Sustained Impact of Task-shifting HCV Treatment to Nonspecialist Providers: 5-Year Follow Up of the ASCEND Investigation

Abstract

Background

Direct-acting antiviral (DAA) therapy for hepatitis C virus (HCV) has ushered in an era of short-duration treatment with high effectiveness across varied patient populations. In the ASCEND investigation, treatment with DAA was efficacious when delivered by nonspecialist and specialist providers. However, long-term outcomes after initial treatment are unknown.

Objective

To determine the long-term outcomes after DAA treatment independently provided by nurse practitioners, primary care physicians, or specialist physicians using DAA therapy.

Design

Retrospective cohort study.

Setting

Twelve urban, federally qualified health centers in the District of Columbia.

Participants

A total of 551 patients treated for HCV in the ASCEND investigation (A Phase IV Pilot Study to Assess of Community-based Treatment Efficacy in Chronic Hepatitis C Monoinfection and Coinfection with HIV in the District of Columbia).

Interventions

None.

Measurements

Sustained viral response (SVR12), reinfection, retreatment, death.

Results

In this large sample of majority Black individuals receiving care at community-based centers, there was an initial 87% rate of SVR, and after 5 years of follow up, an additional 6.5% of participants were found to be cured. This included individuals originally lost to follow up whose subsequent testing confirmed SVR12, and those with successful retreatment after initial treatment failure. There was a 70% rate of testing for reinfection, with 2 identified reinfections. Treatment outcomes were not associated with original treating provider type.

Limitations

As a retrospective analysis, these findings are limited by the availability of data in the electronic medical record.

Conclusions

DAA is an effective treatment for HCV and can safely be prescribed by multiple provider types, with favorable long-term outcomes.

Hepatitis C virus (HCV) is the most common bloodborne infection in the United States. After a significant decline from 1990 to 2005, the annual incidence rate of acute HCV has increased steadily since 2010, with the greatest increases among adults aged 20–40 years [1], largely attributed to the opioid epidemic and associated injection drug use [2–5]. In addition to preventive measures, curative HCV treatment is a major tool for limiting HCV transmission. The introduction of direct-acting antiviral (DAA) therapies represented a major breakthrough, increasing rates of viral clearance from as little as 10% with interferon-based regimens [6] to more than 95% [7].

Despite these advances, an estimated 40%–60% of people living with chronic HCV in the United States are aware of their diagnosis [8, 9], and only 1/3 of insured individuals have been treated and cured [10, 11]. Prohibitive cost, complex prior authorization policies, and ongoing treatment restrictions have limited the widespread efficacy of DAA in the United States [12, 13]. However, a growing body of literature suggests that HCV service delivery models emphasizing decentralization and care integration can help close the current gap from diagnosis to cure [14, 15].

The ASCEND (A Phase IV Pilot Study to Assess of Community-based Treatment Efficacy in Chronic Hepatitis C Monoinfection and Coinfection with HIV in the District of Columbia) investigation, conducted in 2015, was an early study that sought to determine whether patient outcomes differed when treated by 3 provider types—nurse practitioners, primary care physicians, or infectious disease/gastroenterology specialists. The study was one of the first large, community-based, prospective cohorts in a majority Black, urban population, and was conducted within multiple federally qualified health centers. Six hundred patients were referred by their healthcare provider for HCV treatment and were assigned to receive ledipasvir/sofosbuvir from 1 of the 3 provider types. Early results of this study demonstrated high rates of sustained viral response (SVR) regardless of provider type with no major safety concerns [16], thereby demonstrating the value of task-shifting for HCV treatment.

Despite the initial success of this model, it is unclear if long-term outcomes of patients treated by nonspecialist providers are similar to those treated by specialist providers. In particular, it is unclear if the high rate of SVR found in the original investigation was durable, particularly in a predominantly Black, urban cohort treated in a community-based setting. In this study, we sought to examine the long-term outcomes of patients in the 5 years after initial DAA treatment, and to reexamine the evidence for task-shifting longitudinally in the ASCEND cohort.

METHODS

Study Population

The study population included 551 patients treated for HCV with DAA therapy in the ASCEND study at 12 clinics in Washington, D.C., from 2013 to 2015. Of the original 600 patients treated in ASCEND, 49 were not included because they had been treated at a partner institution with a different electronic medical record (EMR) that was not accessible under the initial data use agreement. Four additional patients died before the completion of the ASCEND trial, and 2 were not included because of inaccessibility of the chart within the EMR. In total, 545 patient charts were reviewed, of which 525 had been seen in the clinic system after the completion of ASCEND and thus had extractable outcomes (Figure 1).

