. corporation.  Am J Psychiatry. 2000;157:1274–1278.
  14. Harman JS, Rollman BL, Hanusa BH, et al. Physician office visits of adults for anxiety disorders in the United States, 1985-1998. J Gen Intern Med. 2002;17:165–172.
  15. Nisenson LG, Pepper CM, Schwenk TL, et al. The nature and prevalence of anxiety disorders in primary care. Gen Hosp Psychiatry. 1998;20:21–28.
  16. Spitzer RL, Williams JB, Kroenke K, et al. Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study.  JAMA. 1994;272:1749–1756.
  17. Greenberg PE, Sisitsky T, Kessler RC, et al. The economic burden of anxiety disorders in the 1990s. J Clin Psychiatry. 1999;60:427–435.
  18. Karno M, Golding JM, Sorenson SB, et al. The epidemiology of obsessive-compulsive disorder in five US communities. Arch Gen Psychiatry. 1988;45:1094–1099.
  19. Kessler RC, Chiu WT, Demler O, et al. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:617–627.
  20. DuPont RL, Rice DP, Shiraki S, et al. Economic costs of obsessive-compulsive disorder. Med Interface. 1995;8:102–109.
  21.  Diagnostic and statistical manual of mental disorders, 4th ed, text rev.  Washington DC: American Psychiatric Association; 2000.
  22. Koran L, Leventhal J, Fireman B, et al. Recognition and treatment of obsessive-compulsive disorder in a pre-paid health plan: does adequate treatment reduce costs? (Abstract P.3.056). Eur Neuropsychopharmacol. 1997;7:S243.
  23. Kennedy BL, Schwab JJ. Utilization of medical specialists by anxiety disorder patients. Psychosomatics. 1997;38:109–112.
  24. Hatch ML, Paradis C, Friedman S, et al. Obsessive-compulsive disorder in patients with chronic pruritic conditions: case studies and discussion. J Am Acad Dermatol. 1992;26:549–551.
  25. Demet MM, Deveci A, Taskin EO, et al. Obsessive-compulsive disorder in a dermatology outpatient clinic. Gen Hosp Psychiatry. 2005;27:426–430.
  26. Fineberg NA, O’Doherty C, Rajagopal S, et al. How common is obsessive-compulsive disorder in a dermatology outpatient clinic?  J Clin Psychiatry. 2003;64:152–155.
  27. Nymberg JH, Van Noppen B. Obsessive-compulsive disorder: a concealed diagnosis. Am Fam Physician. 1994;49:1129–1137 1142–1144.
  28. Koblenzer CS. Psychiatric syndromes of interest to dermatologists. Int J Dermatol. 1993;32:82–88.
  29. Koran LM, Hanna GL, Hollander E, et al. Practice guideline for the treatment of patients with obsessive-compulsive disorder. Am J Psychiatry. 2007;164(7 suppl):5–53.
  30. Cullen B, Samuels J, Grados M, et al. Social and communication difficulties and obsessive-compulsive disorder. Psychopathology. 2008;41:194–200.
  31. Kobayashi R, Murata T. Behavioral characteristics of 187 young adults with autism. Psychiatry Clin Neurosci. 1998;52:383–390.
  32. Carcani-Rathwell I, Rabe-Hasketh S, Santosh PJ. Repetitive and stereotyped behaviours in pervasive developmental disorders. J Child Psychol Psychiatry. 2006;47:573–581.
  33. Bejerot S. An autistic dimension: a proposed subtype of obsessive-compulsive disorder. Autism. 2007;11:101–110.
  34. Russell AJ, Mataix-Cols D, Anson M, et al. Obsessions and compulsions in Asperger syndrome and high-functioning autism. Br J Psychiatry. 2005;186:525–528.
  35. Tibbo P, Warneke L. Obsessive-compulsive disorder in schizophrenia: epidemiologic and biologic overlap. J Psychiatry Neurosci. 1999;24:15–24.
  36. Bottas A, Cooke RG, Richter MA. Comorbidity and pathophysiology of obsessive-compulsive disorder in schizophrenia: is there evidence for a schizo-obsessive subtype of schizophrenia?  J Psychiatry Neurosci. 2005;30:187–193.
  37. March JS, Frances A, Kahn DA, et al. Expert consensus guidelines: treatment of obsessive-compulsive disorder. J Clin Psychiatry. 1997;58(suppl 4):1–72.
  38. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383.
  39. Thompson SG, Barber JA. How should cost data in pragmatic randomised trials be analysed?  BMJ. 2000;320:1197–1200.
  40. Matza LS, Revicki DA, Davidson JR, et al. Depression with atypical features in the National Comorbidity Survey: classification, description, and consequences. Arch Gen Psychiatry. 2003;60:817–826.
  41. Ruscio AM, Stein DJ, Chiu WT, et al. The epidemiology of obsessive-compulsive disorder in the National Comorbidity Survey Replication. Mol Psychiatry. 2010;15:53–63.
  42. Fireman B, Koran LM, Leventhal JL, et al. The prevalence of clinically recognized obsessive-compulsive disorder in a large health maintenance organization. Am J Psychiatry. 2001;158:1904–1910.
