Our comparison of the performance of an HIV primary care model for the treatment of HCV among HIV-infected patients in comparison to a subspecialty hepatology model allows the following conclusions to be made: (1) In the HIV primary care model there were more patients treated for HCV and fewer were lost to follow-up; (2) HCV treatment referral rates did not differ during two study periods; (3) discontinuation rates were similar in both models despite the increased prevalence of ongoing substance use in the HIV primary care model; (4) the rate of HCV cure (SVR) was similar in both clinic models.
The idea of using an HIV primary care model for the treatment of HCV is not new [8, 14–17]. However, a novel component of our approach was the integration of HIV clinical pharmacists to enhance protocol adherence and patient safety. Undoubtedly, the role of the pharmacists is becoming more prominent since the introduction of HCV direct acting antivirals in co-infected individuals, where attention to medication interactions is particularly important .
The patient referral rate for HCV treatment did not increase during the HIV primary care model, but referred patients had less obvious clinical or laboratory signs of decompensated liver disease such as thrombocytopenia or coagulopathy (Table 2). It is not clear whether this was the result of the increased awareness among HIV providers about importance of HCV treatment in the HIV primary care model. What is known is that in the HIV primary care model, 33% of patients treated for HCV had history of ongoing substance use and 4 patients were homeless. Despite this ‘high risk population’ treated for HCV by the HIV primary care model, the proportions of HCV treatment discontinuation (either due to adverse events or loss to follow-up) and cure were similar to the hepatology model. Moreover, the treatment outcomes in the high-risk patients were similar to the non-high risk patients treated in the HIV primary care model. In particular there was no difference in the rate of treatment discontinuation due to adverse events (Table 5). We acknowledge that our samples size was small to detect as significant differences between high-risk and non-high risk treated patients. But the message to highlight is that the collaborative monitoring strategy use by the HIV primary care model allowed the HCV cure of patients that may have been rendered unfavorable HCV treatment candidates in other sub-specialty models . The finding that patients with unstable psychiatric conditions, higher HIV viral loads and lower CD4 cell counts were less likely to be treated in the HIV primary care model was in line with the Owen Clinic protocol of working prospectively with patients to link them to care and, once they are more stable and engaged, initiate HCV therapy . This contrasted with the lack of association with any of the studied variables to predict HCV therapy initiation in the hepatology model.
The present study found no significant differences in the rates of HCV SVR and discontinuation of HCV therapy due to adverse events, however, the SVR trend was greater in patients treated on the HIV primary care clinic (44 vs. 35%) despite having almost double discontinuations due to adverse events of patients treated by the hepatology clinic (29 vs. 16%), perhaps due our lack of sample size to detect those differences as significant (e.g. current sample size has 10% power to detect a difference of 9% in SVR as significant). In the absence of a difference in the proportion of patients with viral relapse after prior HCV therapy in both treatment models, we believe that this observation may be explained by unmeasured clinical factors such as: 1) potential better interleukin-28B gene polymorphism allele profile in patients treated in the HIV primary care model , since this test was not available in our institution at the time we performed study; 2) a positive effect of HIV primary care model in motivating patients while on HCV therapy that led to improved adherence and hence chances of cure in those who were able to tolerate HCV therapy . This study has important limitations. First, this was not a randomized clinical trial; rather the aim was to compare the performance of 2 clinic-based models with balanced and overall comparable study populations. Second, the HIV primary care model did not treat patients with advanced cirrhosis given that these individuals are more prone to develop severe adverse events. Thus, our findings do not apply to patients with advanced liver disease, who still stand to benefit from specialized care by hepatologists. Third, we cannot rule out that the increasing rates of HCV treatment initiation over time in the HIV primary care model could be the result of secular trends: (a) aging of cohort and more urgency for treatment, (albeit we found no difference in liver fibrosis scores in both cohorts when liver biopsies were performed); (b) concurrent emphasis in the literature of epidemiologic evidence that liver disease is one of the leading causes of death in HIV-infected people, which may have motivated more treatment . It could be argued that HCV treatment uptake rate in the HIV primary care model was not ideal (25%). However, our reported HCV-treatment rates are higher than a recently reported aggressive program to engage HIV/HCV co-infected patients in care sponsored by the United States National Institutes of Health  and similar to many European countries, despite the fact we had considerably less accrual time than other clinic-based studies . We believe that our study results underestimate the positive impact of the HIV primary care model in the HCV treatment uptake rate, since 23 patients in this model were excluded from the analysis because: (1) Three patients were still receiving HCV therapy and therefore we could not assess their final outcomes; (2) twenty patients who were staged and eligible for conventional HCV therapy within study period elected to wait for HCV protease inhibitors availability and initiated HCV treatment right before or after 30 July 2011, (Figure 1).
The fact that 10% of patients in both cohorts presented with advanced cirrhosis and were not eligible for HCV therapy highlights the importance of reducing disparities in access to HCV care in the HIV population [3, 25]. New HCV therapies offer higher chances of cure, simpler and hopefully less toxic regimens [26, 27]. However, to scale up HCV treatment among the HIV infected population, we will need an inclusive collaborative approach that reduces the negative referral bias of physicians when making HCV treatment decisions in vulnerable populations with ongoing barriers to care [22, 28, 29]. We believe that the HIV primary care model could be useful in other settings and countries burdened by the high prevalence of HCV and difficult to treat urban, poor, marginalized populations that require both more efficacious HCV therapies and newer collaborative models of care such as the one described here .
In conclusion, in this exploratory analysis, the use of an HIV primary care model supported by pharmacists specialized in HIV care increased the number of patients who initiate and successfully finish HCV therapy with comparable virological outcomes to a subspecialty hepatology model, highlighting the importance of increasing the absolute number of HIV-infected patients treated for HCV at any given time.