| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Epidemiology, The University of Iowa College of Public Health, Iowa City, Iowa (L.E.J., C.P.C); and Regenstrief Institute, Indianapolis, Indiana, and the Departments of Psychiatry and Internal Medicine, Indiana University School of Medicine, Indianapolis, Indiana (C.P.C.).
Address correspondence and reprint requests to Caroline P. Carney, MD, Indiana University School of Medicine, 1050 Regenstrief Blvd., RG6, Indianapolis, IN 46202-2213. E-mail: ccarneyd{at}iupui.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
Methods: This was a retrospective analysis of a 100% sample of Blue Cross/Blue Shield of Iowa administrative claims data, 1996 to 2001. Logistic regression was used to calculate unadjusted and adjusted odds ratios (OR) for receipt of angioplasty (PTCA) and bypass graft surgery (CABG) within 30 days of discharge.
Results: A total of 3368 adults, aged 18 to 64 years, were hospitalized for myocardial infarction (MI) and 40% (n = 1342) had a mental disorder. Subjects with mental disorders were more likely to be younger, female, urban residents, and to have increased cardiovascular and medical comorbidity. They were similarly likely as subjects without mental disorders to have received PTCA (OR, 1.10; 95% confidence interval [CI], 0.95–1.29) and CABG (OR, 0.89; 95% CI, 0.71–1.11) in analyses adjusted for demographic and clinical characteristics. Revascularization rates did not differ by mental disorder type, with few exceptions.
Conclusions: Receipt of revascularization was similar for patients with and without mental disorders. Our results may differ from previous findings as a result of the younger population studied and increased comorbidity in people with mental disorders, which may have resulted in a contraindication for surgical intervention. Conversely, the increased burden of comorbidity could suggest that these patients should have received PTCA at higher rates because of the better prognosis associated with revascularization as compared with medical management. Prospective analyses with review of clinical data and behavioral risk factors are necessary to determine why some patients with mental illness may be less likely to receive cardiac interventions.
Key Words: myocardial infarction • revascularization • mental disorders • claims data • comorbidity
Abbreviations: MI = myocardial infarction; CVD = cardiovascular disease; MHD = mental health disorders; PTCA = percutaneous transluminal coronary angioplasty; CABG = coronary artery bypass graft surgery; CPT = Common Procedural Terminology; ICD = International Classification of Diseases; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; OR = odds ratio; COPD = chronic obstructive pulmonary disease; CI = confidence interval.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
Medical therapy and invasive interventions such as percutaneous transluminal coronary angioplasty (PTCA) and coronary artery bypass graft surgery (CABG) are standard treatments for patients hospitalized with MI. Guidelines regarding the use of PTCA and CABG have been published (17,18). A recent metaanalysis suggested that CABG was associated with a decline in mortality at 5 years post-MI, less angina, and fewer repeat revascularization procedures as compared with PTCA (19). If cardiac interventions such as these were not routinely delivered to people with depression and other psychiatric syndromes, one could hypothesize that substandard care may have contributed to the increased mortality in these patients.
Whether people with MHD routinely receive the same invasive cardiac services for angina or after MI has not been uniformly reported (13,15,16,20,21). That people with MHD may have an increased risk for lack of receipt of appropriate cardiac interventions after MI is supported by evidence that these individuals fail to receive other recommended medical interventions (22). People with MHD may fail to receive appropriate medical care as a result of stigmatization from medical providers, perceived noncompliance with care because the presence of psychiatric symptoms may interfere with the necessary postprocedure care (e.g., bedrest after catheterization), or inability to obtain informed consent (23).
We are aware of only four studies to date that have examined receipt of cardiac procedures among people with MHD. Those studies indicate that some people with MHD are less likely to receive PTCA and CABG compared with people without MHD (15,16,20,21). However, these studies have been limited to a Veterans' Administration population of male patients, elders with Medicare coverage, a prospective, population-based record linkage of mental health services users in Western Australia, and an inpatient claims data analysis. Whether the previously reported findings would apply to a younger population of commercially insured men and women with MHD is unknown. We hypothesized that subjects with MHD were less likely to have received cardiac interventions. Because all subjects were similarly insured, the presence or absence of healthcare financing per se may be less likely to confound the analyses.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
Study Population
The study population included all adults ages 18 to 64 years who filed at least 1 claim for medical service from 1996 to 2001. Subjects were classified as residing in an urban or rural Iowa county based on the metropolitan statistical area definition (24). The basic medical insurance coverage was similar among subjects, with only a small proportion (<10%) enrolled in a managed care plan.