  43. Stein MB, Forde DR, Anderson G, et al. Obsessive-compulsive disorder in the community: an epidemiologic survey with clinical reappraisal. Am J Psychiatry. 1997;154:1120–1126.
  44. Rasmussen SA, Eisen JL. The epidemiology and differential diagnosis of obsessive compulsive disorder. J Clin Psychiatry. 1992;53(suppl):4–10.
  45. Marangell LB, Silver JM, Goff DC, et al. Psychopharmacology and electroconvulsive therapy. In: Hales RE, Yudofsky SC, eds. The American Psychiatric Publishing textbook of clinical psychiatry. 4th ed. Arlington, VA: American Psychiatric Publishing, Inc; 2003:1047–1150.
  46. Kando JC, Wells BG, Hayes PE. Depressive disorders. In: DiPiro JT Talbert RE, Yee RL, et al, eds. Pharmacotherapy: a pathophysiologic approach. 4th ed.  Stamford, CT: Appleton and Lange; 1999:1141–1160.
  47. Cole SA, Christensen JF, Cole MR, et al. Mental and behavioral disorders. Depression. In: Feldman MD, Christensen JF, eds. Behavioral medicine: a guide for clinical practice. 3rd ed.  New York, NY: McGraw-Hill Companies, Inc.; 2008:199–226.
  48. Weisberg RB, Dyck I, Culpepper L, et al. Psychiatric treatment in primary care patients with anxiety disorders: a comparison of care received from primary care providers and psychiatrists. Am J Psychiatry. 2007;164:276–282.
  49. Blanco C, Olfson M, Stein DJ, et al. Treatment of obsessive-compulsive disorder by U.S. psychiatrists.  J Clin Psychiatry. 2006;67:946–951.
  50. Koran LM, Leventhal JL, Fireman B, et al. Pharmacotherapy of obsessive-compulsive disorder in a health maintenance organization. Am J Health Syst Pharm. 2000;57:1972–1978.
  51. Loosbrock DL, Tomlin ME, Robinson RL, et al. Appropriateness of prescribing practices for serotonergic antidepressants. Psychiatr Serv. 2002;53:179–184.
  52. Hollander E, Stein DJ, Kwon JH, et al. Psychosocial function and economic costs of obsessive-compulsive disorder. CNS Spectr. 1997;2:16–25.
  53. Eaton WW, Shao H, Nestadt G, et al. Population-based study of first onset and chronicity in major depressive disorder. Arch Gen Psychiatry. 2008;65:513–520.
  54. Himle JA, Muroff JR, Taylor RJ, et al. Obsessive-compulsive disorder among African Americans and blacks of Caribbean descent: results from the National Survey of American Life. Depress Anxiety. 2008;25:993–1005.
  55. Rosario-Campos MC, Leckman JF, Mercadante MT, et al. Adults with early-onset obsessive-compulsive disorder. Am J Psychiatry. 2001;158:1899–1903.
  56. Rasmussen SA, Eisen JL. Clinical and epidemiologic findings of significance to neuropharmacologic trials in OCD. Psychopharmacol Bull. 1988;24:466–470.
  57. Abramowitz JS, Franklin ME, Street G, et al. Effects of comorbid depression on response to treatment for obsessive-compulsive disorder. Behav Ther. 2000; 31:517–528.
  58. Foa EB, Grayson JB, Steketee GS, et al. Success and failure in the behavioral treatment of obsessive-compulsives. J Consult Clin Psychol. 1983;51:287–297.
  59. Poyurovsky M. Exploring OCD subtypes and treatment resistance. Psychiatric Times. 2007;24: September 1, 2007.
  60. Kessler RC, Berglund P, Demler O, et al. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry. 2005;62:593–602.
  61. Murray CJL, Lopez AD. The global burden of disease. Cambridge MA: Harvard University Press; 1996.
  62. Torres AR, Prince MJ, Bebbington PE, et al. Obsessive-compulsive disorder: prevalence, comorbidity, impact, and help-seeking in the British National Psychiatric Morbidity Survey of 2000. Am J Psychiatry. 2006;163:1978–1985.
  63. Eisen JL, Mancebo MA, Pinto A, et al. Impact of obsessive-compulsive disorder on quality of life. Compr Psychiatry. 2006;47:270–275.
  64. Koran LM. Quality of life in obsessive-compulsive disorder. Psychiatr Clin North Am. 2000;23:509–517.
  65. Hembree EA, Riggs DS, Kozak MJ, et al. Long-term efficacy of exposure and ritual prevention therapy and serotonergic medications for obsessive-compulsive disorder. CNS Spectr. 2003;8:363–371,381.
  66. Simpson HB, Liebowitz MR, Foa EB, et al. Post-treatment effects of exposure therapy and clomipramine in obsessive-compulsive disorder. Depress Anxiety. 2004;19:225–233.

CORRESPONDENCE: Cheryl S. Hankin, PhD, President and Chief Scientific Officer, BioMedEcon, LLC, PO Box 129, Moss Beach, CA 94038 USA, E-MAIL: