Posttraumatic stress disorder and odds of major invasive procedures among U.S. Veterans Affairs patients

Journal of Psychosomatic Research 75 (2013) 386–393

Contents lists available at ScienceDirect

Journal of Psychosomatic Research

Posttraumatic stress disorder and odds of major invasive procedures among U.S. Veterans Affairs patients David Sloan Greenawalt a,⁎, Laurel Anne Copeland b,c, Andrea A. MacCarthy d, Fangfang F. Sun b,c, John Edward Zeber b,c a

VISN 17 Center of Excellence for Research on Returning War Veterans, 4800 Memorial Drive, Waco, TX 76711, USA Center for Applied Health Research, Central Texas Veterans Health Care System, Department of Veterans Affairs, 1901 Veterans Memorial Drive, Temple, TX 76504, USA Center for Applied Health Research, Scott & White Healthcare, 2102 Birdcreek Drive, Temple, TX 76502, USA d Center for Improving Veterans Health Through Research, South Texas Veterans Health Care System, South Texas Veterans Health Care System, 7400 Merton Minter Boulevard, San Antonio, TX 78229-4404, USA b c

a r t i c l e

i n f o

Article history: Received 8 February 2013 Received in revised form 15 August 2013 Accepted 17 August 2013 Keywords: Posttraumatic stress disorder Depression Veterans Invasive procedures Surgery

a b s t r a c t Objectives: Although individuals with posttraumatic stress disorder (PTSD) are at heightened risk for several serious health conditions, research has not examined how having PTSD impacts receipt of invasive procedures that may alleviate these problems. We examined whether PTSD, after controlling for major depression, was associated with odds of receiving common types of major invasive procedures, and whether race, ethnicity, and gender was associated with odds of procedures. Methods: Veterans Health Administration patients with PTSD and/or depression were age-matched with patients without these disorders. The odds of invasive hip/knee, digestive system, coronary artery bypass graft/percutaneous coronary intervention (CABG/PCI), and vascular procedures during FY2006–2009 were modeled for the full sample of 501,489 patients and for at-risk subsamples with medical conditions alleviated by the procedures examined. Results: Adjusting for demographic covariates and medical comorbidity, PTSD without depression was associated with decreased odds of all types of procedures (odds ratios [OR] range 0.74–0.82), as was depression without PTSD (OR range 0.59–0.77). In analyses of at-risk patients, those with PTSD only were less likely to undergo hip/ knee (OR = 0.78) and vascular procedures (OR = 0.73) but not CABG/PCI. African-Americans and women atrisk patients were less likely to undergo hip/knee, vascular, and CABG/PCI procedures (OR range 0.31–0.82). Conclusion: With the exception of CABG/PCI among at-risk patients, Veterans with PTSD and/or depression were less likely to undergo all types of procedures examined. Future studies should examine the reasons for this disparity and whether it is associated with subsequent adverse outcomes. Published by Elsevier Inc.

Introduction Posttraumatic stress disorder (PTSD) is a major problem among Veterans of military service, particularly Veterans receiving care in the Veterans Health Administration (VA) where more than 400,000 are treated annually for this disorder. Estimates on the lifetime prevalence of combat-related PTSD for U.S. Veterans range from 15 to 31% for Vietnam Veterans, 6–10% for Gulf War Veterans, and 8–13% for Veterans from Operations Enduring Freedom or Iraqi Freedom (OEF/OIF) [1]. Associated with high levels of functional impairment and poor qualityof-life [2], PTSD is not only a debilitating mental illness but is also associated with significant physical disorders. For example, Vietnam era Veterans with PTSD are more likely to have cardiovascular problems than those without the disorder [3,4]. Several studies have reported greater incidence of physical health problems among Veterans with ⁎ Corresponding author at: Albany Stratton VA Medical Center, 113 Holland Avenue (151), Albany, NY 12208, USA. Tel.: +1 512 795 9566; fax: +1 518 626 5628. E-mail address: [email protected] (D.S. Greenawalt). 0022-3999/$ – see front matter. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jpsychores.2013.08.012

PTSD, especially chronic pain, and gastrointestinal and musculoskeletal disorders [5–8]. Because Veterans with PTSD are likely to have more physical health problems, it would be expected that they use more medical services than Veterans without PTSD, all other factors being equal. Consistent with this premise, a VA study indicated that among younger Veterans, those with PTSD used about twice as many outpatient medical services as those without the diagnosis [9], and several recent studies have found that OEF/OIF Veterans with PTSD had more medical visits than those without PTSD [10–12]. However, one study found no association [13], and another found that a PTSD diagnosis was associated with fewer medical visits [14]. A less studied measure of medical service use is receipt of invasive procedures. Disparities in the receipt of invasive procedures are a cause for concern because not having medically necessary procedures may increase the risk of subsequent serious health problems or death. Because patients with PTSD are more likely to have coronary heart disease than those without the disorder [15,16], these patients likely have an equal or greater need for coronary artery bypass graft (CABG)

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

or percutaneous coronary intervention (PCI; angioplasty). Failure to receive such procedures when indicated may place these patients at increased risk of myocardial infarction and suggests a need for improved quality of care in this area. Several studies have found a link between mental illness and decreased levels of care. A national study of VA patients receiving active medical treatment found that those who had psychiatric disease, who did not, were less likely to receive key preventive services [17]. Further, among patients hospitalized for acute myocardial infarction, those with a comorbid mental disorder were less likely to undergo subsequent coronary revascularization procedures and experienced increased mortality linked to lower quality follow-up care [18,19]. While a few studies have investigated the occurrence of PTSD following invasive procedures [20,21], to our knowledge no published study has examined the association of pre-existing PTSD on the likelihood of undergoing invasive procedures. A recent VA study, however, found that patients with pre-existing major depressive disorder (MDD) were less likely to have common types of major invasive procedures than nondepressed patients [22]. Because MDD and PTSD frequently co-occur [23], however, it is possible that when both disorders are considered, PTSD rather than MDD is associated with reduced odds of surgery. Alternatively, these two disorders may interact in affecting odds of invasive procedures. A further issue which merits continued investigation is possible racial, ethnic, and gender disparities in treatment found in previous research. Studies have identified disparities in cardiac revascularization procedures [24,25], surgical treatment for cancer [26,27], and joint replacement [28]. The primary purpose of this paper was to examine whether PTSD, after controlling for MDD, was associated with the likelihood of having four common types of major invasive procedures at the VA. A secondary purpose was to examine the association of race, gender, and Hispanic ethnicity with common invasive procedures. Methods Study design This study was approved by the local institutional review board and employed retrospective secondary analysis of data gathered for the Surgical Treatment Outcomes of Patients with Psychiatric Disorders (STOPP) project. STOPP assembled data from administrative extracts of the VA's all-electronic medical records system to examine rates, types, and outcomes of major invasive procedures for patients with severe mental illness, including MDD and PTSD. Studies examining diagnoses, types of care received, and demographic variables have demonstrated the validity of VA administrative databases, although race/ ethnicity data are frequently missing [29,30]. Data from fiscal years (FY) 2005–2009 (October 2004–September 2009) were used. Current Procedural Terminology (CPT) and ICD-9-A codes identified the four types of procedures (Appendix A) [31,32]. Study sample and subsamples Patients were eligible for inclusion if they received care in the VA during FY2006; had valid race/ethnicity data, date of birth, and U.S. military Veteran status as indicated by their priority status; and were given a diagnosis of PTSD (ICD-9 code 309.81) or MDD (296.2, 296.3, 311) on at least two different outpatient dates during FY2006, following recommendations of Frayne et al. [33] and Gravely et al. [34]. Each patient with PTSD or MDD was randomly matched with two patients of the same age without PTSD or MDD. Once we identified the entire cohort, we determined if these patients had any of the following types of major invasive procedures most common during fiscal years 2006– 2009: hip/knee repair or replacement, digestive system operations, or cardiovascular procedures—subdivided into CABG/PCI, and vascular operations. Exclusion criteria were: 1) major invasive procedures other than hip/knee repair or replacement, digestive system operations,