Study Design

This study was conducted as a longitudinal continuation of the ASCEND prospective cohort study comparing the long-term treatment outcomes of patients treated by nurse practitioners, primary care physicians, and specialists. The outcomes assessed included rates of SVR after loss to follow up, retreatment and SVR after treatment failure, reinfection, and HCV-related death. Reinfection was defined as laboratory evidence of detectable viral load after SVR in the original ASCEND study.

A review of patient charts in the EMR, E-Clinical Works, was conducted and relevant data were abstracted in the 5-year follow-up period following each patient's final outcome date in 2015 until June 2020. All outcome assessments, including visits and testing, were completed within the same institution in which patients were originally seen and treated on the ASCEND investigation. Outcome data extracted were matched with treatment data from the original ASCEND Study. A REDCap database housed within the University of Maryland, Baltimore, was developed and used for data abstraction from the primary EMR.

This study was reviewed and granted “exempt” status by the institutional review board of the University of Maryland, Baltimore.

Statistical Analysis

Baseline demographic and clinical characteristics were explored to determine whether any differences existed between the population originally treated in the ASCEND population and those for whom 5-year outcome data was available. Two proportion Z-tests (2-sided) and 2-sample t tests (2-sided) were performed.

Similarly, outcome data were explored using McNemar tests (2-sided) to assess for changes in outcomes between the end of treatment during the ASCEND study and the 5-year follow-up period. To assess any effect of provider type on key patient outcomes, descriptive statistics, and Pearson exact confidence intervals were calculated and the frequency of outcomes were compared using 3-sample chi-squared tests for equality of proportions.

All analysis was conducted in R, version 4.1.0 (R Core Team). A P value less than .05 was considered statistically significant in all statistical testing.

RESULTS

Patient Characteristics

Of the 600 patients treated for HCV with ledipasvir-sofosbuvir (LDV-SOF) during the ASCEND study, 525 were included in the longitudinal analysis. The majority of patients were older (mean age, 59.5 years), male (69%), and Black (95%). The most common genotype was genotype 1a (71%) and 82% of patients had never been treated for HCV before the ASCEND study.

Among patients included in the longitudinal cohort, patients treated by primary care providers were less likely to have been HIV-positive and more likely to have reported previous or never using drugs recreationally at baseline (Table 1).

Lost to Follow Up

Twenty-one patients who had been lost to follow up after the ASCEND study returned for care after the end of the trial and had an HCV RNA checked. Of these, 16 (76%) had achieved SVR (Figure 2).

Non-SVR and Retreatment

Of the 36 patients with treatment failure, 26 were re-treated predominantly with Sofosbuvir-Velpatasvir-Voxilaprevir during the five-year follow up period. Of these, 17 patients achieved SVR (Figure 2). Half of those cured (8) were tested at least 1 time for reinfection, of which none were reinfected after achieving SVR.

Of the 10 individuals who were not retreated, 2 were lost to follow up and 5 were not treated because of provider judgment or a medical reason. These included ongoing heavy alcohol use, kidney failure, decompensated cirrhosis, and hepatocellular carcinoma. Two patients declined retreatment, and 1 patient was not retreated because of insurance denial.

SVR

In addition to the 457 patients that had achieved SVR after the original ASCEND study, 33 patients were found to have been successfully cured. After 5 years, 93.5% of the study population was found to have achieved SVR compared to 87% at the end of the ASCEND study, a 6.5% increase in rate of SVR.

Reinfection

Among patients that were cured in the ASCEND study there was a very high rate of screening for reinfection (Figure 3). Ninety-seven percent of the 491 patients who achieved SVR were seen in the clinic again, and 70% of those had HCV RNA assessment of HCV reinfection at least once. Almost half (45%) were tested in the first year after SVR and were not tested again. Forty-one percent have had at least one assessment of reinfection 3 or more years after SVR with nearly 10% tested 5 years after SVR.

Of 336 patients tested for reinfection, only 2 were found to be positive. One of those was subsequently treated and cured a second time.

Death

In total, 25 patients (4.8%) died in the 5 years after ASCEND. The cause of death is unknown for 14 of these patients. Five patients died from HCV-related causes including hepatocellular carcinoma and decompensated cirrhosis.

Provider Type

The primary outcomes of patients did not differ with provider type. At least 90% of patients treated by each of the 3 provider types was successfully cured. Similarly, there was no statistically significant difference in the number of patients who were retreated, experienced reinfection, or who died from a known, HCV-related cause of death (Tables 2 and 3).

