General Hospital Psychiatry 36 (2014) 192–198

Contents lists available at ScienceDirect

General Hospital Psychiatry journal homepage: http://www.ghpjournal.com

Psychiatry and Primary Care Recent epidemiologic studies have found that most patients with mental illness are seen exclusively in primary care medicine. These patients often present with medically unexplained somatic symptoms and utilize at least twice as many health care visits as controls. There has been an exponential growth in studies in this interface between primary care and psychiatry in the last 10 years. This special section, edited by Jürgen Unutzer, M.D., will publish informative research articles that address primary care-psychiatric issues.

Diagnostic specificity and mental health service utilization among veterans with newly diagnosed anxiety disorders☆ Terri L. Barrera, Ph.D a, b, c,⁎, Juliette M. Mott, Ph.D a, b, c, Natalie E. Hundt, Ph.D a, b, c, Joseph Mignogna, Ph.D a, b, c, Hong-Jen Yu, M.S. a, b, c, Melinda A. Stanley, Ph.D a, b, c, Jeffrey A. Cully, Ph.D a, b, c a b c

Center for Innovations in Quality, Effectiveness and Safety, Michael E. DeBakey VA Medical Center, Houston, TX Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX VA South Central Mental Illness Research, Education and Clinical Center

a r t i c l e

i n f o

Article history: Received 31 July 2013 Revised 10 October 2013 Accepted 11 October 2013 Keywords: Anxiety Diagnostic Veteran Mental health service utilization Anxiety disorder not otherwise specified

a b s t r a c t Objective: This study examined rates of specific anxiety diagnoses (posttraumatic stress disorder, generalized anxiety disorder, panic disorder, obsessive-compulsive disorder, social anxiety disorder, and specific phobia) and anxiety disorder not otherwise specified (anxiety NOS) in a national sample of Veterans and assessed their mental health service utilization. Method: This study used administrative data extracted from Veteran Health Administration outpatient records to identify patients with a new anxiety diagnosis in fiscal year 2010 (N = 292,244). Logistic regression analyses examined associations among diagnostic specificity, diagnostic location, and mental health service utilization. Results: Anxiety NOS was diagnosed in 38% of the sample. Patients in specialty mental health were less likely to receive an anxiety NOS diagnosis than patients in primary care (odds ratio [OR] = 0.36). Patients with a specific anxiety diagnosis were more likely to receive mental health services than those with anxiety NOS (OR = 1.65), as were patients diagnosed in specialty mental health compared with those diagnosed in primary care (OR = 16.29). Conclusion: Veterans diagnosed with anxiety NOS are less likely to access mental health services than those with a specific anxiety diagnosis, suggesting the need for enhanced diagnostic and referral practices, particularly in primary care settings. Published by Elsevier Inc.

Anxiety disorders are the most prevalent psychiatric disorders, with a 12 month prevalence rate of 18.1% in the general population [1]. This rate is nearly doubled in Veteran populations (33%) [2,3]. Anxiety disorders are associated with substantial functional impairment and high rates of comorbid psychiatric and medical disorders, and untreated anxiety can result in overutilization of medical services and increased healthcare costs [4–6].

☆ This research was supported by the Department of Veterans Affairs South Central Mental Illness Research Education and Clinical Center (MIRECC) and the Center for Innovations in Quality, Effectiveness and Safety (#CIN 13-413), Michael E. DeBakey VA Medical Center, Houston, TX. The views expressed reflect those of the authors and not necessarily those of the Department of Veterans Affairs, US government or Baylor College of Medicine. ⁎ Corresponding author. Michael E. DeBakey Veterans Affairs (VA) Medical Center, 152, Houston, TX 77030. Tel.: +1 713 440 4490; fax: +1 713 748 7359. E-mail address: [email protected] (T.L. Barrera). 0163-8343/$ – see front matter. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.genhosppsych.2013.10.013

Unfortunately, anxiety disorders often go unrecognized and untreated, particularly in primary care settings [7]. Only 50% of patients with mental health problems are detected by primary care providers, and even fewer are adequately treated or referred for specialty mental health services [5]. Although much work has been done to improve the detection and treatment of depression in primary care settings, far less attention has been given to anxiety disorders, despite their greater prevalence [8]. The under-recognition of anxiety disorders in primary care is likely a multifaceted problem, due in part to the complex overlap of anxiety symptoms with medical conditions, somatization, and lack of physician knowledge, time, and skill in diagnosing anxiety disorders [9]. This latter issue may also affect the specificity of anxiety diagnoses assigned by primary care physicians when they detect clinically significant anxiety. Providers who have less training in assessing and diagnosing mental health disorders or who face structural barriers, such as time and financial disincentives to conducting full psychiatric assessments [10], may use a nonspecific diagnosis such as Anxiety