387

CABG/PCI, or vascular operations; 2) diagnosis of schizophrenia (ICD-9 295 excluding 295.5) or bipolar disorder (296.0–296.1, 296.4–296.8) as indicated by treatment on one inpatient or two outpatient dates in FY2006. In addition, we conducted analyses on subsamples of at-risk patients who had ICD-9 codes for medical conditions alleviated by the invasive procedures we examined (e.g., osteoarthritis for hip/knee replacement; Appendix B). For these analyses we included medical data from FY2005 in order to identify medical preconditions that would indicate a need for later procedures. To be included, patients who underwent a procedure had to have a medical precondition that preceded the procedure date by at least one year. For non-procedure patients, we included medical data for FY2005 and the medical precondition had to precede the date that they qualified for study inclusion (e.g., their second outpatient diagnosis of PTSD or MDD in FY2006) by at least one year. We then repeated our analysis of receipt of subsequent hip/knee, CABG/PCI, or vascular procedures from FY2006–2009 on this subset of at-risk patients. We did not conduct at-risk analyses for digestive surgery because of the heterogeneity of underlying medical conditions that these surgeries address and the difficulty of accurately identifying at-risk patients. Measures For patients who underwent procedures, diagnosis codes during the 12 months preceding major invasive procedures determined MDD/ PTSD status and medical comorbidity burden at baseline. For nonprocedure patients, data from FY2006 were used. Both inpatient and outpatient data informed measures of medical comorbidity. The Charlson score [35], a weighted sum of 19 conditions associated with one-year post-hospitalization mortality, and the Selim physical comorbidity index [36], which sums 30 common chronic medical conditions, were computed following the method of Pugh and colleagues [37]. The Charlson and Selim indices describe different aspects of patient comorbidity burden and are only modestly correlated. Additional measures were age, gender, race, and Hispanic ethnicity. VA priority status, denoting eligibility for VA benefits, was gathered from enrollment files. Because priority status is related to socioeconomic status, disease severity, and use of private health insurance, it is an important covariate [38]. Priority 1 Veterans are at least 50% disabled by a military service-connected condition and have no co-pays for care or drugs; Priority 2–6 Veterans have pharmacy co-pays and have up to 40% service-connected disability, specific service experiences, or low income. Priority 7–8 Veterans incur co-pays for care and drugs. Analysis Demographics were reported for the overall sample and for mental disorder subgroups (PTSD only, MDD only, PTSD and MDD, No PTSD or MDD). Group differences were assessed using chi-square analyses for categorical variables and Student's t-tests or analysis of variance for interval variables. We first conducted unadjusted models without any covariates to examine the relation between the primary predictors of interest (mental health diagnoses) and the outcome variables (relative odds of each type of invasive procedures). Next, we conducted adjusted logistic regression models that controlled for medical morbidity, race, Hispanic ethnicity, age, gender, and Priority 1 status, and included a term for the interaction between PTSD and MDD diagnosis. Because this interaction term was significant for each procedure type, final models for the full sample and for at-risk subsamples estimated effects for PTSD alone, MDD alone, and PTSD and MDD, with no PTSD or MDD as the referent group. Wald chi-square demonstrated statistical significance for each model, and the c-statistic assessed model fit. The c-statistic ranges from 0.5, indicating a model no better than chance, to 1.0, indicating a model with perfect fit.

388

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

Table 1 Baseline characteristics of total sample (n = 501,489) Mean Age Selim physical comorbidity index Charlson comorbidity score

Race White African-American Other Hispanic ethnicity Gender Male Female Marital status Married Not married Missing data VA priority status Priority 1 Priority 2 Priority 3 Priority 4 Priority 5 Priority 6 Priority 7 Priority 8 Physical disorders Hypertension Dyslipidemia Take statins or lipid-lowering drugs Mental health disorders PTSD only MDD only PTSD and MDD Procedure type Hip/knee Digestive CABG/PCI Vascular

56.79 2.22 0.99

Results

SD

Median (min–max)

13.26 1.85 1.57

57.00 (18.00–100.00) 3.00 (0.00–17.00) 0.00 (0.00–21.00)

