Abstract
Purpose:
The new era of direct-acting antivirals (DAAs) against the hepatitis C virus (HCV) has led many primary care clinicians to begin treating HCV. Nevertheless, many patients are referred to specialists due to comorbidities, care complexities, and knowledge gaps of the primary care provider. We compared clinical outcomes for patients treated within a Family Medicine Residency Program (FMRP) affiliated patient-centered medical home (PCMH) with those referred to a specialist.
Methods:
Following didactic education and development of practice resources we conducted a single-center quasi-experimental study of adults with HCV referred for treatment either internally or externally to a specialist between January 2019 and December 2020. The primary outcome was the number of patients with a sustained virologic response at 12 weeks after treatment (SVR12), utilizing an intention-to-treat analysis.
Results:
During the study period 107 patients were assessed by the PCMH, of whom 24 were deemed not a good candidate for treatment. Of the 83 patients referred for treatment, 36 patients were referred externally and 47 were treated internally. While the rate of SVR12 was 100% for both groups when analyzed per protocol (ie, only patients who completed treatment and attended all follow-ups), the rate of SVR12 was 31% for patients referred externally and 62% for patients treated internally when analyzed by intention to treat (relative risk [RR] 2.02, 95% CI 1.18-3.47, P = .01). This difference was entirely attributable to differences in lost to follow-up rates.
Conclusions:
Following education and creation of practice resources, achievement of SVR12 among patients with HCV treated by an internal interdisciplinary family medicine team was superior to those who were externally referred. This was primarily attributable to differences in follow-up rates.
Introduction
The most recent Viral Hepatitis Surveillance Report indicates the rate of acute hepatitis C virus (HCV) infections continues to rise in the United States, with an estimated 57 500 new acute infections in 2019. Newly reported chronic HCV infections more than double that estimate, at 123 312. Tennessee is particularly burdened by chronic HCV, with approximately 1 out of every 1000 residents reporting a new case in 2019. 1 Worsening matters, the COVID-19 pandemic has delayed detection and subsequent treatment of HCV and subsequently the U.S. can expect at best 1 additional case of cirrhosis and liver-related death per 100 000 people by 2030.2,3 This best-case scenario assumes treatment resumes to pre-pandemic levels by the end of 2021—if not, there may be 21 additional cases of cirrhosis and 16 additional liver-related deaths per 100 000 people over the next 10 years. 3
These disheartening statistics underscore the urgent need to expand HCV detection and treatment efforts in the U.S. Even before the COVID-19 pandemic, the Centers for Disease Control encouraged family physicians to treat HCV. 4 The American Academy of Family Physicians, advocated for increased management of HCV within primary care.5,6 Concurrently, family medicine residency program directors have worked over the past 5 years to build capacity for HCV treatment within the specialty by training the next generation of family medicine physicians. 7 Such efforts have produced programs effective at achieving both educational and clinical outcomes within their residents and patients, respectively.8 -10
Though several studies have demonstrated non-inferiority with respect to primary care management of HCV compared to specialist care,11,12 most have evaluated care using per protocol analyses of patients who successfully completed therapy and attended all follow-ups. Such an analysis may overlook the impact of real-world barriers to initiating and completing treatment. Additionally, few comparative studies have been conducted outside of the Veterans Health Administration or federally qualified health centers (FQHCs), with none to our knowledge having been described in a Family Medicine Residency Program (FMRP)-affiliated patient-centered medical home (PCMH).9,11 After providing formal didactic education and creating practice resources to support HCV management, we sought to compare treatment outcomes for patients treated internally by our practice compared with those referred externally.
Methods
The study was a single-center prospective quasi-experimental study comparing the outcomes of adult patients with HCV referred for treatment. The study occurred at the Ascension Medical Group—Saint Louise Family Medicine Center, which is a FMRP-affiliated PCMH in Murfreesboro, Tennessee, which hosts 8 attending physicians, 24 resident physicians, 4 nurse practitioners, and a clinical pharmacist. The study was determined to be quality improvement by the Sterling Institutional Review Board (Sterling IRB. Independent IRB. Atlanta, GA).
In early 2019, a team of 4 clinician champions (an attending physician, clinical pharmacist, resident physician, and resident pharmacist) completed the University of Washington/University of Alabama-Birmingham Hepatitis C Online curriculum and subsequently developed resources for full implementation of the program in the fall of 2019. These resources included a didactic workshop and pocket trifolds as well as note templates and order sets within the electronic health record. During the study period of January 2019 to December 2020, clinicians were free to treat their patients internally or refer the patient externally to a specialist. Clinicians without experience treating HCV patients were encouraged but not required to consult with an HCV champion for their first few patients.
