MMP is an annual cross-sectional survey designed to produce nationally representative estimates of behavioral and clinical characteristics of adults with diagnosed HIV in the United States. MMP data collection is a part of routine public health surveillance, and was thus determined to be nonresearch (Centers for Disease Control and Prevention 2010c). Participating states or territories obtained local institutional review board approval to collect data, when required. Informed consent was obtained from all participants.

An estimated 4.9% (CI 4.4–5.4) of adults with diagnosed HIV had documented active HCV infection during the past 24 months (data not displayed on a table). An estimated 80.2% of persons with HIV/HCV coinfection were male (Table 1), 71.6% were born from 1945–1965, 45.2% were black/African American non-Hispanic, 41.9% had education beyond high school, 44.9% were men who had sex with women only, and 30.2% were MSM. An estimated 75.0% had public insurance only, 60.3% were living below the federal poverty level, 11.8% reported having been homeless in the past 12 months, and 8.2% reported having been incarcerated in the past 12 months. Most (92.4%), had been prescribed antiretroviral treatment, 80.7% had an HIV viral load that was undetectable or <200 copies/mL at last testing, and 52.6% had a mean CD4+ T-lymphocyte cell (CD4) count ≥500 cells/mm³ in the past 12 months. There was evidence of advanced liver fibrosis/cirrhosis in 14.1% who had a FIB-4 score of >3.25 and 10.5% who had an APRI score >1.5.

Overall, 27.0% of persons with HIV/HCV coinfection reported engaging in condomless anal or vaginal sex in the past 12 months (Table 2). Among MSM with coinfection, 45.5% reported condomless anal sex in the past 12 months. Regarding substance use, an estimated 17.5% reported drinking alcohol before or during sex, 10.2% reported binge drinking alcohol (defined as men who drank ≥5 or women who drank ≥4 alcoholic beverages on a single occasion) in the past 30 days, 30.1% reported non-injection drug use, 30.5% reported non-injection drug use before or during sex, 6.7% reported injection drug use, and 6.4% reported injection drug use before or during sex in the past 12 months.

An estimated 73.1% of persons with HIV/HCV coinfection reported receiving at least one prevention service in the past 12 months designed to protect themselves and their partners from HIV and sexually transmitted infections, including receiving informational materials, or having a conversation with an outreach worker, counselor, or prevention worker, with a doctor, nurse, or other healthcare worker, or in an organized small group (Table 3). Among those reporting any anal or vaginal sex, 67.0% reported receiving free condoms in the past 12 months. Among MSM, 42.3% reported receiving free condoms in the past 12 months. Among those reporting injection or non-injection drug use, an estimated 30.8% reported receiving drug or alcohol counseling in the past 12 months. Among PWID, 79.2% received sterile syringes from a needle exchange, pharmacy, or health care facility.

HCV treatment was prescribed during the past 24 months for an estimated 62.1% of persons with HIV/HCV coinfection (Table 4). Prescription of HCV treatment was associated with being prescribed HIV antiretroviral therapy in the past 12 months, having less advanced liver fibrosis, and not using non-injection drugs before or during sex (Supplementary Tables 1, 2). Prescription of HCV treatment was not associated with being born during 1945–1965, race/ethnicity, gender, education, type of health insurance, reported incarceration or homelessness, sexual transmission risk behaviors, or receipt of prevention services (Supplementary Tables 1–3).

An estimated 4.9% of PWHIV had recent medical record documentation of active HCV coinfection. We found substantial missed opportunities in this population to prevent HCV transmission, decrease HCV-related morbidity, and reduce health disparities. Ending HCV will require effective interventions to address substance use disorders, reduce sexual transmission risk behaviors among MSM, and remove barriers to HCV treatment for persons with HIV/HCV coinfection.