n

%

386,079 101,246 14,164 31,146

77.0 20.2 2.8 6.2

468,678 32,811

93.5 6.5

268,712 232,688 89

53.6 46.4 0.0

113,971 40,834 59,197 17,500 164,797 22,455 14,053 68,682

22.7 8.1 11.8 3.5 32.9 4.5 2.8 13.7

265,389 242,818 216,292

52.9 48.4 43.1

55,239 83,456 27,466

11.0 16.6 5.5

4923 3299 2480 17,319

1.0 0.7 0.5 3.5

Demographic and clinical characteristics Mean age for the sample of 501,489 Veterans was 56.8 years (SD = 13.3). The sample was predominantly male and white; 20% were African-American. Roughly one-third qualified for VA care because of low income (Priority 5) whereas more than one-fourth qualified due to greater than 50% service-connected disability (Priority 1) or catastrophic disability of any cause (Priority 4). Approximately 5% (n = 26,196) of the sample underwent major invasive procedures during the study period. Seven percent of procedure patients (n = 1823) had multiple procedures (Table 1). Women were considerably younger (M = 45.9 years; SD = 13.6) than men (M = 57.6 years; SD = 12.9; t(37,024) = − 150.65, p b .001). Hispanic patients (M = 53.4, SD = 41.1) were younger than non-Hispanic patients (M = 57.0, SD = 13.2; t(34,846) = 43.92, p b .001). Mean age also varied by race, F(2, 501,486) = 7714.14, p b .001. African-American patients (M = 52.4, SD = 11.7) were younger on average than patients in the "other" race category (M = 54.7, SD = 13.7), who, in turn, were younger than white patients (M = 58.0, SD = 13.4). Patients with both PTSD and MDD were younger than other mental illness groups (Table 2). Patients without PTSD or MDD were less ill per Charlson and Selim scores than patients with PTSD and/or MDD. Gender, race, ethnicity, and VA priority group differed significantly by mental illness category. Patients with MDD only, PTSD only, or MDD and PTSD were more likely than those without either diagnosis to have dyslipidemia (χ2 (3) = 1452.45, p b .001) or hypertension (χ2 (3) = 685.26, p b .001). Unadjusted models for full sample In unadjusted models, a diagnosis of PTSD only or MDD only was associated with increased odds of hip/knee procedures (Table 3). Having both PTSD and MDD diagnoses was associated with increased odds of having hip/knee, digestive, and vascular procedures. Adjusted models for full sample Preliminary adjusted multiple logistic regression models revealed a significant PTSD × MDD interaction for each outcome. Follow-up contrasts for hip/knee procedures indicated that patients with MDD alone had lower relative odds of undergoing procedures than patients with both PTSD and MDD (Wald χ2 (1) = 6.86, p = .009). The same pattern emerged for odds of digestive procedures (Wald χ2 (1) = 6.63, p = .010), CABG/PCI (Wald χ2 (1) = 8.79, p = .003), and vascular procedures (Wald χ2 (1) = 40.73, p b .001). Final logistic regression models were highly significant for hip/knee (Wald χ2 (10) = 8573.75, c-statistic = 0.83, p b .001); digestive (Wald χ2 (10) = 6066.43, c-statistic = 0.81, p b .001); CABG/PCI (Wald χ2 (10) = 4201.20, c-statistic = 0.83, p b .001); and vascular procedures (Wald χ2 (10) = 28261.92, c-statistic = 0.86, p b .001). Final adjusted models for the full sample revealed that patients with PTSD only were less likely to undergo all types of procedures than were patients without PTSD or MDD

Table 2 Patient characteristics by mental illness status PTSD only n = 55,239

Age Selim physical Charlson

Female Race White African−American Other Hispanic ethnicity Married Priority1 status Medical diagnoses Hypertension Dyslipidemia Procedures⁎ Hip/knee Digestive CABG/PCI Vascular

MDD only n = 83,456

PTSD and MDD n = 27,466

No PTSD or MDD n = 335,328

Mean

SD

Mean

SD

Mean

SD

Mean

SD

56.73b 2.52b 1.02b

12.40 1.93 1.53

57.39c 2.66c 1.18d

14.13 2.00 1.71

54.77a 2.64c 1.06c

11.90 1.97 1.58

56.81b 2.03a 0.93a

13.26 1.76 1.53

n

%

n

%

n

%

n

%

2139−

3.87

8727+

10.46

2711+

9.87

19,234−

5.74

40,986− 12,120+ 2133+ 3944+ 33,196+ 35,459−

74.20 21.94 3.86 7.14 60.10 64.19

67,29+ 14,145− 2014− 5196 39,498− 16,298+

80.64 16.95 2.41 6.23 47.33 19.53

20,718 5736 1012− 1958+ 15,337a 14,932+

75.43 20.88 3.68 7.13 55.84 54.37

257,078+ 69,245− 9005− 20,048− 180,681 47,282−

76.66 20.65 2.69 5.98 53.88 14.10

30,565+ 29,160+

55.33 52.79

46,662+ 42,974+

55.91 51.49

15,017+ 14,594+

54.67 53.13

173,145− 156,090−

51.63 46.55

673 358 291 1919

1.22 0.65 0.53 3.47

817 580 377 2763

0.98 0.69 0.45 3.31

378 228 166 1095

1.38 0.83 0.60 3.99

3055 2133 1646 11,542

0.91 0.64 0.49 3.44

PTSD = posttraumatic stress disorder; MDD = major depressive disorder. Different letter subscripts denote significant differences between groups (PTSD only vs. MDD only vs. PTSD and MDD vs. No PTSD or MDD). + Frequency higher than expected by chance. − Frequency lower than expected by chance. ⁎ Unconditional tests of group differences not conducted.

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

389

Table 3 Relative odds of undergoing invasive procedures Procedure type

Hip/knee PTSD only MDD only PTSD and MDD Hispanic African-American Age in decades Female Priority 1 (no co-pay) Selim physical comorbidity Charlson comorbidity score Digestive system PTSD only MDD only PTSD and MDD Hispanic African-American Age in decades Female Priority 1 (no co-pay) Selim physical comorbidity Charlson comorbidity score

Unadjusted full sample (n = 501,489)

Adjusted full sample (n = 501,489)

OR

95% CI

OR

95% CI

OR

95% CI

1.34⁎ 1.08 1.52⁎

1.23–1.46 1.00–1.16 1.36–1.69

0.82⁎ 0.69⁎ 0.82⁎

0.74–0.90 0.64–0.75 0.73–0.92 0.88–1.13 0.85–0.98 0.78–0.82 0.93–1.19 1.17–1.34 2.01–2.08 0.58–0.61

0.78⁎ 0.62⁎ 0.80⁎

0.69–0.89 0.55–0.70 0.68–0.93 0.82–1.18 0.69–0.85 0.77–0.84 0.69–0.98 0.80–0.96 1.57–1.65 0.68–0.73

0.99 0.91⁎ 0.80⁎ 1.05 1.25⁎ 2.04⁎ 0.60⁎

1.02 1.09 1.31⁎

0.91–1.14 1.00–1.19 1.14–1.50

0.82⁎ 0.77⁎ 0.95 1.16⁎ 1.03 0.85⁎ 1.23⁎ 0.99 1.41⁎ 1.24⁎

Adjusted at-risk subsample (n = 15,896)

0.99 0.77⁎ 0.80⁎ 0.82⁎ 0.88⁎ 1.61⁎ 0.70⁎

0.72–0.92 0.70–0.84 0.82–1.09 1.01–1.33 0.95–1.13 0.83–0.88 1.05–1.43 0.91–1.07 1.38–1.43 1.22–1.26 Adjusted at-risk subsample

CABG/PCI PTSD only MDD only PTSD and MDD Hispanic African-American Age in decades Female Priority 1 (no co-pay) Selim physical comorbidity Charlson comorbidity score Vascular PTSD only MDD only PTSD and MDD Hispanic African-American Age in Decades Female Priority 1 (no co-pay) Selim physical comorbidity Charlson comorbidity score