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

Disorder Not Otherwise Specified (anxiety NOS) if they are unsure whether a given patient meets diagnostic criteria for a specific anxiety disorder. The diagnosis of anxiety NOS refers to clinically significant symptoms of anxiety or phobic avoidance that do not meet criteria for any of the other Diagnostic & Statistical Manual, Fourth Edition, Text Revision anxiety disorders or when a clinician concludes that an anxiety disorder is present, but has not determined whether it is primary, due to a medical condition, or substance-induced [11]. The prevalence of anxiety NOS in the general population is unknown, as the diagnosis has not been estimated in epidemiological studies. However, anxiety NOS is the most commonly diagnosed anxiety disorder among active-duty service members [12] and primary care patients [13,14]. The high rate of diagnosis of anxiety NOS in primary care may reflect the true incidence of anxiety symptoms that are not classifiable using more specific anxiety disorder diagnoses, but anxiety NOS also is likely used frequently as a provisional diagnosis, with the expectation that the diagnosing physician or other provider will be able to classify the symptoms more accurately at a later date. In either case, a nonspecific diagnosis such as anxiety NOS may be a barrier to receiving mental health services. No specific treatment guidelines for anxiety NOS exist [15], and although general anxiety treatment approaches may be beneficial, the lack of specific treatment recommendations may make referral options for anxiety NOS less clear. Similarly, if anxiety NOS is being used as a provisional diagnosis, this may indicate that the provider does not yet have enough information to determine which mental health services would be beneficial, making it less likely that the provider will initiate referral. The goal of the current study was to determine the rates of specific and nonspecific anxiety diagnoses in a national sample of Veterans receiving care at VA medical centers and examine patterns of mental health service utilization in the year following diagnosis. The VA is the largest healthcare system in the United States, and the availability of VA national databases allows an examination of diagnostic and mental health service utilization patterns in a sizeable sample of patients. We hypothesized that anxiety NOS would be a common diagnosis, particularly in primary care settings where providers tend to have less time and expertise in diagnosing psychiatric disorders relative to specialty mental health settings. To determine whether patient factors are related to diagnostic specificity, we examined the role of patient demographic and clinical characteristics in predicting the diagnosis of specific vs. nonspecific anxiety disorders. We expected anxiety NOS to be used primarily as a placeholder, and therefore expected that most patients initially diagnosed with anxiety NOS would receive a diagnosis of a more specific anxiety disorder within the following 12 months. Finally, because diagnostic specificity may facilitate treatment use, we predicted that patients with specific anxiety diagnoses would be more likely to receive mental health services in the 12 months following diagnosis than patients diagnosed with anxiety NOS. 1. Methods This retrospective database cohort study used patient data from the Veterans Health Administration (VHA) National Patient Care Database (NPCD) outpatient encounter files for Veterans receiving care during the fiscal year 2010 (October 1, 2009, to September 30, 2010). The NPCD contains encrypted patient identifiers that are associated with broad patient and service characteristics and are available for health services research. The accuracy and validity of the NPCD data are monitored by the VHA Veteran Information Resource Center [16]. 1.1. Patient population This study focused on patients who received a new anxiety diagnosis in VHA outpatient facilities during the specified year.

193

Patients were categorized using the International Classification of Diseases, Ninth Edition, Clinical Modification. Anxiety disorder NOS was identified by the ICD-9 codes 300.00 and 300.09, while specific anxiety diagnoses included posttraumatic stress disorder (PTSD; 309.81), generalized anxiety disorder (GAD; 300.02), panic disorder (PD; 300.01 and 300.21), obsessive-compulsive disorder (OCD; 300.3), social anxiety disorder (SAD; 300.23), and specific phobia (SP; 300.29). Restricting the sample to patients with new diagnoses reduced the likelihood of contamination from patients who previously received a nonspecific anxiety diagnosis that was subsequently clarified and provided greater assurance that patients would likely benefit from mental health services. A new diagnosis was defined as one occurring after a 6-month period without a related diagnosis before the index date (the date of first diagnosis during the study period). No patients received both a new anxiety NOS diagnosis and a specific anxiety diagnosis on their index date. However, several patients received multiple new-onset specific anxiety diagnoses on their index date and were therefore classified into more than one specific anxiety diagnostic category (e.g., a patient with a new PTSD diagnosis and a new GAD diagnosis would be classified into both categories for descriptive purposes). We excluded patients with 60 or more inpatient hospital days in the 180days following the index date to limit the sample to patients with adequate opportunities to use outpatient mental health services. Thus, the final sample included 292,244 patients with a new anxiety diagnosis (180,646 diagnosed with a specific anxiety disorder and 111,598 diagnosed with anxiety NOS). 1.2. Diagnostic care settings Within the VHA, specific identifiers known as stop codes are used to indicate the primary work group responsible for each patient encounter. We used VA clinic stop codes to identify diagnostic patterns across the following care settings: primary care, specialty mental health, integrated primary care mental health (PC-MHI), and “other” specialty settings. Primary care visits included stop codes 301, 322, 323, 348 and 350; and PC-MHI included stop codes 531, 534, 539, 571 and 572. Specialty mental health care was defined by VA clinic stop codes 500–599, excluding the above-mentioned PC-MHI stop codes. Finally, anxiety diagnoses associated with any other VA clinic stop codes were categorized as “other” specialty settings, and most frequently included laboratory and emergency care services. 1.3. Mental health service utilization Patient use of mental health services was assessed during the 12 months following each patient’s index date. The full spectrum of mental health Current Procedural Terminology (CPT) codes (90801– 90911, 96100–96155) were used to assess mental health service use. Psychotherapy use was classified using the following CPT codes for fiscal year 2010: 90804, 90806, 90808, 90810, 90812, 90814, 90845, 90846, 90847, 90849, 90853, 90857, 90875, 90876, 96152, 96153, 96154, 96155). Psychiatric medication visits (90862) and psychotherapy with medication-management visits (90805, 90807, 90809, 90811, 90813, 90815) were also examined. Descriptive statistics related to mental health service visits were calculated for the 12 months following each patient’s index date. 1.4. Patient characteristics We examined patient sociodemographic characteristics, including age, gender, marital status, income (estimated using the average adjusted gross income for each patient zip code, based on 2008 Internal Revenue Service data), and distance in miles to the nearest VA facility (calculated using patient zip code). Patient race was not

