The vast majority of largely urban minority HIV-positive subjects in care in this study were diagnosed with non-HIV–related medical comorbidities, with prevalence rates being higher in women than in men and associated with older age and intravenous drug use as route of HIV transmission. The comorbidities found to be most prevalent (hepatitis C, pulmonary disease, high blood pressure, and high cholesterol) were largely consistent with earlier studies of HIV-positive persons.13–15
Those coinfected with hepatitis C were found to have a greater number of other non-HIV–related comorbid conditions than those without hepatitis C coinfection.
The high prevalence rates of non-HIV–related medical comorbidities found in this population are noteworthy given that persons with HIV are living longer and increasingly dying of medical problems that are no longer AIDS defining. Patients should no longer be educated solely about their HIV disease; they must also be made aware of their other medical problems in order to provide effective comprehensive care. Preventive care is an essential part of providing HIV care. This includes the promotion of behavioral change (e.g., smoking cessation, exercise, weight loss, dietary modification) as well as ensuring that patients undergo recommended screening tests.28
This is consistent with recently issued primary care guidelines for the management of HIV-positive persons.29
While this is the first study to comprehensively investigate the concordance between self-report and medical record data of non-HIV–related medical comorbidities in HIV-positive persons, three previous studies of HIV-positive persons have investigated this specifically as it relates to hepatitis C,11,30
and hepatitis B and C.31
In a sample of 182 marginally housed HIV-positive persons with comorbid hepatitis C infection, 64% self-reported having hepatitis C infection.11
In a sample of 681 HIV-positive women with comorbid hepatitis C infection enrolled in the Women's Interagency HIV Study, 77% reported knowing their hepatitis C diagnosis.30
In a sample of 970 HIV-positive patients in clinical care, the sensitivity of self-reported hepatitis B infection was 27% (correct self-report by those who do have hepatitis B infection) and the sensitivity of self-reported hepatitis C infection was 72%.31
Our finding of sensitivity of 66% for self-reported hepatitis C infection is in line with the results of these three studies and our finding of 63% for self-reported hepatitis B infection is considerably higher than in the one other study to date.
While absolute concordance rates were 67% or higher for all 15 non-HIV–related medical comorbidities, there is still much room for improvement. The most clinically relevant metric of patient awareness is the sensitivity value per non-HIV– related medical comorbidity. Fully one third (34%) of those with hepatitis C, the most prevalent comorbid medical non-HIV–related medical comorbidity, were unaware that they had this viral infection. Ideally we would like HIV-positive persons to be fully aware of all of their non-HIV–related medical comorbidities. Given this lack of adequate awareness on the part of patients, it is important to develop strategies to better integrate HIV care into the broader context of care for other non-HIV–related medical comorbidities.
Priorities need to be set to begin developing programs and interventions to increase patient awareness of non-HIV–related medical comorbidities. In establishing priorities, the following dimensions can be taken into account: (1) prevalence; (2) associated morbidity and mortality; and (3) chronicity. It is most important that patients become aware of the non-HIV–related medical comorbidities, which are: (1) most prevalent; (2) most likely to increase morbidity and mortality; and (3) require chronic disease management. Of the 15 non-HIV–related medical comorbidities examined, the 7 comorbidities that best meet these criteria are: hepatitis C, pulmonary disease, high blood pressure, high cholesterol, renal insufficiency, cardiovascular disease, and diabetes mellitus. Among these 7 non-HIV–related medical comorbidities, high cholesterol, renal insufficiency, cardiovascular disease, and hepatitis C had the lowest sensitivity and are important disease states to target in educational programs for HIV-positive persons if choices need to be made.