Participants
All patients with an active, chronic HCV infection (as evidenced by a positive HCV RNA) were included in the analysis. Patients were excluded if they declined treatment, spontaneously cleared their infection, expired prior to treatment initiation, were unable to be reached after laboratory diagnosis, moved out of state, were deemed to not be a treatment candidate (eg, poor appointment adherence and ongoing unstable substance use) or were referred to subspecialty care due to decompensated liver disease.
Variables
Patients were classified as an internal referral if they were referred to a family medicine clinician within AMG Saint Louise for treatment and as an external referral if they were referred to a gastroenterologist (the only specialty to which patients were referred to during our study period). Sustained virologic response was defined as a negative HCV viral load 12 weeks post-treatment.
Data Collection
Population health reports were run throughout the study period to identify all patients with a positive HCV RNA referred for treatment, including their disposition and clinical outcome. We collected demographic, laboratory, and treatment-related data. Additionally, we collected data on disposition of the patient, carefully classifying the stage of treatment the patient was in if lost to follow-up. Variables were assessed equally for both groups by utilizing only information documented within the electronic health record.
Analysis
The primary outcome was the proportion of patients achieving a sustained viral response at 12 weeks (SVR12) using an intention-to-treat (ITT) approach. Secondary outcomes included the proportion of patients achieving SVR12 per protocol as well as disposition within the care cascade, which we measured as the proportion who (1) accepted referral, (2) attended their first visit, (3) were prescribed treatment, (4) received and began their first month of treatment, (5) completed treatment, and (6) returned to check HCV RNA viral load at 12 weeks posttreatment. The ITT population included all patients for whom a decision to refer was documented in the chart whereas the per protocol population included only those patients who completed treatment and returned to check HCV RNA viral load at 12 weeks posttreatment.
Fisher’s exact tests and risk ratios were used to compare differences in SVR12 responses across the study arms. Missing data was handled conservatively by assuming that patients lost to follow-up remained infected; however, a sensitivity analysis was conducted to assess the impact of missing SVR12 data for patients who completed treatment. Logistic regression was used to assess the impact of potential factors on SVR12 response, including age, sex, race, diagnosis of cirrhosis, and insurance coverage. Data were entered into REDCap software and analyzed using MedCalc for Windows, version 19.4 (MedCalc Software, Ostend, Belgium). 13
Results
In total, 107 patients with active HCV were assessed during the study period. Of these, 24 were excluded, most commonly due to ongoing unstable substance use (n = 6) or being unable to reach patient after laboratory analysis revealed active HCV (n = 4). Additional exclusions are listed in Figure 1. Patient age followed a bimodal distribution similar to the overall U.S. population. 14 Patients were mostly white and uninsured. Intravenous drug use was the most common risk factor reports in both groups. Nearly all patients were treatment-naïve. Overall, characteristics were similar in both groups though patients treated externally were more likely to have insurance. See Table 1 for additional patient characteristics and Table 2 for a summary of the primary and secondary outcomes discussed below.
Enrollment and disposition.
Patient Characteristics.
Note patients may have more than 1 risk factor documented.
Treatment Outcomes.
Secondary outcomes not adjusted for multiple comparisons.
Disposition
Among patients referred for treatment, 39 (83%) in the internally treated study arm began direct-acting antiviral (DAA) therapy compared with 20 (55%) of patients referred externally. Among patients who began therapy, 38 (97%) patients treated internally completed treatment compared with 16 (80%) of patients referred externally completed treatment. Of these, 29 (76%) in the internal treatment group and 11 (69%) in the external referral group had an HCV RNA viral load checked at 12 weeks post treatment.
Primary Outcome
Using an ITT analysis of participants, the proportion of patients achieving SVR12 in the internally treated arm (62%, 29/47) was significantly higher than the proportion among those referred externally (31%, 11/36; relative risk [RR] 2.02, 95% CI 1.18-3.47, P = .01). Accordingly, only 3 patients would need to be treated internally to achieve an SVR12 that would not have been achieved had the patient been referred externally.
Secondary Outcomes
Using a per protocol analysis of patients who completed treatment and returned to check HCV RNA viral load at 12 weeks posttreatment, 100% of patients in both study arms achieved SVR12.