194

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

consistently populated in these administrative data files and, therefore, was not included. Because of the increased access to care provided to Veterans with service-connected disabilities, patients were categorized into dichotomous disability groups (service-connected disability vs. no service-connected disability). Illness burden was assessed using a diagnosis-based risk-adjustment methodology (DxCG Company, Boston) [17], validated in the VA population [18], that provides a Veteran’s Relative Risk Score. Relative risk is the total predicted healthcare cost of an individual, compared with the average predicted cost of the population, with a score of 1.0 reflecting average risk [17,18]. 1.5. Analyses Analyses were conducted using SAS version 9.2 (SAS Institute, Inc., Cary, NC). Chi-square and analysis of variance (ANOVA) were used to test the relationship between patient demographic and clinical variables and specificity of anxiety diagnosis (anxiety NOS vs. specific anxiety disorder). Descriptive analyses examined the diagnostic rates of each anxiety disorder across care settings and rates of mental health service utilization during the 12 months following patients’ index date. The relative proportion of patients receiving specific anxiety diagnoses vs. anxiety NOS in primary care and specialty mental health settings was tested using chi square. To determine whether anxiety disorder NOS may be used as a “placeholder,” subsequent analyses examined the rate at which patients with an initial diagnosis of anxiety disorder NOS went on to receive a later diagnosis of a specific anxiety disorder. Logistic regression was used to ascertain demographic and clinical variables significantly associated with specificity of anxiety diagnosis (anxiety NOS vs. specific anxiety disorder). All variables that significantly discriminated diagnostic specificity in the prior chi square and ANOVA analyses were included. Demographic variables included dichotomous variables for gender (male vs. female), marital status (married vs. not married), and service-connected disability (service-connected disability vs. no service-connected disability), as well as continuous variables for age, income, and relative-risk score. Clinical variables included dichotomous presence of comorbid mental health diagnoses (depression, substance use, schizophrenia, cognitive disorder, and bipolar disorder), and care setting associated with initial anxiety diagnosis. Predictor variables were entered using a block method involving entry of all variables into the model at the same time to account for the relationships among potential predictor variables. To test whether diagnostic specificity was related to subsequent receipt of treatment, a second logistic-regression analysis was used to ascertain whether demographic and clinical variables, including diagnostic specificity, were significantly associated with receipt of mental health services. Mental health service use was defined as having any mental health CPT code during the 12 months following each patient’s index date. Demographic variables were the same as described above. Clinical variables included type of anxiety diagnosis (anxiety NOS vs. specific anxiety disorder), care setting of initial anxiety diagnosis (primary care, specialty mental health, etc.), and comorbid mental health diagnoses (depression, substance use, schizophrenia, cognitive disorder, and bipolar disorder). Predictor variables were entered using a block method involving entry of all variables into the model at the same time to account for the relationships among potential predictor variables.