Myocardial Infarction Population
As validated by other studies, we used ICD-9 code 410 to identify subjects with a hospitalization for MI occurring in any diagnostic position (15,20,25–27). To increase specificity of MI, subjects who were not hospitalized overnight for an MI were excluded. Because we did not have access to medical records, we also excluded hospitalizations for other acute coronary syndromes (e.g., unstable angina) to ensure the specificity of the MI claim.
Mental Disorders
Subjects were classified as having an MHD if ICD-9 codes (290–319, 607.84, 608.89, 625.00, 625.80, 780.09, 780.52, 780.54, 780.59, 787.60) for MHD were identified in the claims data before or within the 30-day period after discharge. We included subjects first diagnosed with a MHD within the 30-day period after discharge because they may be at risk for lack of revascularization. Subjects with MHD claims appearing only after the 30-day period after discharge were excluded given that we were unable to determine if the MHD was a result of the MI/revascularization or was not coded in the claims data before the MI. Results did not differ when only subjects with MHD coded before the MI admission date were included (data not shown).
The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) has organized MHD into 17 major diagnostic categories based on ICD-9 codes (28). We examined specific MHD (adjustment, anxiety, cognitive, mood, psychotic, sexual, substance, and "other" disorder) and assigned subjects into a single diagnostic category based on the most clinically prominent disorder in the claims data. Categorization was determined by a combination of the most frequently occurring MHD diagnosis and the severity of the MHD, based on the hierarchy of MHD hospitalization, psychiatrist diagnoses, and by diagnoses made by any other provider type. We did not restrict the diagnostic classification to claims filed only in the period before the discharge date given the chronic and/or recurrent nature of many of these conditions.
Outcome Measures
The primary outcome of interest was receipt of PTCA (CPT 92980–92982, 92984, 92995, 92996; ICD-9 procedural codes 36.00–36.09) and CABG (CPT 33510–33536, 33560, 33570; ICD-9 procedural codes 36.10–36.99) during the hospitalization or within 30 days of discharge. We also examined receipt of these procedures within 90 days and 180 days of discharge (data not shown), and revascularization receipt rates were similar to rates for the 30-day period. We examined receipt of cardiac catheterization (CPT 93526–93529, 93539–93545, 93555–93556; ICD-9 procedural codes 37.21–37.23), and because more than 90% of subjects in both the MHD and non-MHD populations received this procedure, we did not further analyze this variable.
Confounders
The analyses were controlled for age, gender, number of days hospitalized, residence (rural, urban), hospital transfer (yes, no, unknown), cardiovascular risk factors, and other medical comorbidity. Information on hospital characteristics such as location, size, and availability of cardiac procedures was unavailable in this data.
We adapted the Elixhauser Comorbidity Index to assess medical comorbidity using both inpatient and outpatient claims data (29). Comorbid conditions were counted if they occurred in the inpatient setting or in multiple outpatient claims for a period of greater than 30 days before the discharge date, as validated by Klabunde (30). This time period was used to ensure that acute or miscoded outpatient comorbidities were not included in the total comorbidity count. We excluded mental health comorbidities (n = 4) and comorbidities with prevalence rates less than 1% (n = 8). Additionally, we adjusted for angina and hypercholesterolemia given that these are risk factors for MI and may influence the decision to pursue revascularization.
