Hepatitis C Virus Testing in Adults Living with HIV: A Need for Improved Screening Efforts

Objectives Guidelines recommend hepatitis C virus (HCV) screening for all people living with HIV (PLWH). Understanding HCV testing practices may improve compliance with guidelines and can help identify areas for future intervention. Methods We evaluated HCV screening and unnecessary repeat HCV testing in 8,590 PLWH initiating care at 12 U.S. HIV clinics between 2006 and 2010, with follow-up through 2011. Multivariable logistic regression examined the association between patient factors and the outcomes: HCV screening (≥1 HCV antibody tests during the study period) and unnecessary repeat HCV testing (≥1 HCV antibody tests in patients with a prior positive test result). Results Overall, 82% of patients were screened for HCV, 18% of those screened were HCV antibody-positive, and 40% of HCV antibody-positive patients had unnecessary repeat HCV testing. The likelihood of being screened for HCV increased as the number of outpatient visits rose (adjusted odds ratio 1.02, 95% confidence interval 1.01–1.03). Compared to men who have sex with men (MSM), patients with injection drug use (IDU) were less likely to be screened for HCV (0.63, 0.52–0.78); while individuals with Medicaid were more likely to be screened than those with private insurance (1.30, 1.04–1.62). Patients with heterosexual (1.78, 1.20–2.65) and IDU (1.58, 1.06–2.34) risk compared to MSM, and those with higher numbers of outpatient (1.03, 1.01–1.04) and inpatient (1.09, 1.01–1.19) visits were at greatest risk of unnecessary HCV testing. Conclusions Additional efforts to improve compliance with HCV testing guidelines are needed. Leveraging health information technology may increase HCV screening and reduce unnecessary testing.


Introduction
Up to 25% of people living with HIV (PLWH) in the United States (U.S.) are coinfected with hepatitis C virus (HCV). [1] HIV/HCV coinfection accelerates progression to liver fibrosis, end-stage liver disease, and death compared to HCV monoinfection. [2] Consequently, guidelines recommend that all PLWH be screened for HCV infection upon initiation of HIV care [3][4][5][6].
Despite these recommendations, HCV screening rates have varied for PLWH. [7][8][9][10] Among a random sample of 1,329 HIVinfected men who have sex with men (MSM) in care at 8 U.S. HIV clinics between 2004 and 2007, only 54% were ever tested for HCV. [7] Conversely, the U.S. Veterans Health Administration (VA) reported that 96% of the 23,463 HIV-infected veterans in care in 2008 received HCV screening [9].
Screening for HCV is performed by testing for HCV antibody; positive results should be followed by measurement of HCV RNA to differentiate chronic infection from resolved infection. [11] HCV antibody testing in individuals with a prior positive result is unnecessary, as the antibody test remains reactive regardless of whether the infection has cleared. [11] However, PLWH may be at increased risk of unnecessary HCV testing due to the high prevalence of HCV infection and their more frequent use of medical services compared to the general population [3].
Understanding HCV testing practices is important to improving compliance with national guidelines and can help identify areas for future intervention. The goal of this study was to estimate the proportion of PLWH screened for HCV and to identify groups at risk of not being screened. We also assessed unnecessary repeat HCV testing among patients with a prior positive HCV antibody test.

Methods
The HIV Research Network (HIVRN) is a consortium of clinics that provide care to people living with HIV. [12,13] Data from 12 sites treating adults, located in the Northeastern (6), Midwestern (1), Southern (2), and Western (3) U.S., were included in this analysis. All patients establishing care at these sites are offered enrollment in the HIVRN; 99% of patients consent to participate. Sites abstract data from medical records and send them to a datacoordinating center after removing personal identifying information. Following quality control and verification, data are combined across sites to produce a uniform database. Institutional review boards (IRBs) at each site (complete list of sites can be found in the acknowledgments) and the data coordinating center at Johns Hopkins University approved the collection and analyses of these data. IRBs at some clinics required written informed consents, while others waived the requirement because only existing anonymized and de-identified data were collected.
HIV-infected adults (age $18 years) new to care at HIVRN sites and seen at least twice between January 1, 2006 and December 31, 2010 were included and followed through December 31, 2011. The two-visit criterion was employed to capture patients establishing regular care. To exclude those who may have received care previously, we excluded patients with first recorded HIV RNA #400 copies/mL (n = 2,665) and individuals with a history of outpatient HIV visits or use of antiretroviral therapy (ART) prior to HIVRN enrollment (n = 916).

