In an analysis of US data from 1996, researchers have found poorer levels of antiretroviral adherence in people of non-white race. While there were links between adherence and discriminatory healthcare experiences, greater distrust of healthcare providers, and strength of belief in treatment outcomes, these factors were not sufficient to explain the racial differences in adherence levels. The study was reported in the Journal of Acquired Immune Deficiency Syndromes.
Race is often associated with other factors, such as depression and health literacy, which can impede adherence. Yet studies have produced inconsistent conclusions as to whether racial/ethnic status per se has a significant, independent effect after controlling for these other factors.
Since it is also well-established that better personal relationships with healthcare providers are associated with better adherence, the researchers in this study investigated the links between race, adherence, trust in healthcare providers, and experiences of discrimination within the healthcare system.
This study was a secondary analysis of data from the HIV Cost and Services Utilization Study (HCSUS), which collected data on a representative sample of 2864 HIV-positive people across the US (urban and rural) between 1996 and 1997. The present study used data from 1911 HCSUS participants who had completed three study interviews and were taking combination therapy by the time of the last interview. Of these, 993 were non-Hispanic whites, and 918 were identified as "racial/ethnic minorities" (participants of all other backgrounds). The "minority" participants were more likely to be female, heterosexual, have less social support, have less than a high school degree, earn under $25,000, have no insurance, have a nadir (lowest-ever) CD4 cell count below 200 cells/mm3, and have used heroin or cocaine in the past year.
Importantly, reported "discriminatory experiences" were not necessarily racially based, but were based on patient reports of unfair treatment, including hostility, disrespect, discomfort, or refusal of service since they had become HIV-positive. "Provider trust" was based on measures of expectation that providers would act in the patient's best interest, including offering high-quality care, providing information, respecting confidentiality, and being non-judgmental. "Beliefs about antiretroviral therapy" summarised attitudes about treatment efficacy, difficulty accessing and/or scheduling medication, and psychological medication burden.
Results
Overall, minority participants were less likely to report "perfect" adherence than non-minority participants (40.3% vs. 45.3%, p < 0.001). In unadjusted bivariate analysis, poorer adherence was associated with minority status, female sex, younger age, less than high school education, lower income, having Medicaid, being heterosexual, exposure to HIV through injecting drug use or heterosexual contact, higher viral load, depression, and heroin or cocaine use (p < 0.001 for all).
The research team then used a technique known as structural equation modelling (SEM) to attempt to fit these variables into an interrelated network that allows for "mediating factors". In other words, chains of association were developed in which, for instance, minority status might significantly predict healthcare provider distrust, which in turn could predict poorer medication efficacy beliefs, finally influencing adherence. The final model described a network of eight such variables, beginning with race and ending with adherence.
Several statistically significant associations were identified, but other expected associations were not found, or were counter to expectations. For instance, more than a third (41%) of the participants reported discriminatory experiences with the healthcare system. However, minorities were actually less likely to report discrimination than white people (p < 0.001). Minority participants were more sceptical of treatment efficacy (p < 0.001) and reported more difficulty accessing medication (p < 0.001) and psychological medication burden (p < 0.001). Higher levels of discrimination predicted greater distrust; however, the level of trust in providers did not vary between white people and minorities.
Interpretation
Overall, the model could not explain the effects of race on adherence by way of intermediate (mediating) factors, as had been expected. The unexpected finding of less reported discrimination by minority participants was not satisfactorily explained.
There were a number of caveats in this analysis. First, the data are twelve years old and reflect the onset of combination therapy rather than present-day standards of antiretroviral therapy. Perceived "discriminatory experiences" could have been based on many factors, including HIV status itself, rather than race. The SEM model used to explore relations between variables "does not allow researchers to identify a 'true' model… rather, multiple models may fit… and researchers must use their own judgment". And other potentially significant variables such as health literacy and housing stability were not included in the analysis.
Despite these rather confusing results, the researchers believe that future research should continue to explore minority status as a contributor to other multifactoral models of adherence.
Reference:
Thrasher AD et al. Discrimination, distrust, and racial/ethnic disparities in antiretroviral therapy adherence among a national sample of HIV-infected patients. J Acquir Immune Defic Syndr 49:84–93, 2008.