Chi-square and ANOVA tests indicated that all variables examined were significantly associated with diagnostic specificity. Patients diagnosed with anxiety NOS were more likely to be older, female, and unmarried; to have a lower income, and lower relative-risk scores; and to have a diagnosis of a comorbid cognitive disorder. Patients diagnosed with a specific anxiety disorder were more likely to live closer to a VA facility; to have a service-connected disability; and to have a diagnosis of comorbid psychiatric disorders, including depression, substance use, schizophrenia, and bipolar disorder. As shown in Table 2, 62% of patients received a specific anxiety diagnosis and 38% received a diagnosis of anxiety NOS. PTSD was the most common specific anxiety disorder diagnosis (55% of the total sample), followed by GAD (8%) and PD (5%). The majority of patients were diagnosed with anxiety conditions in primary care or specialty mental health settings, with PC-MHI settings and other nonmental health specialty clinics diagnosing substantially fewer anxiety disorders. Notably, anxiety NOS was diagnosed much more frequently in primary care (67% of anxiety NOS diagnoses) than in specialty mental health settings (18% of anxiety NOS diagnoses), χ 2(1, N= 254056) = 14312.49, Pb.0001. 2.2. Mental health service utilization Rates of mental health service utilization, including both psychotherapy and medication management visits, in the 12 months following diagnosis are presented in Table 3. Most patients with a specific anxiety diagnosis received mental health services, with treatment rates for patients with the most frequently diagnosed specific anxiety disorders (PTSD, GAD, and PD) ranging between 60 and 67%. In contrast, only 32% of patients with anxiety NOS received mental health services following their index date. These patterns of service utilization were similar across psychotherapy, medication management, and combined medication and psychotherapy visits. 2.3. Anxiety NOS follow-up We examined the 111,598 patients with an initial diagnosis of anxiety NOS to determine how many received a diagnosis of a specific anxiety disorder within the following 12 months. Contrary to our hypothesis, this study observed that only 12% went on to receive a diagnosis of a specific anxiety disorder in the following year. These subsequent diagnostic patterns mirrored those of the full sample, with PTSD, GAD, and PD being the most frequently diagnosed specific anxiety conditions in this group. Most anxiety NOS patients who later received a diagnosis of a specific anxiety disorder (87%) received mental health services in the year following their index date, compared with 29% of those who did not receive a subsequent diagnosis of a specific anxiety disorder. 2.4. Prediction of diagnostic specificity Logistic regression (see Table 4) showed that participants who were older, female, or unmarried had increased odds of receiving a diagnosis of anxiety NOS, as did patients who did not have a serviceconnected disability. Significant clinical predictors of anxiety NOS included the absence of comorbid substance use, schizophrenia, or bipolar disorders. Additionally, patients who were diagnosed in primary care clinics were more likely to receive a diagnosis of anxiety NOS than patients diagnosed in integrated primary care mental health (OR=0.70) or specialty mental health settings (OR=0.36).

2. Results 2.5. Prediction of Mental Health Service Utilization 2.1. Diagnostic rates and location Table 1 lists the demographic and clinical characteristics of patients diagnosed with a new anxiety disorder in fiscal year 2010.

Results of a second logistic regression (see Table 5) revealed that patients diagnosed with a specific anxiety disorder were more likely to receive mental health services than those diagnosed with anxiety

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

195

Table 1 Sample Descriptive Characteristics Characteristic

Age in years, mean (SD) Gender, n (%) Male Female Marital status, n (%) Married Not married Unknown Income 0–30 30,001–40,000 40,001–50,000 50,001–60,000 N=60,001 Unknown Relative risk score, n (%) b0.26 0.26–1.00 1.01–2.00 N2.00 Distance to care facility in miles, mean (SD) Service-connected disability, n (%) No Yes Comorbid mental health diagnosis, n (%) Depression Substance use Schizophrenia Cognitive disorder Bipolar disorder

Total sample n=292,244

Specific anxiety diagnosis n=180,646

Anxiety NOS n=111,598

n (%)

n (%)

n (%)

55.2 (17.1)

53.4 (16.6)

57.9 (17.4)

267,626 (91.6) 24,618 (8.4)

168,085 (93.1) 12,561 (7.0)

99,541 (89.2) 12,057 (10.8)

163,275 (55.9) 122,961 (42.1) 6008 (2.1)

102,287 (56.6) 74,051 (41.0) 4308 (2.4)

60,988 (54.7) 48,910 (43.8) 1700 (1.5)

16,740 (5.7) 78,047 (26.7) 89,635 (30.7) 46,974 (16.1) 54,835 (18.8) 6013 (2.1)

65,369 123,776 53,928 49,171

(22.4) (42.4) (18.5) (16.4)

14.3 (15.6) 133,532 (45.7) 158,712 (54.3) 57,991 (19.8) 38,256 (13.1) 1729 (0.6) 2888 (1.0) 4569 (1.6)

10,847 48,293 54,883 29,170 34,094 3359

35,933 80,156 34,152 30,405

(5.9) (26.7) (30.4) (16.2) (19.0) (1.9)

(19.9) (44.7) (18.9) (16.8)

14.4 (16.6) 61,977 (34.3) 118,669 (65.7) 37,181 24,639 1232 1694 3249

NOS (OR=1.65). Additionally, patients who received their initial anxiety diagnosis in PCMHI (OR=41.91), specialty mental health (OR=16.29), or other specialty settings (OR=1.37) were more likely to receive mental health services than those diagnosed in primary care. Demographic characteristics associated with greater odds of receiving at least one mental health visit included younger age, male gender, service-connected disability status, and higher relative-risk scores. The presence of comorbid mental health disorders, including depression, schizophrenia, cognitive disorder, and bipolar disorder, also increased the likelihood of mental health service use; however, the presence of comorbid substance use resulted in decreased odds of outpatient mental health services. 3. Discussion This study used VHA administrative databases to evaluate diagnostic and mental health service use patterns among Veterans who received a diagnosis of a new anxiety disorder in fiscal-year 2010. Approximately 62% of the sample received a diagnosis of a specific anxiety disorder (PTSD, GAD, PD, SAD or SP), with PTSD being the most common index diagnosis (55% of the sample). Consistent with research indicating high rates of anxiety NOS diagnoses in active-duty service members [12], over one third of the sample received an anxiety NOS diagnosis. Multiple factors can explain the high rates of anxiety NOS; these rates may reflect that a large number of Veterans are presenting with significant anxiety symptoms that are not consistent with a specific anxiety disorder or may reflect overdiagnosis of anxiety NOS because of factors such as providers' lack of familiarity with criteria of anxiety diagnosis or insufficient time to assess patients comprehensively. Providers may also be using the NOS diagnosis as a placeholder until further assessment can determine the precise diagnosis; however,