Statistical Analyses
Demographic and clinical characteristics were compared using chi-squared tests for categorical variables and t tests for continuous variables. Logistic regression was used to determine if receipt of PTCA and CABG differed among subjects with any MHD, as well as the specific MHD diagnoses, as compared with subjects without MHD. Unadjusted and adjusted odds ratios (ORs) are presented. The ORs were adjusted for all confounders mentioned previously. Alpha was set at 0.05 (two-sided) for all comparisons. All analyses were performed with SAS version 8.2 (SAS Institute, Inc., Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
As shown in Table 1, subjects with any MHD were more likely to be younger, women, urban residents, and to have more cardiovascular risk factors (e.g., peripheral vascular disorders, hypertension, chronic obstructive pulmonary disease [COPD], obesity, hypercholesterolemia) and other medical comorbidity (e.g., neurological disorders, deficiency anemias). More than 33% of subjects with MHD had 3 or more medical/cardiovascular comorbidities as compared with 27% of subjects without MHD. Given that difference in follow up may contribute more opportunities for comorbidities to be coded in the claims data, we compared the follow-up periods from the first claim date to the discharge date for both groups. Subjects with MHD had a longer follow up than those without (26 versus 16 months; p < .01). A trend toward greater likelihood for rehospitalization for MI was noted for subjects with MHD as compared with those without MHD (11.6% versus 9.6%; p = .06).
|
Approximately 25% (n = 830) of all subjects did not receive revascularization during the study period. Sixty percent (n = 2020) received PTCA only and 436 (13.0%) received CABG only. Less than 3% (n = 82) received both procedures, and the majority (n = 55) had claims for PTCA and CABG occurring on the same day. Because we cannot distinguish the temporal order of the procedures with claims filed for the same date of service, we included subjects who received both procedures (n = 82) in both the PTCA and CABG regression models.
Table 2 shows the unadjusted and adjusted ORs and 95% confidence intervals (CIs) for receipt of PTCA and CABG during the index hospitalization or within 30 days of discharge. The effect of MHD on receipt of cardiac services was not marked in this population. People with and without MHD had a similar likelihood of receiving PTCA in both unadjusted and adjusted analyses. In unadjusted analyses, people with any type of MHD were 19% less likely (OR, 0.81; 95% CI, 0.67–0.99) to have received CABG as compared with people without a MHD, but this difference did not persist in adjusted analyses (OR, 0.89; 95% CI, 0.71–1.11). Few exceptions to these generalizations regarding receipt of revascularization were noted for subjects with specific types of mental disorders. However, subjects with substance disorders were statistically more likely (OR, 1.36; 95% CI, 1.10–1.69) to have received PTCA. Subjects with "other" disorders were 1.77 times more likely (95% CI, 1.11–2.82) to have received PTCA and were 0.44 times less likely (95% CI, 0.20–0.95) to have received CABG in analyses adjusted for demographic and clinical factors.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
Reasons for lack of receipt of revascularization in those with MHD are plentiful. Patients with MHD may be unfairly stigmatized because physicians believe that these patients are not appropriate candidates for revascularization as a result of a presumption of patient noncompliance with postsurgical medical/behavioral recommendations. Patients diagnosed with depression are less likely to adhere to a low-fat diet, maintain regular exercise, reduce stress, and have adequate social support after MI (31). Moreover, another study has indicated that depressed patients are 3 times more likely to be noncompliant with medical care. Because noncompliance with medical and behavioral advice/recommendations is a significant problem among patients with MHD, the use of more definitive revascularization procedures may be rationalized to be superior to medical management alone (23). Certainly, postprocedure care and medication compliance is essential, and the risks of postoperative noncompliance may outweigh any benefits of revascularization interventions.
People with MHD in our study had more medical comorbidity than people without MHD, contrary to the results of a previous analysis (15). Hypertension was almost 5% more common in this group, and smoking-related comorbidities such as COPD and peripheral vascular disorders were twice as common for people with MHD as compared with those without. In general, people with COPD may be less likely to receive CABG as a result of the increased time needed for mechanical ventilation, suggesting that these patients may be better suited to receive PTCA. However, in our analyses, people with COPD were no more or less likely to have received any type of revascularization as compared with people without COPD. Although COPD is associated with early mortality and morbidity after CABG, guidelines suggest that these patients may be appropriate candidates if preoperative treatment of the pulmonary abnormality is offered (17). Furthermore, patients with congestive heart failure, cardiac arrhythmias, and valvular disease were significantly less likely to have received PTCA, but not significantly more likely to have received CABG. It would be expected that people who were not candidates for 1 particular revascularization procedure (e.g., PTCA) would be candidates for another type of revascularization (e.g., CABG) based on evidence that suggests that revascularization is associated with better prognosis as compared with medical management. It is unclear why people with these specific comorbidities were not more likely to have received CABG, although we hypothesize that other specific medical comorbidity (e.g., renal failure) may have been an important factor in deciding if any revascularization was warranted. Regardless, overall increased medical comorbidity was associated with greater likelihood of CABG and decreased odds of PTCA in this population.