(n = 63,697) 1.07 0.92 1.23⁎

0.95–1.22 0.82–1.03 1.05–1.45

0.81⁎ 0.63⁎ 0.84⁎

0.99

0.70–0.93 0.56–0.71 0.71–0.99 0.74–1.06 0.49–0.62 0.76–0.81 0.13–0.28 0.85–1.04 1.61–1.68 0.96–1.01

0.74⁎ 0.59⁎ 0.76⁎ 0.87⁎ 0.80⁎ 0.78⁎ 0.45⁎ 0.89⁎ 1.74⁎ 1.14⁎

0.70–0.78 0.56–0.62 0.71–0.82 0.80–0.93 0.77–0.84 0.77–0.80 0.40–0.50 0.85–0.92 1.73–1.76 1.13–1.15

0.88 0.55⁎ 0.78⁎ 0.19⁎ 0.94 1.64⁎

1.01 0.96 1.17⁎

0.96–1.06 0.92–1.00 1.09–1.24

0.94 0.63⁎ 0.87 1.06 0.56⁎ 0.57⁎ 0.31⁎ 0.85 1.44⁎ 0.92⁎ 0.73⁎ 0.55⁎ 0.66⁎ 1.12 0.82⁎ 0.60⁎ 0.60⁎ 0.84⁎ 1.51⁎ 1.03⁎

0.74–1.19 0.52–0.76 0.65–1.17 0.76–1.48 0.45–0.70 0.53–0.61 0.16–0.60 0.72–1.01 1.39–1.49 0.88–0.96

0.66–0.81 0.51–0.60 0.58–0.75 0.98–1.28 0.76–0.89 0.59–0.62 0.49–0.73 0.78–0.90 1.49–1.53 1.01–1.05

⁎ 95% confidence interval excludes 1.0.

(Table 3). Similarly, patients with MDD only were less likely to have all types of procedures than those without PTSD or MDD. Having PTSD and MDD was associated with lower odds of hip/knee, CABG/PCI, and vascular procedures, but not digestive procedures. Ethnicity, race, and gender were significantly related to likelihood of procedures. Hispanic Veterans were more likely than non-Hispanic Veterans to have digestive procedures but less likely to undergo vascular procedures. African-Americans were less likely than other Veterans to receive CABG/PCI, hip/knee, and vascular procedures. Female gender was associated with higher odds of digestive procedures but much lower odds of CABG/ PCI or vascular procedures. Increasing age was associated with lower odds of undergoing all types of procedures examined. Each decade increase in age was associated with 15%–22% cumulative reduced odds of hip/knee, digestive, CABG/PCI, or vascular procedures. Medical comorbidity had mixed associations with receipt of procedures. Higher Selim scores were associated with greater odds of all types of procedures whereas higher Charlson scores were associated with greater likelihood of digestive and vascular procedures, but much lower odds of hip/knee procedures. Adjusted models for at-risk patients Logistic multiple regression models including only at-risk patients were highly significant for hip/knee procedures (n = 15,896; Wald χ2 (10) = 1339.39, c-statistic = 0.73,

p b .001); CABG/PCI procedures (n = 63,697; Wald χ2 (10) = 592.92, c-statistic = 0.75, p b .001); and vascular procedures (n = 63,697; Wald χ2 (10) = 4435.75, c-statistic = 0.77, p b .001). For hip/knee procedures for at-risk patients (e.g., those with osteoarthritis), patients with PTSD only, MDD only, or both PTSD and MDD had significantly lower odds of undergoing these procedures. Among at-risk patients diagnosed with cardiovascular disease, patients with PTSD only, MDD only, or both PTSD and MDD had lower relative odds of having vascular procedures relative to patients without these diagnoses. For CABG/PCI, patients with MDD only had lower odds of surgery, but those with PTSD only or both MDD and PTSD did not (Table 3). African-American and women at-risk patients were less likely to undergo hip/knee, vascular, and CABG/PCI procedures. Consistent with results for the whole sample, increasing age was associated with lower odds of procedures. Higher Selim scores were associated with higher odds of procedures whereas higher Charlson scores were associated with increased odds of vascular but decreased odds of hip/knee and CABG/PCI procedures.

Discussion For all four types of invasive procedures examined in the full sample, Veterans with PTSD only were less likely to have the procedures than

390

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

patients without a PTSD or MDD diagnosis. Vascular procedures showed the strongest effect; patients with PTSD only were about 25% less likely to receive the procedures. Significant results for CABG/PCI for the entire sample were tempered by analyses of at-risk patients, which did not show reduced odds of CABG/PCI among patients with PTSD, although they showed reduced odds of vascular and hip/knee procedures. Patients with PTSD are more likely to have coronary heart disease [15,16], and patients with PTSD in our sample were more likely to be diagnosed with dyslipidemia and hypertension than those with neither PTSD nor MDD. For these reasons, disparities for CABG/PCI and vascular procedures are of concern, and raise the possibility that patients with PTSD at high risk for heart attack and stroke may be less likely to receive needed treatment. This study also found that patients with MDD had lower odds of undergoing all four types of procedures examined in the full sample. The relative odds of undergoing CABG/PCI or vascular procedures among patients with MDD were particularly low—35 to 40% lower than for patients with neither PTSD nor MDD. In addition, MDD was associated with lower odds of invasive procedures in all analyses of at-risk patients. As with PTSD patients, patients with MDD in our sample were more likely to have dyslipidemia and hypertension, and MDD has also been linked to cardiovascular disease in general population studies [39]. Therefore, the lower odds of receiving these types of procedures may result in increased risk of heart attack or stroke, a possibility to be determined in future research. Results are consistent with two other large analyses based on STOPP data [22,40] and raise the question of why VA patients with PTSD or MDD appear to be less likely to undergo common invasive procedures after accounting for medical comorbidity and other demographic covariates. Physicians may be more hesitant to recommend major procedures for patients with a mental illness diagnosis because they believe that the outcomes for these patients will be poor. For example, pre-existing depression is associated with worse outcomes following surgery [41] and CABG in particular [42]. Alternatively, patients with PTSD or MDD may be more reluctant to undergo procedures. Patients with MDD or military-related PTSD may have difficulty trusting authority figures [43,44], and may be less likely to provide consent for invasive procedures for this reason. A third possibility is that physicians are less likely to identify Veterans with PTSD and MDD who are appropriate candidates for procedures. Combat Veterans who screen positive for PTSD are much more likely to report difficulty taking care of their health than those who screen negative [45], and thus may be less likely to follow up with their physicians when symptoms appear or get worse. Veterans with severe PTSD or MDD may be more likely to be homebound and to lack energy needed to resolve medical issues. A further possibility is that patients with PTSD or MDD in this sample had higher levels of physical comorbidities than those without one of these diagnoses. As a result, they may not remember or have adequate time during appointments to discuss all current symptoms. Finally, knowledge of a patient's past psychiatric history may affect a physicians' estimation of the probability of a disease. For example, when a patient with a past psychiatric history presents with severe pain symptoms, physicians may be more likely to attribute the symptoms to the psychiatric disorder rather than another cause [46]. Although patients with both MDD and PTSD in the full sample had reduced odds of three types of procedures, analyses revealed that odds of having procedures were higher among those with PTSD and MDD than those with MDD alone. It may be that patients with both diagnoses experience more somatic symptoms and greater distress, which leads them to seek more frequent medical care, increasing the odds that physicians will recommend invasive procedures. The study found that African-Americans had lower odds of having three types of procedures than did other Veterans. Given that AfricanAmericans are more likely than non-Hispanic whites to die of heart