(20.6) (13.6) (0.7) (0.9) (1.8)

Test statistic

P

F=4835.2

b.0001

χ2=1325.9

b.0001

χ2=428.8

b.0001

χ2=171.1

b.0001

χ2=1804.6

b.0001

F=31.87 χ2=24703.5

b.0001 b.0001

χ2=162.4 χ2=125.3 χ2=65.7 χ2=12.3 χ2=169.9

b.0001 b.0001 b.0001 .0004 b.0001

5893 (5.3) 29,754 (26.7) 34,752 (31.1) 17,804 (16.0) 20,741 (18.6) 2654 (2.4)

29,436 43,620 19,776 18,766

(26.4) (39.1) (17.7) (16.8)

14.1 (13.8) 71,555 (64.1) 40,043 (35.9) 20,810 (18.7) 13,617 (12.2) 497 (0.5) 1194 (1.1) 1320 (1.2)

contrary to our hypothesis, this study observed that most Veterans who received an anxiety NOS diagnosis did not go on to receive a disorder-specific diagnosis within the subsequent 12 months. Furthermore, patients diagnosed with anxiety NOS were significantly less likely to receive mental health care than patients diagnosed with a specific anxiety disorder. Another possibility is that anxiety NOS is being improperly diagnosed in patients whose symptoms may be better accounted for by another condition (such as a mood disorder), which may explain the lack of subsequent specific anxiety disorder diagnosis; however, if this were the case, we would not expect a lower rate of treatment utilization in this group, since patients would be likely to receive mental health services for the primary disorder. These findings suggest that diagnostic specificity may be an important factor in facilitating the provision of mental health services. Several patient factors increased the likelihood of an anxiety NOS diagnosis, including female gender, older age, the absence of specific comorbid diagnoses (i.e., substance-use disorders, bipolar disorder), and absence of service-connected disability. Patients with these characteristics may present more often with anxiety symptoms that do not fit a specific anxiety diagnosis; or, these findings may indicate disparities in care such that patients with these characteristics are less likely to receive comprehensive diagnostic evaluation. Additionally, patients diagnosed in primary care were more likely to receive a diagnosis of anxiety NOS than patients diagnosed in other care settings. Similarly, Veterans with PTSD and GAD were most commonly diagnosed in primary care settings; whereas Veterans with less prevalent diagnoses, such as OCD, PD, SP, and SAD, were most often diagnosed in specialty care settings. Diagnosis of these less common anxiety disorders often requires a detailed familiarity with diagnostic criteria, which may be more common among mental health specialists. Comparatively, primary care physicians, who represent the most frequent first line of care for patients

196

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

Table 2 Diagnostic rates by care setting Anxiety diagnosis

Anxiety NOS Specific anxiety Dx PTSD GAD PD OCD SAD SP

Total no. of patients diagnosed with condition

No. diagnosed in primary care

No. diagnosed in specialty MH

N

%

N

%

N

111,598 180,646 161,412 22,700 13,717 2901 1131 1106

38.2 61.8 55.2 7.8 4.7 1.0 0.4 0.4

75,050 88,892 79,702 9742 5316 1177 213 496

67.3 49.2 49.4 42.9 38.8 40.6 18.8 44.9

19,636 70,478 62,742 10,061 6477 1415 782 370

No. diagnosed in primary care MH

Other specialty care

%

N

%

N

%

17.6 39.0 38.9 44.3 47.2 48.8 69.1 33.5

4300 7234 6498 1389 807 134 103 74

3.9 4.0 4.0 6.1 5.9 4.6 9.1 6.7

12,612 14,042 12,470 1508 1117 175 33 166

11.3 7.8 7.7 6.6 8.1 6.0 2.9 15.0

Note: Patients with more than 1 specific anxiety diagnosis are included in each specific anxiety diagnostic category for which they were diagnosed.

with mental health diagnoses and the largest source of mental health care overall [19,20], often have less training in assessing and diagnosing mental health disorders and have many structural barriers, such as time and financial disincentives [10], to conducting full mental health assessments. It is also possible that Veterans in primary care present with primarily somatic symptoms of anxiety that are difficult to classify, and that these Veterans may be more likely to minimize their symptoms and need for subsequent mental health treatment [9]. Whereas prior studies have examined service use across anxiety disorders (i.e., grouping disorders into one diagnostic category) [21–23], this study is the first to compare mental health service use patterns associated with individual anxiety diagnoses. The proportion of patients in each diagnostic group that received mental health services in the year following diagnosis ranged from 32% (anxiety NOS) to 89% (SAD). Demographic and clinical predictors of service use were largely consistent with prior research and included male gender [22], young age [24], service-connected disability [25], high illness burden [26], and psychiatric comorbidity [27]. The majority of the