The increased rate of receipt of PTCA among subjects with substance disorders in this study was unexpected, particularly because noncompliance with risk factor reduction (i.e., smoking cessation) after revascularization may be problematic in this population. Tobacco abuse is a significant risk factor for MI, and patients with MHD are 2 times more likely to smoke as compared with patients without MHD (32,33). It is imperative that people who smoke receive smoking cessation counseling to lessen the likelihood of future complications that may result from revascularization. Smoking cessation results in less recurrent angina, improved physical functioning, fewer admissions, and improved survival after CABG (17). High rates of smoking may have accounted for some of the increased rates of hospitalization in the population of subjects with MHD in this study. Conversely, substance abuse may have been ascertained for subjects receiving revascularization as a result of the detailed history and examination, resulting in a detection bias.
The results of our study are both similar and dissimilar to previous studies (15,16,20,21). Appendix 1 gives an overview of the study populations analyzed in prior studies as compared with our study and the general results of these studies, which may be helpful in the understanding of similarities and differences of study findings. In contrast to our study, 3 of the 4 previous studies conducted have reported that subjects with any MHD were statistically less likely to have received revascularization (15,16,21). However, 1 of these studies did not adjust for potential confounders and was not limited to patients hospitalized for MI, which may limit the comparison of results between studies (20). Our results likely differ from results from these studies because of differences in demographics, degree of medical comorbidity and cardiovascular risk factors, and the classification of MHD. However, the use of administrative claims data, as opposed to medical records, does not invalidate the results of our study based on evidence from Duke University using medical records. That study has also reported that older age and people with a greater burden of medical comorbidity such as diabetes mellitus are associated with higher rates of CABG but lower rates of PTCA (34). Finally, differences in results may be attributable to small sample size in our study, which may have contributed to low statistical power. However, the CIs reported in Table 2 were fairly narrow. Regardless, a further description of similarities and differences between our study and the previously conducted studies of patients with MHD is warranted.
First, age is an important predictor of the incidence of MI. However, age, in combination with other factors, influences whether revascularization should be offered (17,18). As shown in Appendix 1, the mean age of our study population was approximately 54 years, whereas the other studies focused on older (>65 years) populations. Our results are generalizable to younger populations with other similar characteristics, whereas the previously mentioned studies may be better generalizable to older populations. Given that our population was younger, revascularization rates may have been higher than those reported in other studies of subjects with MHD, especially given that family history of early onset of MI is a strong predictor of early-onset MI (35). Compared with the results for the population of people less than 65 years of age (40.5% of the sample) in the Young study, subjects in our study were more likely to receive PTCA with the exception of people with psychotic disorders, but less likely to receive CABG (21). The ORs in the Young study and those in our study were not similar, in general. This suggests that underlying differences in the populations contributed to varying receipt rates of PTCA and CABG. Geographic differences have been observed in differences in revascularization receipt in other studies and may be a contributing factor to the differences between the 2 studies given that our study population only includes Iowa residents, whereas the Young study was a national sample of people who were hospitalized (36).
Second, differences in gender may have influenced disparity in rates of revascularization, particularly because older women with cardiac comorbidity are more likely to undergo PTCA (18). As shown in Appendix 1, our study was comprised of significantly fewer women (23.8%) than the percentage of women included in other studies. Women in our study were more than 30% less likely to have received any type of revascularization, in contrast to the results reported in the Western Australia study for which no differences in receipt of revascularization were found for women with MHD as compared with women without MHD (16). Women may have been less likely to have received CABG in our study given that they are reported to have increased morbidity and mortality after CABG, possibly as a result of the presence of comorbid conditions. However, this does not explain the decreased likelihood of PTCA among these women, especially because the acute and long-term outcomes of PTCA are similar for both men and women (17,18). Our results are consistent with the lack of uniformity in reports of revascularization receipt among women (37–39). In general, with regard to gender, our results can be generalized to primarily younger populations of commercially insured men.