Our estimate that nearly 5% of PWHIV had documented active HCV infection in the past 24 months compares with 1% of non-institutionalized US civilians as reported by The National Health and Nutrition Survey (NHANES) between 2013 and 2016 (Hofmeister, 2018). NHANES did not report estimates of HIV/HCV coinfection during this time period. Previous information about HCV infection among PWHIV in the United States is limited. In 2009, an estimated 21% of HIV patients who were tested for past or present HCV infection tested positive (Garg, Brooks, Luo, Skarbinski, 2014). However, prevalence varies widely by subgroup: 1–12% among MSM, 9–27% among heterosexuals, and 72–95% among PWID (Alter, 2006). HIV cohort studies have reported HCV viremia among approximately 12% of cohort members (Cacahy, 2014; Vellozzi, 2011). Several factors could explain why our estimate of active HCV coinfection is lower than these earlier reports. First, our data describe HCV viremia in the era of DAAs, during which some persons with coinfection would presumably have been cured prior to our 2-year observation period. HIV cohort data were obtained prior to or early after the introduction of DAAs. Second, we recorded HCV lab data from the most recent 2-year period. HIV cohorts recorded viremia at any time since cohort enrollment. Third, previous studies may have included a higher percentage of PWID.

There are similarities in the sociodemographic characteristics of the HIV/HCV coinfected population and the overall HCV-infected population in the United States. We estimated that persons born from 1945–1965 accounted for 72% of persons with coinfection, which is comparable to previous estimates among all persons with HCV-infection in the United States (Smith, 2012; Hofmeister, 2018). Males accounted for 80% of persons with coinfection compared with 64% of all persons with HCV infection. However, compared with persons with HCV overall, double the percentages of persons with coinfection were living in poverty (60% vs. 29%) and were black (45% vs. 25%) (Hofmeister, 2018). The increased risk of poor outcomes among persons with coinfection compared with HCV monoinfected persons, including end-stage liver disease, hepatocellular carcinoma, and death (Rein et al 2011), compounds the burden of these income and racial disparities. To reduce the risk of morbidity and mortality from these complications and to reduce health disparities, access to curative treatment for persons with coinfection is urgently needed.

Injection drug use remains the leading cause of HCV transmission and the incidence of acute HCV infection among young PWID has been increasing since the mid-2000s, especially in nonurban settings including Appalachia (Suryaprasad, 2014; Zibebell, 2015). However, we found that only 1 in 13 persons with HIV reported injection drug use in the past 12 months and, although our relatively small sample size did not permit us to generate a stable estimate of the percentage of PWID who shared needles or other injecting equipment, 79% of those who injected reported receiving sterile syringes. Although we did not ascertain the percentage that used drugs intranasally, this route of drug use is considered a risk factor for HCV transmission and warrants HCV screening per United States Preventative Services Task Force recommendations (Moyer, 2013). We found that among persons with HIV/HCV coinfection, 31% reported non-injection drug use in the past 12 months. Receipt of substance use counseling among persons who used drugs was low. Given the increasing number of reported acute HCV cases among PWID (Suryaprasad, 2014; Zibebell, 2015), access to substance use services and sterile injecting equipment for all PWID is vital.

Although sexual transmission of HCV is generally inefficient (Terrault, 2013), sexual transmission among MSM with HIV has been identified as an emerging challenge for HCV prevention (Suryaprasad, 2014; Zibbell, 2015; Hagan, 2015; Fierer, 2011; Kaplan-Lewis & Fierer, 2015). Two US studies have reported an HCV incidence rate of 0.21–0.51 per 100 person-years among MSM with HIV (Witt, 2013; Taylor, 2011). Factors associated with incident HCV infection in this population include condomless anal intercourse, antecedent syphilis, gonorrhea, or chlamydial infection, history of injection drug use, having sex with concurrent methamphetamine use, and douching prior to anal intercourse (Fierer, 2011; Witt, 2013; Apers, 2015). We found that among MSM with HIV/HCV coinfection, an estimated 46% reported engaging in condomless anal sex in the past 12 months. A case-control study in New York City found that condomless insertive and receptive anal sex increase the odds of becoming HCV-infected 8-fold and 25-fold, respectively among HIV-positive MSM (Fierer, 2011). The high prevalence of condomless anal sex among MSM with HIV underscores the importance of testing this population for HCV and counseling MSM with HIV/HCV coinfection about sexual risk reduction to prevent HCV transmission to their partners. Providers must explain to HIV patients that highly effective biomedical strategies for reducing HIV transmission, including HIV treatment as prevention and PrEP, have no biologic activity against HCV and cannot be relied upon to prevent HCV transmission.