demographic and clinical characteristics predictive of greater service use were also associated with lower likelihood of anxiety NOS diagnosis; although not explicitly examined in the present study, it is possible that the absence of these factors in patients with anxiety NOS contributed to decreased access to services among this group. Also consistent with previous research [28], Veterans diagnosed in specialty mental health and PC-MHI settings. were more likely to receive subsequent mental health services than those diagnosed in primary care. This finding may be due in part to a greater likelihood that patients presenting in specialty care settings are treatment seeking as compared to those in primary care. Although PC-MHI accounted for only a small proportion of anxiety diagnoses, Veterans who were diagnosed in PC-MHI were much more likely to receive subsequent mental health services than those diagnosed in traditional primary care settings. While it is possible that differences between the primary care and PC-MHI patient populations may, in part, account for these differences, the comparatively higher rates of service use among those diagnosed in PC-MHI also likely reflects the success of the PCMHI initiative in improving Veterans’ access to mental health care.

Table 3 Mental health visits by diagnosis Type of visit

Diagnosis

No. of patients receiving visits

% of Patients receiving visits

Mean (SD) no. of visits

Any MH visit

PTSD GAD PD OCD SAD SP AnxNOS PTSD GAD PD OCD SAD SP AnxNOS

101,754 13,617 9125 1951 1002 494 36,179 50,396 5835 3828 866 493 263 16,685

63.0 60.0 66.5 67.3 88.6 44.7 32.4 31.2 25.7 27.9 29.9 43.6 23.8 15.0

5.59 (7.5) 2.80 (3.2) 2.94 (3.4) 3.24 (3.8) 2.98 (3.8) 2.72 (4.2) 2.75 (3.2) 6.03(8.3) 2.17 (3.7) 2.12 (3.8) 2.63 (4.3) 2.34 (4.1) 2.55 (5.3) 2.40 (3.6)

PTSD GAD PD OCD SAD SP AnxNOS PTSD GAD PD OCD SAD SP AnxNOS

30,587 4858 3211 683 340 123 10,076 54,360 7625 5483 1227 605 230 16,388

19.0 21.4 23.4 23.5 30.1 11.1 9.0 33.7 33.6 40.0 42.3 53.5 20.8 14.7

2.23 1.48 1.54 1.60 1.28 1.02 1.49 2.60 1.72 1.80 1.89 1.57 1.23 1.65

Psychotherapy

Medication visits Medication only

Medication+psychotherapy

MH, mental health; AnxNOS=anxiety not otherwise specified.

(1.8) (1.4) (1.4) (1.7) (1.5) (1.2) (1.4) (2.2) (1.6) (1.7) (1.8) (1.7) (1.3) (1.5)

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

197

Table 4 Multivariate logistic regression results predicting specificity of anxiety diagnosis

Demographic factors

Comorbid MH diagnosis

Care setting of initial anxiety diagnosis

Age Female gender Married status Service-connected disability Relative-risk score Income Distance to VA Depression Substance use Schizophrenia Cognitive disorder Bipolar disorder PC-MHI Specialty Mental Health Other

B

SE

P

Odds ratio

95% CI

0.02 0.69 −0.09 −1.47 0.01 −0.01 −0.01 −0.01 −0.18 −0.15 0.03 −0.22 −0.36 −1.02 −0.04

0.01 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.06 0.04 0.04 0.01 0.01 0.01

b.0001 b.0001 b.0001 b.0001 b.0001 .02 .01 .4 b.0001 .01 .48 b.0001 b.0001 b.0001 .0003

1.22 1.99 0.92 0.23 1.02 1.00 1.00 0.99 0.83 0.86 1.03 0.80 0.70 0.36 0.97

1.21–1.22 1.94–2.06 0.90–0.93 0.23–0.24 1.01–1.02 1.00–1.00 1.00–1.00 0.97–1.01 0.81–0.86 0.77–0.97 0.95–1.12 0.75–0.86 0.66–0.73 0.35–0.37 0.95–0.98

Note: Specific anxiety diagnosis was set as the reference category for the dependent variable. Primary care was set as the reference category for the diagnostic setting predictor variable.

Notably, patients with anxiety NOS were less likely than patients with a specific anxiety disorder to receive mental health care following diagnosis. These disparate rates may be attributable, in part, to the emphasis on disorder-specific treatments for anxiety disorders. For example, most existing psychotherapy protocols, such as Exposure and Response Prevention for OCD [29], Panic Control Treatment [30], and Cognitive Processing Therapy for PTSD [31] focus on the symptoms of a single anxiety disorder. No treatment has been designed specifically for anxiety NOS; and, to the authors’ knowledge, only a single case study examines treatment of anxiety NOS [32]. Therefore, providers may be less likely to initiate treatment with anxiety NOS patients due to a lack of clarity regarding effective treatment approaches for this population. The growing body of research on transdiagnostic anxiety treatments [33,34], which use common principles of disorder-specific, evidence-based treatments to target a range of related disorders, may hold promise for anxiety NOS [15]. An alternative explanation for the relatively low rates of service use among patients with anxiety NOS is that these patients may present with less disabling symptoms and, therefore, a lower need for treatment. A diagnosis of anxiety NOS, however, requires the presence of clinically significant impairment and does not necessarily imply lesser symptom severity or functional impact. Indeed, research on the reliability of DSM-IV diagnoses indicates that clinicians frequently disagree in their differential diagnosis of anxiety NOS versus other