disease [47], their reduced odds of receiving CABG/PCI or vascular procedures are potentially problematic. Other research has found that African-Americans appear to be less likely to get CABG or other revascularization surgery, both within and outside of the VA, even when the procedures are needed and appropriate [24]. Women in the sample were much less likely to undergo CABG/PCI or vascular procedures. This finding is consistent with literature reviews that indicate that women are less likely than men to receive PCI [48] or CABG [49]. Although women have lower rates of heart disease than men, coronary heart disease is the number one cause of death among women [48], so disparities in receipt of effective treatments are potentially problematic. One possible explanation for the gender disparity is that physicians are more hesitant to recommend CABG for women. Some past research has indicated that they are more likely than men to die after coronary bypass procedures, although more recent studies that account for age and comorbidities suggest similar mortality rates [50]. An unexpected result was that increasing age was associated with lower odds of procedures. The finding that the odds of hip and knee procedures decreased with age was especially surprising and was at odds with a large study of the U.S. Medicare population [51]. To better understand this finding, we divided the sample into five age groups (18–35; 36–49; 50–60; 61–72; 73 years and older) following the methods employed by Noel et al. [52], who examined a large Veteran sample. Veterans 50–60 years old comprised 47% of our sample and far outnumbered any other age group. In addition, this age group had significantly higher odds of procedures. Because more Veterans in our sample were older than younger compared to this group (150,308 versus 114, 387), it likely resulted in a negative association between age and procedure odds. Another factor that may be contributing to the negative association of invasive procedures with age is the tendency for older Veterans to undergo invasive procedures outside the VA using their Medicare coverage [53]. To examine this possibility, we conducted analyses excluding Veterans over age 65 who were Medicare-eligible. In contrast to earlier results, we now found higher odds of undergoing hip/knee, CABG/PCI, and vascular procedures. However, for digestive procedures, the original finding of lower odds with increasing age persisted. Also noteworthy is that a higher Charlson score was associated with lower odds of hip/knee procedures. It may be that physicians judged the mortality risk for patients with high Charlson scores to be too high to justify these procedures. Alternatively, patients with life-threatening conditions may see less value in undergoing and recovering from hip/ knee procedures. Limitations and strengths This study relied on administrative data compiled from electronic medical records. Severity ratings of psychiatric illnesses or indicators of the relative need for major procedures were not available. Whereas current diagnoses of mental disorders during the study period were captured, lifetime histories are not encoded in VA administrative databases. Diagnoses of PTSD and MDD were dependent on administrative coding, which may misclassify some cases [33]. Medicare and other out-of-system data were not available, and any major procedures and/ or mental health treatment conducted outside of the VA were not included in analyses. However, results are consistent with a large study of Medicare patients that found lower rates of revascularization procedures among those with mental health diagnoses [18]. A further limitation is that a large percentage of VA patients are missing data on race/ ethnicity and it is not clear how this may impact results in research examining healthcare disparities [54], although a recent study suggests the impact may be minimal [55]. Although we included women and controlled for gender, the sample was mostly male and results may not generalize to women. Finally, this VA sample of patients, compared to a general population sample, was likely to be older, poorer, have a higher proportion of African-Americans, and experience more illnesses

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

[56]. As a result, the study sample was likely biased toward sicker patients. Strengths of the study include its large scale, encompassing half a million individuals and the longitudinal design which enabled us to track surgeries over a four-year period (FY2006–2009). Further strengths were the inclusion of large numbers of patients with both PTSD and MDD, which enabled us to control for possible effects of each mental illness as well as the combination of both, and analyses of at-risk patients who were most likely to need invasive procedures. Future research should examine possible causes for lower odds of major procedures among VA patients with PTSD and MDD. Explanations may include physician judgments that patients with PTSD or MDD are at greater risk for complications following procedures; reluctance of PTSD or MDD patients to undergo invasive procedures; physician beliefs that physical pain may be a consequence of mental illness rather than medical problems; patients with PTSD or MDD underreporting physical symptoms; or high numbers of comorbid psychological and physical problems that result in some symptoms being overlooked. Future studies should also address whether patients with PTSD are eschewing invasive treatments in favor of pharmacological approaches to managing cardiovascular disease.

Conclusion Veterans with PTSD and/or MDD appear to be at heightened risk for several serious health problems, such as coronary heart disease and other cardiovascular disease. However, the current study found that they were less likely to undergo common types of major invasive procedures that may be indicated for these problems. Further research is needed to investigate possible explanations for these disparities. Future

391

studies should also examine whether disparities are associated with subsequent adverse outcomes for those with PTSD, MDD, or both. The current study also found that African-American and female Veterans were less likely to receive CABG/PCI, suggesting that vigilance is needed to ensure that these groups receive appropriate care while optimizing resource deployment of publicly funded health care services.

Conflict of interest The authors have no competing interests to report.

Disclaimer The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.

Acknowledgments This study was funded by a grant from the Veterans Health Administration, Health Services Research and Development #IIR-09-335 (PI: Laurel A. Copeland). Preparation of this manuscript was supported by the Department of Veterans Affairs Office of Academic Affiliations Advanced Fellowship Program in Mental Illness Research and Treatment, VISN 17 Center of Excellence for Research on Returning War Veterans, Central Texas Veterans Health Care System (CTVHCS), and the Center for Applied Health Research, a research center jointly sponsored by CTVHCS and Scott & White Healthcare, Temple, TX.