Third, people with MHD in our study were more likely to have increased overall medical comorbidity, including cardiovascular risk factors, in contrast to the medical comorbidity reported in previous studies of people with MHD (15,20). However, the increased medical comorbidity among people with MHD in our study did not translate into overall higher rates of revascularization procedures in this population. This finding merits further attention in future studies. It is unknown if the same findings are true of other studies given that results regarding receipt of revascularization for people with specific comorbidities were not presented.
Finally, differences in the classification of MHD may have influenced the results. We classified a subject into a single DSM-IV category based on the most clinically prominent MHD for a subject's entire claims history during 1996 to 2001, similar to the method used in the Petersen study (20). Druss classified people with a MHD only if the MHD was deemed to be "current and ongoing" at the time of the MI hospitalization (15). Although we cannot verify that our population of people with mental illness truly had a "current and ongoing" MHD, more than half of this population sought care for their MHD on at least 2 occasions before the MI hospitalization. More severe MHDs are likely to be "current and ongoing" illnesses, which further helps to ensure that our classification system is adequate. Furthermore, we chose to include people first diagnosed with a MHD in the 30-day period after discharge given that these people may be at risk for lack of revascularization, possibly as a result of a psychosis or intoxication. This method was not used in other studies. Regardless, without access to medical records, our study and the other studies using administrative data may only be generalizable to patients with well-documented MHD who use medical services and have such information recorded. Future studies will need to obtain more detailed clinical information to validate the results reported.
Previous studies reported that people with substance disorders were less likely to have received revascularization procedures as compared with people without MHD (15,16,20). Contrary to those results, we found that people with substance disorders were more likely to have received PTCA, but not CABG, which was an unexpected result, particularly because noncompliance with risk factor reduction (i.e., smoking cessation) after revascularization may be problematic in this population. Tobacco abuse is a significant risk factor for MI, and patients with MHD are 2 times more likely to smoke as compared with patients without MHD (32,33). It is imperative that people who smoke receive smoking cessation counseling to lessen the likelihood of future complications that may result from revascularization. Smoking cessation results in less recurrent angina, improved physical functioning, fewer admissions, and improved survival after CABG (17). High rates of smoking may have accounted for some of the increased rates of hospitalization in the population of subjects with MHD in this study. Conversely, substance abuse may have been ascertained for subjects receiving revascularization as a result of the detailed history and examination, resulting in a detection bias. One possible explanation for why people with substance abuse in our study may have been more likely to receive revascularization may be attributable to the demographic characteristics, as previously mentioned, of our population as compared with that of the other studies conducted. It has been reported that substance disorders are often undetected in the elderly, which may help to explain why people with substance disorders in our younger population were more likely to have received PTCA (40).
Although we did not find that people with psychotic disorders in our study were statistically less likely to have received revascularization, 2 prior studies have shown that people with psychotic disorders may be less likely to receive PTCA or CABG (15,16). This is an important finding and suggests that this population may be at risk for lack of appropriate cardiac care because of an inability to give informed consent, a misdiagnosed delirium, or an inappropriate bias by physicians against the severely mentally ill. We conducted analyses (data not shown) for subjects who were first diagnosed with a mental disorder during the hospitalization given that this should indicate that the physicians were aware of the mental disorder diagnosis and can help ascertain whether physician bias may have been a potential reason for underuse of revascularization for people with MHD. In general, results were similar to those reported, although subjects with cognitive disorders or psychosis were much less likely to have received revascularization when the cognitive disorder was first diagnosed during the MI hospitalization as opposed to when it was diagnosed before the hospitalization. This may indicate that informed consent could not be obtained or that physicians may differentially offer revascularization to patients without mental illness. This serves to promote future research to better understand why people with severe mental illness are less likely to receive services. This would help in the development of improved systems of medical care that can better serve this population.