The predictive variables found to be related to having a greater number of comorbid conditions in multivariate analyses are older age, female gender, and intravenous drug use as route of HIV transmission, with the group of women older than 50 years strongly driving these associations. The finding that age is associated with increased non-HIV related medical comorbidity replicates previous case-control studies conducted pre-highly active antiretroviral therapy (HAART)32
While intravenous drug use is highly associated with coinfection with hepatitis C, the most prevalent non-HIV–related comorbidity in the study, we reported on post hoc
analyses that found that the hepatitis C coinfected subjects also had significantly more of the other 14 non-HIV–related medical comorbidities than did those without hepatitis C coinfection. This finding is thus likely explained by multiple aspects of intravenous drug use that negatively impact overall health status including food insecurity and nutritional deficits.34,35
This is the first study to document higher rates of non-HIV–related medical comorbidities in HIV-positive women than in men. Further research is needed to investigate whether this finding can be replicated in other HIV-positive patient populations and what the underlying reasons are for these gender differences. A direction for such research on gender disparities is provided by a recent study of over 400 severely disadvantaged HIV-positive persons in New York City, which found that women were less likely to use HIV primary care services and more likely to have an emergency department visit than were men.36
Our findings may reflect a tendency for women to wait until they are more ill to come to HIV primary care as compared to men.
This study has several limitations. First, we have taken the medical record as the gold standard for each index condition. As is the case in most chart review studies, it is likely that there are non-HIV–related medical comorbidities that were not documented in the medical records as well as the possibility that there were incorrectly documented non-HIV–related medical comorbidities. It could be that medical documentation was incomplete in our studies or that providers minimized a medical complaint and did not pursue further diagnostic evaluation for it. It would have been more complete to interview each provider regarding each individual subject to further supplement the information obtained from the chart review. Furthermore, there are reasons why a subject's self-report may differ from the medical record other than lack of knowledge of having a medical condition. These may include the subject failing to report a diagnosis due to social stigma and the subject's lack of familiarity with the terms used.
Another limitation of this study is that we did not collect information on educational level, literacy, or health literacy. These unexplored constructs may play a key role in determining subject awareness37
and we recommend that this be examined in future research. We did not undertake substudies on the subjects who were in specialized integrated care programs to analyze if there was improved awareness. Future studies should investigate whether integrated programs (e.g., hepatitis C coinfection clinics) empower patients with problem-specific health literacy, and whether this has an effect on disease outcomes. This study was conducted at only one site limiting generalizability and the potential for sampling bias exists as clinic patients who did not attend any appointments during the summer of 2005 were not eligible for participation. Finally, while absolute concordance rate, sensitivity, and positive predictive value were used to compare the accuracy of self-report across the 15 non-HIV–related medical comorbidities, all three of these values are affected by the prevalence of each condition as well as the accuracy of the chart documentation for each condition.
In order for patients to optimally access and adhere to care for a medical condition, awareness of having the condition is essential. Interventions and educational campaigns to increase the awareness of HIV-positive patients of their non-HIV–related medical comorbidities that provide them with culturally appropriate information aimed at their level of literacy and education should be developed, implemented, and evaluated. Future studies should investigate the complex relationship between multiple medical morbidities, patient awareness, and adherence to care and medication. It is important to study adherence to antiretroviral medication and HIV care in the context of the entire medication and treatment regimen of the HIV-positive patient.
In order to guide the development of educational materials, strategies, and campaigns to increase accurate patient awareness of non-HIV–related medical comorbidities, research is needed on the underlying reasons for the discordance between patient self-report and chart documentation. This study did not investigate the patient, provider, or clinic factors that are related to the level of patient awareness of non-HIV–related medical comorbidities. This is a vitally important topic for future research to address. Such research should be designed to control for the potentially confounding impact of the number of comorbid conditions on level of awareness. Guidance for such research comes from a recent study of Sohler and colleagues38
investigating the factors related to disagreement between self-report of number of HIV primary care visits and chart documentation in a sample of marginalized, HIV-infected patients in 10 sites across the United States. The factors they found to be associated with disagreement included younger age, non-Hispanic black race/ethnicity, lower education, and substance use. These factors should be investigated to see if they also predict low levels of HIV-positive patient awareness of non-HIV–related medical comorbidities.
Future research should also investigate the relationship between patient awareness of non-HIV–related medical comorbidities and adherence to medication and medical care and whether medication adherence differs for antiretroviral medication, psychotropic medication, and medication regimens for other non-HIV–related medical comorbidities in HIV-positive patients with multiple morbidities, particularly those with psychiatric and substance use disorders. As HIV-positive patients live longer and present with increasing numbers of non-HIV–related medical comorbidities that are being treated, it is imperative to educate patients about these non-HIV–related medical comorbidities in order to improve their adherence to comprehensive health care and effectively provide chronic disease management.