DISCUSSION

The results of this analysis affirm the long-term safety and efficacy of HCV treatment with DAA by nonspecialist providers, emphasizing the durability of task shifting as a mechanism to address the HCV epidemic. In this large sample of majority Black individuals receiving community-based care, there was an initial 87% rate of SVR, and after 5 years of follow up, an additional 6.5% of participants were found to be cured. This included individuals originally lost to follow up whose subsequent testing confirmed SVR, and those with successful retreatment after initial treatment failure. Further, there were high rates of testing for reinfection, with almost no identified reinfection.

In this analysis, we demonstrate continued support for task-shifting to nonspecialist providers, and to HCV treatment in community-based settings. Similar to our initial findings, long-term outcomes after DAA treatment were not impacted by provider type. These findings emphasize that task shifting remains a critical strategy to address the HCV epidemic. With expanded HCV screening guidelines in younger populations driving an increased demand for HCV treatment, and a rising HCV incidence driven by the opioid epidemic, the ability for this therapy to be administered by non-specialist providers is critical. As insurance-based restrictions around provider type are eased in the United States [17], a focus on expanded education for providers and removal of administrative burdens is required for task shifting to fully take root.

This analysis also demonstrates the critical nature of longitudinal care that is often found in community-based primary care settings. Of individuals who did not achieve SVR in the original ASCEND investigation, half were re-treated during the 5-year follow-up period. And of individuals who were lost to follow up, nearly half were found to have SVR at a later timepoint. Without the setting of treatment in community-based sites focused on longitudinal, wraparound services, it is unclear if these levels of durable SVR would be achieved.

Furthermore, in the original ASCEND investigation, loss to follow up was the primary reason for non-SVR. However, of individuals originally lost to follow up who later had HCV RNA assessment, 76% were revealed to have achieved SVR. As loss to follow up and adherence concerns are still utilized to limit, delay, or deny access to HCV treatment [13], these data support high rates of efficacy even in individuals with inconsistent engagement with the healthcare system. Overall, these findings add to the growing body of evidence supporting a low-barrier, minimal monitoring approach to HCV treatment [18].

Five-year follow-up data also revealed high rates of retreatment for individuals who failed initial treatment on ASCEND. As 71% of patients with NSVR were able to access retreatment, with 76% achieving SVR, these data support the feasibility and efficacy of HCV retreatment in community-based settings.

In addition to being efficacious, we found that the benefits of HCV treatment were durable, with only 2 cases of reinfection identified during the 5-year follow-up period. This finding also supports existing research demonstrating the cost-effectiveness of DAAs, as the vast majority of patients may only ever require treatment for a single infection. It is worth noting that while the biggest risk factor for acquiring HCV in the United States is injection drug use, only 14% of the ASCEND population self-reported any type of current substance use at baseline. As such, these findings may not be generalizable to a population with higher rates of active substance use. Consistent with HCV treatment guidelines [19], decisions around retesting should be guided by individual risk factors.

This study had several strengths. The inclusion criteria for the baseline ASCEND study paralleled the LDV-SOF label, ensuring that the results are generalizable to persons living with HCV in the community, outside of a research setting. To our knowledge, the ASCEND cohort is also the largest single study to date of LDV-SOF therapy in Black patients. Furthermore, the subset of the ASCEND cohort followed in this longitudinal analysis had very similar demographic and treatment characteristics to the full baseline cohort, suggesting minimized bias due to censoring.

The study also had some limitations. The biggest limitation is incomplete data on patient outcomes available for consideration and analysis, including provider type for retreatment data. We attempted to minimize any bias associated with the limited data availability by only considering variables that were more commonly captured and would likely be captured in many patient scenarios.

CONCLUSION

In this 5-year follow-up evaluation of the ASCEND cohort, we found increased rates of SVR, high rates of retreatment and retreatment SVR, and minimal reinfection, independent of provider type. These data emphasize the efficacy and durability of task-shifting for HCV treatment. In light of the low uptake of HCV treatment globally, and the current momentum to develop a national HCV elimination plan [12], a task-shifting model should serve as a foundational investment.

Acknowledgments

The authors acknowledge the contributions and support of Sandra Juarez, Ahmad Al-Hadidi, Gregory Carey, PhD, and Qui Cao.

Financial support. This analysis was funded in part by the University of Maryland School of Medicine Program for Research Initiated by Students and Mentors (PRISM) Program.

References

Author notes