In total, 18 (38%) of patients treated internally were lost to follow-up, compared with 25 (69%) of patients referred externally. Among patients lost to follow-up, those treated internally were most likely to be lost to follow-up after completing treatment but before returning to confirm SVR12 (9, 50%) whereas those referred externally were most likely to be lost to follow-up before starting treatment (7, 64%).
A sensitivity analysis was conducted to examine the impact of missing SVR12 data among patients who had completed treatment by assuming all patients achieved an SVR12. Under this alternative scenario, the proportion of patients achieving SVR12 in the internally treated arm (81%, 38/47) remained significantly higher than the proportion among those referred externally (44%, 16/36; relative risk [RR] 1.82, 95% CI 1.23-2.69, P = .003). Univariate analysis found no statistically significant factors associated with SVR12 response and thus multivariate analysis was not undertaken.
Discussion
Our experience adds to the growing body of literature demonstrating that patients with uncomplicated HCV infection can be safely and effectively treated within primary care.15,16 In our study, treating patients internally results in a higher rate of SVR12 due to a lower risk of patients becoming lost to follow-up. Indeed, this finding was resilient regardless of whether missing viral loads at 12 weeks post-treatment for patients who had completed treatment were handled as having achieved SVR12 or not. This finding is expected given that the vast majority of patients lost to follow-up in the study became lost prior to starting treatment. That many more patients were lost to follow-up in the externally referred group is likely due to patient concerns regarding cost, specialist access, unfamiliarity with a new clinician, or simply the inconvenience of receiving care from multiple clinics.
The findings of our study are consistent with the results of many studies of community and primary care-based HCV testing and treatment, including a recent systematic review and meta-analysis which graded the strength of evidence in these settings as medium for both treatment uptake and achieving SVR12 response. 17 However, few of these studies compared their intervention to specialist care and most reported their results utilizing a per protocol analysis. Syed et al 11 found no difference between primary care or specialty care among 377 veterans with HCV; however patients lost to follow-up were excluded per protocol. The ASCEND study was a nonrandomized, open-label trial that referred patients in a specified manner to specialist, primary care physician, or nurse practitioner care. 12 Their ITT analysis did include patients lost to follow-up but found no difference in SVR12 response rate. However, study medications in the ASCEND study were provided by Gilead Sciences, removing one of the chief barriers to treatment uptake, medication access. Most recently, Wade et al 18 randomized 136 people who inject drugs (PWID) infected with HCV to receive treatment at their primary care site or hospital-based specialist care. Similar to our study, their intention-to-treat analysis found 75% of patients in the primary care arm achieved an SVR12 response compared with 34% of those referred for specialist care (relative risk [RR] 2.48, 95% CI 1.54-3.95). In contrast to their intervention, our practice did not offer opioid substitution therapy during the study period, suggesting that—while valuable—the absence of such therapy should not discourage health systems from implementing primary-care based HCV management.
Key strengths of our study include the pragmatic nature of our implementation model. Using a clinician champion model, we were able to open services to patients of 36 primary care clinicians through the intensive training of only 4 clinicians. As members of the practice itself, the champions were able to promote resources and address barriers to implementing the new service in a way that led to quick uptake. Encouraging new clinicians to consult with a champion and utilize guideline-driven note templates and order sets increased clinician confidence and maintained safe and effective treatments. Furthermore, since clinicians were free to continue referring patients externally, our observational study provides real-world evidence of the effectiveness of HCV care provided by an interdisciplinary family medicine team.
There are several limitations to this study. This study occurred in a safety net clinic with a sliding scale fee schedule, a patient advocate, and an embedded clinical pharmacist. Thus our findings may not be applicable to clinics without such support for vulnerable patients. Next, while the observational nature of the trial provides real-world evidence of efficacy, it also introduces the possibility of undocumented confounders not captured in the analysis. For example, fewer risk factors were reported overall in the external referral group; this likely reflects better documentation of risk factors for patients treated internally due to use of a note template. However, it is notable that significantly more patients treated internally were uninsured and numerically more patients had a diagnosis of cirrhosis or were treatment-experienced—risk factors expected to worsen treatment outcomes. Accordingly, we feel that referral bias did not make a significant impact on our study findings.
Conclusion
Treatment of HCV patients internally within a family medicine PCMH results in increased SVR12 response rates compared with patients referred externally due to fewer patients becoming lost to follow-up. Utilization of an interdisciplinary clinician champion model may be utilized to increase the number of primary care clinicians trained to manage patients with HCV, especially in medium to large group practices or health systems.
Footnotes
Authors’ Note
Prior Presentation: This research has not been previously presented in any form.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