Appendix A. Definitions of types of invasive procedures

Procedures Code type

Values included

Hip/Knee

27090, 27091, 27125, 27120, 27130, 27132, 27134, 27137, 27138, 27299, 27310, 27437, 27438, 27440, 27441, 27442, 27443, 27445, 27446, 27447, 27486, 27487, 27488, 29850, 29851, 29870, 29871, 29873, 29874, 29875, 29876, 29877, 29880, 29881, 29882, 29884, 29885, 29886, 29887, 29888 77.66, 77.70, 77.76, 78.06, 78.16, 78.17, 78.46, 78.56, 79.26, 79.36, 80.05, 80.26, 80.46, 80.6, 80.76, 80.96, 81.47, 81.51, 81.52, 81.53, 81.54, 81.55 42410, 42415, 43030, 43280, 43289, 43310, 43320, 43610, 43621, 43632,43644, 43653, 43659, 43846, 44005, 44055, 44120, 44121, 44130, 44139, 44141, 44145, 44147, 44200, 44201, 44238, 44602, 44603, 44604, 44620,44850, 45110, 45111, 45112, 45170, 46020, 46040, 47560, 47612, 48140,48150, 48520, 49000, 49002, 49020, 49085, 49255, 49320, 49321, 49323, 49329 42.0, 42.1, 42.19, 42.41, 42.42, 42.5, 42.6, 42.7, 42.81, 42.82, 42.83, 42.84, 42.85, 42.89, 42.91, 42.92, 42.99, 43.0, 43.5, 43.6, 43.7, 43.8, 43.9, 44.0, 44.21, 44.29, 44.40, 44.41, 44.42, 44.43, 44.44, 44.45, 44.46, 44.47, 44.48, 44.49, 44.5, 44.60, 44.61, 44.62, 44.63, 44.64, 44.65, 44.66, 44.67, 44.68, 44.69, 44.90, 44.91, 44.92, 44.93, 44.94, 44.95, 44.96, 44.97, 44.98, 44.99, 45.01, 45.02, 45.03, 45.10, 45.15, 45.19, 45.26, 45.31, 45.41, 45.43, 45.49, 45.50, 45.51, 45.52, 45.61, 45.62, 45.63, 45.90, 45.91, 45.92, 45.93, 45.94, 45.95, 46.01, 46.02, 46.03, 46.1, 46.21, 46.22, 46.23, 46.24, 46.30, 46.31, 46.32, 46.33, 46.39, 46.40, 46.41, 46.42, 46.43, 46.44, 46.45, 46.46, 46.47, 46.48, 46.49, 46.51, 46.52, 46.6, 46.64, 46.71, 46.73, 46.74, 46.75, 46.76, 46.79, 46.81, 46.82, 46.85, 46.93, 46.94, 46.99, 47.93, 47.99, 48.0, 48.1, 48.4, 48.60, 48.66, 48.67, 48.68, 48.70, 48.71, 48.72, 48.73, 48.74, 48.75, 48.76, 48.77, 48.78, 48.79, 48.81, 48.82, 48.9, 48.93, 48.99, 51.00, 51.01, 51.02, 51.03, 51.04, 51.05, 51.06, 51.07, 51.08, 51.09, 51.10, 51.11, 51.12, 51.13, 51.14, 51.16, 51.17, 51.18, 51.19, 51.21, 51.32, 51.36, 51.37, 51.39, 51.40, 51.41, 51.42, 51.43, 51.44, 51.45, 51.46, 51.47, 51.48, 51.49, 51.51, 51.59, 51.60, 51.61, 51.62, 51.63, 51.64, 51.65, 51.66, 51.67, 51.68, 51.69, 51.7, 51.80, 51.81, 51.82, 51.83, 51.84, 51.85, 51.86, 51.87, 51.88, 51.89, 51.90, 51.91, 51.92, 51.93, 51.94, 51.95, 51.96, 51.98, 52.00, 52.01, 52.02, 52.03, 52.04, 52.05, 52.06, 52.07, 52.08, 52.09, 52.10, 52.12, 52.15, 52.16, 52.17, 52.18, 52.22, 52.11, 52.3, 52.4, 52.5, 52.6, 52.7, 52.8, 52.90, 52.91, 52.92, 52.93, 52.94, 52.95, 52.96, 52.97, 52.98, 52.99, 54.0, 54.1, 54.3, 54.4, 54.5, 54.6, 54.7 33510, 33511, 33512, 33513, 33514, 33516, 33517, 33518, 33519, 33521, 33522, 33523, 33533, 33534, 33535, 33536, 92975, 92977, 92980, 92981, 92982, 92984, 92995 36.03, 36.11, 36.12, 36.13, 36.14, 36.15, 36.16, 36.17, 36.19, 36.31 33200, 33206, 33207, 33208, 33210, 33212, 33214, 33218, 33222, 33223, 33233, 33246, 33322, 33405, 33422, 34111, 34201, 34800, 34803, 34804, 34808, 34812, 34825, 34826, 35081, 35102, 35131, 35141, 35151, 35152, 35190, 35207, 35221, 35301, 35363, 35390, 35450, 35470, 35471, 35473, 35474, 35475, 35476, 35493, 35495, 35509, 35456, 35556, 35558, 35566, 35571, 35582, 35646, 35656, 35661, 35840, 35860, 35879, 35881, 36818, 36819, 36821, 37720, 37730, 37785, 37799 00.50, 00.51, 00.56, 00.57, 00.58, 00.59, 00.60, 00.62, 00.66, 00.67, 00.68, 00.69, 35.0, 35.11, 35.12, 35.13, 35.14, 35.21, 35.22, 35.23, 35.24, 35.27, 35.28, 35.33, 33.39, 35.4, 35.50, 35.51, 35.53, 35.54, 35.55, 35.56, 35.57, 35.58, 35.59, 35.61, 35.62, 35.71, 35.72, 35.8, 35.91, 35.93, 35.95, 35.99, 36.01, 36.02, 36.04, 36.05, 36.06, 36.07, 36.09, 36.32, 36.39, 36.91, 36.99, 37.10, 37.11, 37.12, 37.13, 37.14, 37.15, 37.16, 37.17, 37.18, 37.19, 37.21, 37.22, 37.23, 37.24,37.25, 37.26, 37.27, 37.28, 37.29, 37.31, 37.32, 37.33, 37.34, 37.4, 37.61, 37.62, 37.64, 37.65, 37.80, 37.81, 37.82, 37.83, 37.94, 39.61, 39.99, 38.0, 38.1, 38.3, 38.40, 38.42, 38.43, 38.44, 38.45, 38.46, 38.47, 38.48, 38.49, 38.5, 38.6, 38.7, 38.8, 39.0, 39.1, 39.22, 39.23, 39.24, 39.27, 39.28, 39.29, 39.25, 39.26,39.30, 39.31, 39.32, 39.41, 39.42, 39.43, 39.49, 39.50, 39.51, 39.52, 39.53,39.55, 39.56, 39.57, 39.58, 39.59, 39.71, 39.72, 39.79, 39.8, 39.90, 39.91, 39.98

CPT ICD-9-A CPT

Digestive

ICD-9-A

CABG/PCI

CPT

Vascular

ICD-9-A CPT

ICD-9-A

CPT = Current Procedural Terminology; ICD-9 = International Classification of Disease, Ninth Revision.