anxiety disorders, particularly GAD, despite consistency regarding symptom severity [35]. Lack of diagnostic clarity may also contribute to low service use for patients with anxiety NOS; in cases where anxiety NOS is a provisional diagnosis, the diagnosing provider may not yet have enough information to determine which mental health services would be most appropriate for that patient, which may act as a barrier to referral. Further research clarifying the factors that contribute to low service use among this diagnostic group can inform efforts to increase access to services for those who present with significant anxiety symptoms but do not meet criteria for a specific anxiety disorder. 3.1. Limitations Limitations of the study include the use of CPT codes, developed by the American Medical Association, to assess mental health service use. These codes are entered by the provider (therefore reflecting the provider’s perspective of the care provided) but are not independently verified. CPT codes do not include information regarding the content of the session (e.g., the specific type of psychotherapy received) or the type of medication prescribed. Given our focus on mental health CPT codes, this study does not capture services provided in nonmental health visits (e.g., psychotropic medication prescribed during a non-mental health appointment with a primary care provider). The validity of the patient diagnoses, recorded by VHA

Table 5 Multivariate logistic regression results predicting receipt of mental health services following initial anxiety diagnosis

Demographic factors

Anxiety diagnostic specificity Comorbid MH diagnosis

Care setting of initial anxiety diagnosis

Age Female gender Married status Service-connected disability Relative-risk score Income Distance to VA Specific anxiety diagnosis Depression Substance use Schizophrenia Cognitive disorder Bipolar disorder PC-MHI Specialty mental health Other

B

SE

P

Odds Ratio

95% CI

−0.03 −0.15 −0.01 −0.1 0.13 −0.01 −0.01 0.5 0.53 −0.12 0.58 0.23 0.67 3.74 2.79 0.32

0.01 0.02 0.01 0.01 0.01 0.01 0.01

b.0001 b.0001 .94 b.0001 b.0001 .01 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001

0.71 0.86 0.10 0.91 1.14 1.00 0.99 1.65 1.70 0.89 1.79 1.26 1.95 41.91 16.29 1.37

0.70–0.71 0.83–0.89 0.98–1.02 0.89–0.93 1.14–1.15 1.00–1.00 0.99–0.99 1.62–1.68 1.67–1.74 0.86–0.91 1.57–2.05 1.15–1.39 1.79–2.12 36.68–47.88 15.80–16.80 1.35–1.40

0.01 0.01 0.07 0.05 0.04 0.07 0.02 0.01

Note: Lack of mental health service utilization following receipt of the index diagnosis was set as the reference category. MH, mental health.

198

T.L. Barrera et al. / General Hospital Psychiatry 36 (2014) 192–198

providers and possibly not based on a structured diagnostic interview, is similarly unknown. However, this procedure represents standard clinical practice and, thus, may increase the generalizability of the findings. We did not assess patients’ use of mental health services after the 1-year follow-up period and information regarding Veterans’ use of non-VA services was not available to us. As our sample included only Veterans enrolled in the VHA, replication in a non-VA setting is warranted. 4. Conclusions This is the first study to examine the link between diagnostic specificity and mental health service utilization for Veterans with newly diagnosed anxiety disorders. Our results suggest that Veterans who receive a diagnosis of anxiety NOS are less likely to access mental health services than Veterans who receive specific anxiety diagnoses. Given that the majority of anxiety NOS diagnoses were initiated in primary care and few of these diagnoses were later clarified, the need for enhanced diagnostic and referral practices in these settings is highlighted by these findings. References [1] Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 2005;62:617–27. [2] Jordan BK, Schlenger WE, Hough R, et al. Lifetime and current prevalence of specific psychiatric disorders among Vietnam veterans and controls. Arch Gen Psychiatry 1991;48:207–15. [3] Hoge CW, Castro CA, Messer SC, et al. Combat duty in Iraq and Afghanistan, mental health problems, and barriers to care. N Engl J Med 2004;251:13–22. [4] Roy-Byrne PP, Katon W. Generalized anxiety disorder in primary care: the precursor/modifier pathway to increased health care utilization. J Clin Psychiatry 1997;58(Suppl. 3):34–8 discussion 39–40. [PMID: 9133491]. [5] Speer DC, Schneider MG. Mental health needs of older adults and primary care: opportunity for interdisciplinary geriatric team practice. Clin Psychol: Sci Pract 2003;10(1):85–101. [6] Stein MB, Roy-Byrne PP, Craske MG, et al. Functional impact and health utility of anxiety disorders in primary care outpatients. Med Care 2005;43:1164–70. [7] Stein MB. Attending to anxiety disorders in primary care. J Clin Psychiatry 2003;64(Suppl. 15):35–9. [8] Katon W, Roy-Byrne P. Anxiety disorders: efficient screening is the first step in improving outcomes. Ann Intern Med 2007;146(5):390–2. [9] Tylee A, Walters P. Underrecognition of anxiety and mood disorders in primary care: why does the problem exist and what can be done? J Clin Psychiatry 2007;68(Suppl. 2):27–30. [10] American Academy of Family Physicians. Mental health care services by family physicians (position paper). Available at http://www.aafp.org/about/policies/a; 2011. Accessed July 30, 2013. [11] Psychiatric Association American. Diagnostic & Statistical Manual of Mental Disorders, Fourth Edition – Text Revision; 2000. Arlington, VA.