Despite widely disseminated recommendations to treat persons with HIV/HCV coinfection for HCV (American Association for the Study of Liver Disease-Infectious Disease Society of America 2018), more than one-third were not prescribed HCV treatment during the period following the introduction of DAAs. We found that a higher percentage of untreated persons reported risky drug use suggesting that providers may be deferring treating patients with active drug use or that Medicaid sobriety requirements in some states may be a barrier to prescribing treatment (Canary, Klevens, Holmberg, 2015; Barua, Greenwald, Grebely, Dore, Swan, Taylor, 2015), Despite treatment eligibility requirements of some payers that patients must have more advanced liver fibrosis, treated patients had lower fibrosis scores suggesting that patients with more advanced fibrosis may be less likely to receive treatment. Although HCV monoinfected patients face Medicaid restrictions and cumbersome pre-authorization requirements limiting access to treatment in some states, we did not find differences in prescription of HCV treatment for coinfected persons based on insurance type. Further research is needed to identify health system factors that could explain why some patients are treated and others are not, including facility type, availability of onsite supportive services designed to increase care engagement, and the use of decision support including clinical reminders in electronic health records.

Although nearly all persons with HIV/HCV coinfection can be cured with DAAs (Wyles, 2015; Naggie, 2015), reinfection rates are high. Incidence of reinfection of MSM with HIV ranges from 2.9–15.2 per 100 persons-years and 2-year cumulative rates of reinfection ranging from 25–33% (Ingiliz, 2016; Lambers, 201; Chaillon, 2017) Incidence of reinfection of PWID ranges from 0–5.3 per 100 person-years for those following completion of interferon-based treatment to as high as 28.8 per 100 person-years among a prospective cohort of active users (Grady, Schinkel, Thomas, Dalgard, 2013, Weir, 2016; Midgard, 2016, Aspinall, 2013; Sacks-Davis, 2013). The high reinfection risk among persons with HIV/HCV coinfection highlights the challenges of applying a treatment as prevention strategy for HCV without concurrently addressing ongoing sexual and drug use risk behaviors. Theoretical models suggest that scaling up treatment with DAAs for PWID coupled with opioid substitution therapy and access to needle and syringe services could reduce HCV transmission and achieve reductions in prevalence of >50% over 10 years (Martin, Hickman, Hutchinson, Goldberg, Vickerman, 2013; Hickman, De Angelis, Vickerman, Hutchinson, Martin, 2015; Hellard, 2015).

Our analysis is subject to the following limitations. First, our definition of HCV coinfection as documentation of detectable RNA during the study period allowed us to conservatively estimate active HCV coinfection among persons who were tested during the study period, but did not capture persons with coinfection who did not have an HCV RNA test during that period. Also, because of uncertainty about the HCV status of persons not tested for HCV RNA during the study period, we were unable to confidently classify persons as HIV monoinfected for comparison with coinfected persons across sociodemographic characteristics and risk behaviors. Second, we were unable to evaluate the temporality of HIV and HCV infections or identify incident HCV infection because of the cross-sectional design of MMP. Third, since we lack longitudinal data on respondents, we were unable to evaluate treatment outcomes, or cases of possible reinfection in this population. Fourth, we were unable to evaluate the infection status and risk behaviors of the sexual or drug using partners of our population. Fifth, since data collected on risk behaviors relied on interview questions, reported risk behaviors may be underestimated due to socially desirable responding. Finally, our sampling and weighting design did not allow geographic comparisons.

Recent injection drug use is relatively uncommon among persons with coinfection and most who injected received sterile syringes. However, nearly half of MSM with coinfection reported sexual behaviors placing their partners at high risk of acquiring HCV. Despite recommendations to treat all persons with HCV, over one-third of persons with HIV/HCV coinfection were not prescribed curative HCV treatment. Health care providers should evaluate persons with HIV/HCV coinfection for transmission risk behaviors, link them to appropriate prevention services to reduce HCV transmission, and prescribe treatment for HCV. Improving access to HCV prevention services and treatment for persons with HIV/HCV coinfection is essential for achieving HCV elimination goals and reducing health disparities.