392

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393

Appendix B. ICD-9-A codes for medical conditions identifying at-risk patients

Procedures

Values included

Hip/knee

170.6, 715.5, 715.35, 733.42, 718.05, 719.95, 726.5, 730.05, 730.15, 730.25, 808.0, 808.1, 808.2, 808.3, 808.40, 808.41, 808.42, 808.43, 808.44, 808.49, 808.51, 808.52, 808.53, 808.54, 808.59, 808.8, 808.9, 715.96, 732.7, 730.06, 730.16, 730.26, 717.3, 717.4, 836.0, 836.1, 836.2, 717.83, 844.2, 822.0, 822.1 402, 402.0, 402.00, 402.02, 402.1, 402.10, 402.11, 402.9, 402.90, 402.91, 404, 410, 410.0, 410.1, 410.2, 410.3, 410.4, 410.5, 410.6, 410.7, 410.8, 410.9, 411.0, 411.1, 411.81, 412, 413, 413.0, 413.1, 413.9, 414.19, 414.2, 414.8, 425.4, 427, 427.0, 427.1, 427.2, 427.3, 427.31, 427.32, 427.4, 427.41, 427.42, 427.5, 427.8, 427.81 427.89, 427.9, 428, 428.0, 428.1, 428.2, 428.20, 428.21, 428.22, 428.23, 428.3, 428.30, 428.31, 428.32, 428.33, 428.4, 428.40, 428.41, 428.42, 428.43, 428.9, 428.9, 429.1, 429.4, 429.7, 429.71, 429.79, 430, 431, 432.0–432.9, 433.0–433.91, 434–434.91, 435–435.9, 436, 437–437.9, 438–438.9, 997.1, V12.53

CABG/PCI and vascular

Appendix C. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.jpsychores.2013.08.012. References [1] Richardson JD, Frueh BC, Acierno R. Prevalence estimates of combat-related PTSD: a critical review. Aust N Z J Psychiatry 2010;44:4–19. [2] Schnurr PP, Lunney CA, Bovin MJ, Marx BP. Posttraumatic stress disorder and quality of life: extension of findings to veterans of the wars of Iraq and Afghanistan. Clin Psychol Rev 2009;29:727–35. [3] Boscarino JA. Posttraumatic stress disorder and physical illness: results from clinical and epidemiologic studies. Ann N Y Acad Sci 2004;1032:141–53. [4] Boscarino JA. A prospective study of PTSD and early-age heart disease mortality among Vietnam Veterans: implications for surveillance and prevention. Psychosom Med 2008;70:668–76. [5] Andersen J, Wade M, Possemato K, Ouimette P. Association between posttraumatic stress disorder and primary care provider-diagnosed disease among Iraq and Afghanistan veterans. Psychosom Med 2010;72:498–504. [6] Moeller-Bertram T, Keltner J, Strigo IA. Pain and post traumatic stress disorder— review of clinical and experimental evidence. Neuropharmacology 2012;62:586–97. [7] Richardson JD, Pekevski J, Elhai JD. Post-traumatic stress disorder and health problems among medically ill Canadian peacekeeping veterans. Aust N Z J Psychiatry 2009;43:366–72. [8] Schnurr PP, Spiro A, Paris AH. Physician-diagnosed medical disorders in relation to PTSD symptoms in older male military veterans. Health Psychol 2000;19:91–7. [9] Calhoun PS, Bosworth HB, Grambow SC, Dudley TK, Beckham JC. Medical service utilization by veterans seeking help for posttraumatic stress disorder. Am J Psychiatry 2002;159:2081–6. [10] Cohen BE, Gima K, Bertenthal D, Kim S, Marmar CR, Seal K. Mental health diagnoses and utilization among returning Iraq and Afghanistan veterans. J Gen Intern Med 2009;25:18–24. [11] Frayne SM, Chiu VY, Iqbal S, Berg MS, Laungani KJ, Cronkite RC, et al. Medical care needs of returning veterans with PTSD: their other burden. J Gen Intern Med 2010;26:33–9. [12] Hoge CW, Terhakopian A, Castro CA, Messer SC, Engel CC. Association of posttraumatic stress disorder with somatic symptoms, health care visits, and absenteeism among Iraq war veterans. Am J Psychiatry 2007;164:150–3. [13] 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. [14] Craddock-O'Leary J, Young AS, Yano EM, Wang M, Lee ML. Use of general medical services by V.A. patients with psychiatric disorders. Psychiatr Serv 2002;53:874–8. [15] Ahmadi N, Hajsadeghi F, Mirshkarlo HB, Budoff M, Yehuda R, Ebrahimi R. Post-traumatic stress disorder, coronary atherosclerosis, and mortality. Am J Cardiol 2011;108:29–33. [16] Kubzansky LD, Koenen KC, Spiro III A, Vokonas PS, Sparrow D. Prospective study of posttraumatic stress disorder and coronary heart disease in the Normative Aging Study. Arch Gen Psychiatry 2007;64:109–16. [17] Druss BG, Rosenheck RA, Desai MM, Perlin JB. Quality of preventive medical care for patients with mental disorders. Med Care 2002;40:129–36. [18] Druss BG, Bradford DW, Rosenheck RA, Radford MJ, Krumholz HM. Mental disorders and use of cardiovascular procedures after myocardial infarction. JAMA 2000;283:506–11. [19] Druss BG, Bradford DW, Rosenheck RA, Radford MJ, Krumholz HM. Quality of medical care and excess mortality in older patients with mental disorders. Arch Gen Psychiatry 2001;53:565–72.