Variables
Demographic and clinical characteristics (age, sex, race/ ethnicity, HIV transmission risk, insurance coverage, CD4 count) at the time of the first outpatient visit were collected for each patient. Patients' age was divided into 4 groups: 18-29, 30-39, 40-49, and over 50 years old. Race/ethnicity was categorized as non-Hispanic white, non-Hispanic black, Hispanic, and other/ unknown. Self-reported HIV transmission behavior categories were MSM, heterosexual transmission (HET), injection drug use (IDU), and other/unknown. Patients who had IDU in combination with another risk factor (e.g. MSM, heterosexual transmission) were classified as IDU. Insurance was categorized as private, Medicaid, Medicare, uninsured, or other/unknown. Patients whose care was funded by Ryan White, those recorded as selfpay, and those covered by local governmental programs were considered to be uninsured. CD4 cell count was categorized as # 350, 351-500, .500 cells/mm 3 , or missing.
The numbers of outpatient, inpatient, and emergency department visits over the entire study period were collected for each patient. Outpatient visits refer only to primary HIV care appointments made to an HIVRN clinic, and do not include nursing or pharmacy visits, consultations, or other types of appointments.
Dichotomous outcomes were (1) performance of any HCV screening and (2) unnecessary repeat HCV testing, during the observation period. We defined the observation period as the time from date of enrollment to date of last outpatient visit, transfer of care, or death, whichever occurred first. Patients were considered screened for HCV if they had one or more HCV antibody test(s) during the observation period. Unnecessary repeat HCV testing was defined as one or more HCV antibody tests in patients with a prior positive test result.

Statistical Analysis
We compared demographic and clinical differences in the proportion of patients screened for HCV, HCV antibody positive, and unnecessarily tested for HCV using x 2 tests. Multivariable logistic regression was used to estimate the association between patient factors and the two outcomes. Analyses of unnecessary repeat HCV testing were restricted to patients with a prior positive HCV test, for whom unnecessary testing could be measured. In secondary analyses, multivariable logistic regression was used to estimate the association between patient factors and repeat HCV testing (.1 HCV antibody tests) among HCV antibody negative patients. All regression models included indicators for each HIVRN site and length of observation period. Analyses were conducted in STATA 12.1 (College Station, TX). Table 1 presents descriptive information for the 8,590 patients new to care at the 12 HIVRN sites in 2006-2010. Median age at presentation was 39 years, with 16% aged 50 years or older. The sample comprised high proportions of patients who were male (75%), of minority race/ethnicity (71%), and who had either no health insurance or Medicaid at the time of their first outpatient visit (66%). The most common HIV transmission behavior was MSM (46%), followed by HET (37%) and IDU (11%). Excluding patients with missing data (N = 607), the mean first CD4 cell count was 318 cells/mm 3 .
Among patients screened for HCV, 1,283 (18% of 7,023) had a positive test result. ( Table 2) The percentage of patients identified as HCV antibody positive varied (P,0.05) by age, sex, and HIV transmission behavior. Persons with public insurance and greater numbers of outpatient and inpatient visits had higher proportions of HCV antibody positivity than those with private insurance and lower numbers of outpatient and inpatient visits, respectively (P, 0.05).
Of patients with a positive HCV antibody, 510 (40% of 1,283) had an unnecessary repeat HCV test. ( Table 2) Sixty-five percent had one unnecessary test, 19% had two, 10%, had three, and 6% had four or more unnecessary tests. Unnecessary testing varied by HIV risk behavior, with 35% of MSM, 46% of HET, and 39% of IDU undergoing repeat testing (P,0.05). (Figure 1) Individuals with greater numbers of outpatient, inpatient, and emergency department visits had higher proportions of unnecessary testing than those with lower numbers of outpatient, inpatient, and emergency department visits, respectively (P,0.05) (Figure 2).
We identified 5,740 patients who were HCV antibody negative throughout the observation period. Of these, 2,447 (43% of 5,740) had a repeat HCV test. Fifty-five percent had one repeat test, 24% had two, 11%, had three, and 10% had four or more repeat tests. In multivariate analyses, the likelihood of having a repeat HCV test increased as the number of outpatient visits (AOR = 1.03, 95%