Study Strengths
Our study analyzed a large population-based sample of younger adults with commercial insurance. We believe this is the first study to analyze revascularization among adults with and without mental illness in this type of population. We examined 6 years of claims data, with a median follow-up period of 26 months for subjects with MHD and 16 months for subjects without. This difference in follow up may have contributed to more opportunities for comorbidities to be coded in the claims data for subjects with MHD. We do not believe this would differentially affect the 2 groups, however, given that all subjects were hospitalized for the MI, and any comorbidity not previously coded in outpatient claims data would likely have been coded during the inpatient stay. Furthermore, unlike other studies that have examined only primary MI diagnoses, we included all subjects with a hospitalization for MI occurring in any diagnostic position. This enabled us to include subjects who were diagnosed with a MI while hospitalized for other medical causes. The use of rigorous case-finding methodology further ensures accuracy and generalizability of these findings. One study has reported that the positive predictive value of acute MI and revascularization procedures in administrative claims data are high (>90%), further supporting the use of this type of data for these analyses (41). Approximately 25% of our study population did not undergo revascularization during the index hospitalization or within 30 days of discharge, similar to the proportion of subjects that did not receive revascularization in a study using medical records (39).
Study Limitations
The nature of administrative claims data are such that important clinical information that might influence decision-making is not available. For instance, although codes for renal failure and congestive heart failure are present, information regarding creatinine level and left ventricular ejection fraction was not available. We were unable to control for the actual acuity level, involvement of the left main coronary artery, the number of coronary arteries involved, or the number of skilled interventionalists at given practice locations. We have no reason to believe that these factors differentially affected people with MHD other than the increased risk factor burdens discussed previously.
This study included insured adults from Iowa. Because Iowa is a racially homogeneous state, our results are generalizable to similar insured populations. Limitations inherent to the analysis of claims data may have affected the reported findings. A misclassification bias may have occurred if subjects lacked supporting MHD claims. Subjects with claims for MHD occurring more than 30 days after MI discharge were excluded (n = 502). However, given that the median number of days after this 30-day period to the first MHD claim was 273.5 days for these subjects, we do not believe that many subjects were misclassified. We believe that we have appropriately classified subjects with MHD because the majority of these subjects had at least 2 claims for an MHD before the MI hospitalization. Other immeasurable factors such as hospital location and availability of cardiac procedures likely influence the rate of revascularization. Future studies that examine a younger population of adults with commercial insurance should obtain that type of data.
Clinical Implications
This study suggests the need for prospective cohort studies using medical record abstraction in a population of younger, commercially insured adults to determine if receipt of revascularization is different among people with and without MHD. Prospective analyses, including review of clinical data, behavioral risk factors, and ability to provide informed consent, might elucidate why some people with mental illness are less likely to receive cardiac interventions.
|
APPENDIX 1. |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
This study was funded by the National Institute of Mental Health K08 MH001932-03 "Epidemiology of Cancer and Mental Illness in Rural Areas" (Dr. Carney) and the American Psychiatric Institute for Research and Education Health Services Research Scholars Program Award 2002 (Dr. Carney). We acknowledge the support of Dr. Sheila Riggs, Vice-President for Healthcare Measurement and Reporting, Wellmark Blue Cross/Blue Shield, for allowing access to this data.
DOI:10.1097/01.psy.0000170336.87544.74
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONAPPENDIX 1.NOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Kisely, L. A. Campbell, and Y. Wang Treatment of ischaemic heart disease and stroke in individuals with psychosis under universal healthcare The British Journal of Psychiatry, December 1, 2009; 195(6): 545 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Mitchell, D. Malone, and C. C. Doebbeling Quality of medical care for people with and without comorbid mental illness and substance misuse: systematic review of comparative studies The British Journal of Psychiatry, June 1, 2009; 194(6): 491 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. E. Abrams, M. Vaughan-Sarrazin, and G. E. Rosenthal Psychiatric Comorbidity and Mortality After Acute Myocardial Infarction Circ Cardiovasc Qual Outcomes, May 1, 2009; 2(3): 213 - 220. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