[20] Favaro A, Gerosa G, Caforio ALP, Volpe B, Rupolo G, Zarneri D, et al. Posttraumatic stress disorder and depression in heart transplantation recipients: the relationship with outcome and adherence to medical treatment. Gen Hosp Psychiatry 2011;33:1–7. [21] Schelling G, Roozendall B, Krauseneck T, Schmoelz M, Quervain DE, Briefel D, et al. Efficacy of hydrocortisone in preventing posttraumatic stress disorder following critical illness and major surgery. Ann N Y Acad Sci 2006;1070:46–53. [22] Copeland LA, Zeber JE, Pugh MJ, Phillips KL, Lawrence VA. Ethnicity and race variations in receipt of surgery among veterans with and without depression. Depress Res Treat 2011;2011:370962. [23] Kinder LS, Bradley KA, Katon WJ, Ludman E, McDonnell MB, Bryson CL. Depression, posttraumatic stress disorder, and mortality. Psychosom Med 2008;70:20–6. [24] Epstein AM, Weissman JS, Schneider EC, Gatsonis C, Leape LL, Piana RN. Race and gender disparities in rates of cardiac revascularization. Do they reflect appropriate use of procedures or problems in quality of care? Med Care 2003;41:1240–55. [25] Freund KM, Jacobs AK, Pechacek JA, White HF, Ash AS. Disparities by ethnicity, race, and sex in treating acute coronary syndromes. J Womens Health 2012;21:126–32. [26] Morris AM, Billingsley KG, Baxter NN, Baldwin L-M. Racial disparities in rectal cancer treatment: a population-based analysis. Arch Surg 2004;139:151–5. [27] Fairley TL, Cardinez CJ, Martin J, Alley L, Friedman C, Edwards B, et al. Colorectal cancer among U.S. adults younger than 50 years of age, 1998–2001. Cancer 2006;107:1153–61. [28] Dunlop DD, Manheim LM, Song J, Sohn M, Feinglass JM, Chang HJ, et al. Age and racial/ ethnic disparities in arthritis-related hip and knee surgeries. Med Care 2008;46:200–8. [29] Borzecki AM, Wong AT, Hickey EC, Ash AS, Berlowitz DR. Identifying hypertensionrelated comorbidities from administrative data: what's the optimal approach? Am J Med Qual 2004;19:201–6. [30] Kressin NR, Chang BH, Hendricks A, Kazis LE. Agreement between administrative data and patients' self reports of race/ethnicity. Am J Public Health 2003;93:1734–9. [31] American Medical Association. CPT: Current Procedural Terminology. Chicago, IL: American Medical Association; 2008. [32] Buck CJ. 2010 ICD-9-CM for Hospitals, vols. 1, 2, & 3. Maryland Heights, MO: Saunders; 2011 [Standard ed.]. [33] Frayne SM, Miller DR, Sharkansky EJ, Jackson VW, Wang F, Halanych JH, et al. Using administrative data to identify mental illness: what approach is best? Am J Med Qual 2010;25:42–50. [34] Gravely AA, Cutting A, Nugent J, Grill K, Carlson K, Spoont M. Validity of PTSD diagnoses in VA administrative data: comparison of VA administrative PTSD diagnosis to self-reported PTSD checklist scores. J Rehabil Res Dev 2011;48:21–30. [35] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83. [36] Selim AJ, Fincke G, Ren XS, Lee A, Rogers WH, Miller DR, et al. Comorbidity assessments based on patient report: results from the Veterans Health Study. J Ambul Care Manage 2004;27:281–95. [37] Pugh MJ, Copeland LA, Zeber JE, Cramer JA, Amuan ME, Cavazos JE, et al. The impact of epilepsy on health status among younger and older adults. Epilepsia 2005;56:1820–7. [38] Shen Y, Hendricks A, Zhang S, Kazis LE. VHA enrollees' health care coverage and use of care. Med Care Res Rev 2003;60:253–67. [39] Rugulies R. Depression as a predictor for coronary heart disease: a review and metaanalysis. Am J Prev Med 2002;23:51–61. [40] Copeland LA, Sako EY, Wang CP, Mortensen EM, Pugh MJ, Zeber JE, et al. Severe mental illnesses are associated with receipt of surgery and postoperative outcomes in a retrospective analysis of patients in the Veterans Health Administration; 2013 [Manuscript submitted for publication]. [41] Rosenberger PH, Jokl P, Ickovics J. Psychosocial factors and surgical outcomes: an evidence-based literature review. J Am Acad Orthop Surg 2006;7:397–405. [42] Burg MM, Benedetto MC, Rosenburg R, Soufer R. Presurgical depression predicts medical morbidity 6 months after coronary artery bypass graft surgery. Psychosom Med 2003;65:111–8. [43] Kravitz RL, Paterniti DA, Epstein RM, Rochlen AB, Bell RA, Cipri C, et al. Relational barriers to depression help-seeking in primary care. Patient Educ Couns 2011;82:207–13. [44] Dekel R, Monson CM. Military-related post-traumatic stress disorder and family relations: current knowledge and future directions. Aggress Violent Behav 2010;15:303–9. [45] Sayer NA, Noorbaloochi S, Frazier P, Carlson K, Gravely A, Murdoch M. Reintegration problems and treatment interest among Iraq and Afghanistan combat veterans receiving VA medical care. Psychiatr Serv 2010;61:589–97. [46] Graber MA, Bergus G, Dawson JD, Wood GB, Levy BT, Levin I. Effect of a patient's psychiatric history of physicians' estimation of probability of disease. J Gen Intern Med 2000;15:204–6. [47] Agency for Healthcare Research, Quality. National Healthcare Disparities Report 2011. Rockville, MD: AHRQ; 2012 [Retrieved from http://www.ahrq.gov/qual/ nhdr11/nhdr11.pdf]. [48] Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, et al. American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics — 2011 update: a report from the American Heart Association. Circulation 2011;123:e18-209. [49] Hussain KMA, Kogan A, Estrada AQ, Kostandi G, Foschi A. Referral pattern and outcome in men and women undergoing coronary artery bypass surgery: a critical review. Angiology 1998;49:243–50. [50] Lundberg G, King S. Coronary revascularization in women. Clin Cardiol 2012;35:156–9. [51] Mahomed NN, Barrett JA, Katz JN, Phillips CB, Losina E, Lew RA, et al. Rates and outcomes of primary and revision total hip replacement in the United States Medicare population. J Bone Joint Surg Am 2003;85:27–32. [52] Noel PH, Wang C-P, Bollinger MJ, Pugh MJ, Copeland LA, Tsevat J, et al. Intensity and duration of obesity-related counseling: association with 5-year BMI trends among obese primary care patients. Obesity (Silver Spring) 2012;20:773–82.

D.S. Greenawalt et al. / Journal of Psychosomatic Research 75 (2013) 386–393 [53] Hynes DM, Koelling K, Strope K, Arnold N, Mallin K, Sohn M-N, et al. Veterans access to and use of Medicare and Veterans Affairs health care. Med Care 2007;45:214–23. [54] Long JA, Bamba MI, Ling B, Shea JA. Missing race/ethnicity data in Veterans Health Administration based disparities research: a systematic review. J Health Care Poor Underserved 2006;17:128–40.

393

[55] Copeland LA, Pugh MJ, Hicks PB, Noel PN. Use of obesity-related care by psychiatric patients. Psychiatr Serv 2012;63:230–6. [56] Agha Z, Lofgren RP, VanRuiswyk JV, Layde PM. Are patients at Veterans Affairs medical centers sicker? A comparative analysis of health status and medical resource use. Arch Intern Med 2000;21:352–7.