Discussion
Despite recommendations that all HIV-infected patients be screened for HCV infection upon initiation of HIV care, [3][4][5][6] only 82% of patients were screened. This screening rate is higher than the 16-54% found in some prior studies but is lower than the VA (96%). [7][8][9][10] The VA has prioritized HCV screening, [14] as veterans have a higher prevalence of HCV infection compared to the general population. [15] The VA also uses an integrated electronic medical record system that is easily accessible by practitioners at multiple sites and generates preventive care and screening reminders. This combination of prioritized HCV screening and sophisticated health information technology may serve as a model for other HIV clinics. In addition, alerting patients and providers when screenings are due, evaluating provider performance and offering feedback, delivering one-onone and group education to patients, and using patient and provider incentives have been identified as additional tools for improving compliance with screening guidelines [16][17][18][19][20].  Similar to unnecessary HCV testing rates reported among non-HIV infected individuals, [21] 40% of HIV-infected HCV antibody positive patients in this study had unnecessary duplicate HCV tests. Multiple factors could contribute to the high number of inappropriate tests, including failure to check previous test results, lack of awareness that a test has already been performed, and inadequate distribution of patient information across multiple medical record systems. [22,23] In a survey of 283 primary care physicians and residents, 17-32% reported having no reliable method to track test results, suggesting the need to develop effective and user-friendly methods for monitoring test results [23].
Computer provider order entry (CPOE) with embedded decision-support tools may provide one solution for reducing repetitive testing. [24,25] In one study, point-of-care prompts decreased redundant testing by 24%, demonstrating the feasibility of CPOE decision-support tools to reduce unnecessary testing. [24] However, given the widespread adoption of electronic health records and the U.S. government's emphasis on health information technology, many practices may already have access to these resources [26,27]. Injection drug users are at increased risk of HCV and disproportionally account for 80% of new infections and 60% of existing cases in developed countries. [28] Despite this risk, in our cohort, injection drug users were less likely to be screened for HCV than MSM. This finding may be due, in part, to the multiple comorbidities and psychosocial barriers facing many people with IDU. Providers are challenged with managing drug addiction; treating co-occurring conditions, including HIV infection and mental illness; and addressing competing needs (e.g. lack of housing, insurance, social support) during a 20-30 minute office visit. [29][30][31][32][33] Deferral of preventative care services, as typically occurs during complicated visits, and the need to review lengthy medical records to verify prior test results may lead to lower HCV screening. Conversely, the recent emergence of sexually transmitted acute HCV infection may influence HCV testing in MSM. [34] Increased awareness of HCV transmission in this group by both patients and providers may have contributed to the higher screening rate [35].
In our study, greater use of health care services was significantly associated with increased HCV testing, both appropriate and unnecessary. Prior research demonstrates that the number of laboratory tests performed rises with the number of health system interactions. [29] In addition, higher health care utilization can compromise care coordination, which may lead to over treatment and testing. [36][37][38][39] Accordingly, improving care coordination and addressing the needs of high users of health care services have been identified as key targets for reducing waste in the U.S. health care system [36,[40][41][42].
The current analysis has several limitations. First, sites in our sample are not nationally representative, although they do encompass a broad geographic distribution and a variety of demographic and clinical characteristics of patients, thereby improving generalizeability compared to single-site studies. Second, our data only capture laboratory tests performed at or reported to the site of care. It is possible that patients may be tested for HCV during consultation with other physicians or during hospitalization at outside institutions. Third, some patients may have been tested for HCV prior to entering HIV care. If the results of these test(s) are accurately documented and reported, some providers may not pursue HCV testing. Lastly, our analyses did not focus on the cost of unnecessary HCV testing. Additional studies are needed to estimate how repeat HCV testing impacts health expenditures. These studies should not only consider the added cost of HCV antibody testing but also potential downstream costs (e.g. HCV viral load testing, genotype testing, radiologic studies, and provider referrals).
In summary, 82% of HIV-infected patients were screened for HCV, 18% of those screened were HCV antibody-positive, and 40% of HCV antibody-positive patients had unnecessary repeat HCV testing. Additional efforts are needed to increase HCV screening, particularly for PLWH with IDU risk. In low prevalence populations, serum-pooling strategies may represent cost-effective techniques to improve screening. [43,44] Leveraging health information technology may be useful for both increasing screening and reducing unnecessary testing.

Supporting Information
Appendix S1 Proportion of HCV Antibody Negative Patients with Repeat HCV Testing and Factors Associated with Repeat HCV Testing. (DOC)