[12] Lovering ME, Proctor SP, Heaton KJ. A retrospective study of anxiety disorder diagnoses in the military from 2000 to 2009. J Anxiety Disord 2013;27:25–32. [13] Spitzer RL, Kroenke K, Linzer M, et al. Health-related quality of life in primary care patients with mental disorders. JAMA 1995;274:1511–7. [14] Stanley MA, Roberts RE, Bourland SL, Novy DM. Anxiety disorders among older primary care patients. J Clin Geropsychol 2001;7(2):105–16. [15] McManus F, Shafran R, Cooper Z. What does a ‘transdiagnostic’ approach have to offer the treatment of anxiety disorders? Br J Clin Psychol 2010;49:491–505. [16] Veterans Affairs Information Resource Center (VIReC). Data Quality. Available at http://www.virec.research.va.gov/DataQuality/Overview.htm; 2012. [17] Ellis RP, Ash AS, Ayanian JZ, et al. Diagnosis based risk adjustment for Medicare capitation payments. Health Care Financ Rev 1996;17:101–28. [18] Petersen LA, Pietz K, Woodard L, et al. Comparison of the predictive validity of diagnosis-based risk adjusters for clinical outcomes. Med Care 2005;43:61–7. [19] Wang PS, Aguilar-Gaxiola S, Alonso J, et al. Use of mental health services for anxiety, mood, and substance disorders in 17 countries in the WHO world mental health surveys. Lancet 2007;370(9590):841–50. [20] Wang PS, Demler O, Olfson M, et al. Changing profiles of service sectors used for mental health care in the united states. Am J Psychiatry 2006;163(7): 1187–98. [21] Cully JA, Tolpin L, Henderson L, et al. Psychotherapy in the Veterans Health Administration: missed opportunities? Psychol Serv 2008;5(4):320–31. http: //dx.doi.org/10.1037/a0013719. [22] Mott J, Hundt NE, Sansgiry S, Mignogna J, Cully JC. Changes in psychotherapy use among veterans with depression, anxiety, and PTSD. Psychiatr Serv 2013 in press. [23] Burnett-Zeigler IE, Pfieffer P, Zivin K, et al. Psychotherapy utilization for acute depression within the Veterans Affairs health care system. Psychol Serv 2012;9: 325–35. [24] Hunt MG, Rosenheck RA. Psychotherapy in mental health clinics of the Department of Veterans Affairs. J Clin Psychol 2011;67(6):561–73. http://dx.doi.org/10.1002/jclp.20788. [25] Sayer NA, Spoont M, Nelson DB. Disability compensation for PTSD and use of VA mental health care. Psychiatr Serv 2004;55:589. [26] Possemato K, Wade M, Andersen J, Ouimette P. The impact of PTSD, depression, and substance use disorders on disease burden and health care utilization among OEF/OIF veterans. Psychol Trauma 2010;2:218–23. [27] Chermack ST, Zivin K, Valenstein M, et al. The prevalence and predictors of mental health treatment services in a national sample of depressed veterans. Med Care 2008;46(8):813–20. [28] Spoont M, Murdoch M, Hodges J, Nugent S. Treatment receipt by veterans after a PTSD diagnosis in PTSD, mental health, or general medical clinics. Psychiatr Serv 2010;61(1):58–63. [29] Kozak M, Foa E. Mastery of obsessive-compulsive disorder: a cognitive-behavioral approach. San Antonio: Psychological Corp; 1997. [30] Barlow DH, Craske MG. Mastery of your anxiety and panic II. Albany (NY): Graywind Publications; 2000. [31] Resick PA, Monson CM, Chard KM. Cognitive processing therapy, veteran/military version: therapist’s manual. Boston, MA: Department of Veterans Affairs, National PTSD Center; August 2008. [32] Shafran R, McManus F, Lee M. A case of anxiety disorder not otherwise specified: a transdiagnostic approach. Inter J Cogn Ther 2008;1:256–65. [33] Barlow DH, Allen LB, Choate ML. Towards a unified treatment for emotional disorders. Behav Ther 2004;35:205–30. [34] Norton PJ, Barrera TB. Transdiagnostic versus diagnosis-specific CBT for anxiety disorders: a preliminary randomized controlled non-inferiority trial. Depress Anxiety 2012;29:874–82. [35] Brown TA, DiNardo PA, Lehman CL, Campbell LA. Reliability of DSM-IV anxiety and mood disorder disorders. implications for the classification of emotional disorders. J Abnorm Psychol 2001;110:49–58.