Sustained Virologic Response Rates Before and After Removal of Sobriety Restriction for Hepatitis C Virus Treatment Access

Abstract

Objectives:

Until November 1, 2018, Illinois Medicaid restricted coverage of hepatitis C virus (HCV) medication to patients with sobriety from alcohol and illicit substances for ≥12 months. This policy limited treatment access for patients with a high risk of HCV transmission, despite clinical trial and real-world data demonstrating high sustained virologic response (SVR) rates among patients with substance use. The objective of this study was to compare HCV SVR rates between patients treated before and after removal of the Illinois Medicaid sobriety restriction.

Methods:

We performed a retrospective cohort study of Medicaid-insured patients who completed direct-acting antiviral treatment at an urban, academic medical center in Illinois from January 1, 2014, through October 21, 2020. The primary endpoint was SVR. We compared group characteristics using χ² or Fisher exact tests for categorical variables and Wilcoxon rank-sum tests for continuous variables. We used logistic regression to compare SVR rates before and after the policy change, adjusting for differences between groups.

Results:

A total of 496 patients (348 pre–policy change; 148 post–policy change) started treatment; excluding loss to follow-up/early discontinuation, SVR rates were 95.4% (309 of 324) pre–policy change and 97.1% (134 of 138) post–policy change. SVR rates did not differ after adjusting for the use of historic HCV regimens and the higher cirrhosis rate in the pre–policy change group compared with the post–policy change group (odds ratio = 0.98; 95% CI, 0.32-3.67).

Conclusion:

HCV SVR rates were similar before and after removal of the Illinois Medicaid sobriety restriction, regardless of group differences. Results support HCV treatment regardless of documented sobriety to facilitate progress toward HCV elimination.

The World Health Organization and the United States have established the goal of hepatitis C virus (HCV) elimination by 2030.^1-3 The US Department of Health and Human Services has publicized local and state hepatitis elimination projects working toward the Viral Hepatitis National Strategic Plan; yet, the most recent website update documented that fewer than half of states had formal HCV elimination plans. ⁴ A recently published modeling study purports that the United States is on track to eliminate HCV in 2037, and Illinois in 2044, in contrast to the global and national goals. ⁵

Direct-acting antivirals (DAAs) have revolutionized HCV treatment, with sustained virologic response (SVR) rates >95%. ³ Despite high cure rates, DAAs are useful for elimination efforts only if they are accessible to HCV-infected patients. In the United States, many HCV-infected patients are insured by a state Medicaid program, several of which have strict clinical, behavioral (sobriety), and/or administrative criteria for HCV treatment coverage to control drug expenditures. Sobriety restrictions decrease HCV treatment access for patients with high HCV transmission risk, despite clinical trial and real-world data demonstrating high SVR rates among HCV-infected patients with substance use.^6-10 Lower HCV treatment initiation rates have been observed among Medicaid-insured patients compared with Medicare-insured patients, regardless of stage of fibrosis. ¹¹ In addition, both Medicaid patients and Medicare patients have lower rates of DAA initiation compared with patients with private health insurance. ¹²

The landscape of HCV-infected patients has shifted in the past decade to include more patients affected by the HCV–opioid syndemic. ¹³ The joint HCV guidance of the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America (AASLD-IDSA) recommends against barriers to HCV treatment and recommends treatment of people who inject drugs (PWID) to assist with HCV elimination efforts. ¹⁴ Despite this national HCV guidance, most state Medicaid programs had sobriety requirements for HCV treatment access in 2014, and 19 state Medicaid programs maintained restrictions until January 2019.^15-17 Reports in August 2020 and May 2021 documented that 13 state Medicaid programs still had sobriety requirements that mandated abstinence from alcohol and illicit substances for 1, 3, or 6 months prior to HCV treatment approval. ¹⁸ As of January 2022, 8 state Medicaid programs still had sobriety requirements for 3 or 6 months. ¹⁹ No states still require 12 months of sobriety; Illinois Medicaid removed its 12-month sobriety restriction for HCV medication coverage on November 1, 2018, following several other states’ policy changes that allowed treatment of patients with substance use. State Medicaid HCV treatment restrictions undermine public health efforts to achieve HCV elimination. Evidence of treatment success among populations in states that have removed restrictions may spur removal of policy barriers to HCV treatment access in other states.

SVR is the clinical outcome that aligns with global and national HCV elimination goals, and published data on SVR rates before and after state Medicaid policy changes are lacking. The objective of this study was to evaluate and compare an urban academic medical center’s HCV SVR rates among Medicaid-insured patients in Illinois before and after the 2018 sobriety policy change. It is important to understand the impact of this policy change on SVR as the nation works toward the goal of HCV elimination by 2030.

Methods

Study Design

The payor mix of the University of Illinois Hospital and Health Sciences System (UI Health) includes 45% Medicaid, 31% Medicare, 21% commercial, and 3% self-pay or other coverage. We performed a retrospective cohort study of HCV-infected adult (defined as aged ≥18 y) Medicaid-insured patients with HCV who initiated a dual- or triple-DAA regimen treatment course prescribed from the Liver Clinic at UI Health from January 1, 2014, through October 21, 2020. The UI Health Liver Clinic uses a clinical pharmacist–driven HCV treatment model within its interdisciplinary health care team. The clinical pharmacist selects treatment, manages drug interactions, offers extensive education at treatment initiation, orders and monitors laboratory values for treatment efficacy, manages patients during treatment, and encourages adherence and treatment completion. ²⁰

We stratified patient groups according to initiation of treatment pre–policy change (January 1, 2014, through October 31, 2018) or post–policy change (November 1, 2018, through October 21, 2020). Patients were excluded if 12 weeks had not elapsed after treatment completion by April 9, 2021. The primary outcome was SVR, defined as an undetectable HCV RNA polymerase chain reaction test result at ≥12 weeks after HCV treatment completion, which indicates cure and treatment success. The study was reviewed and approved by the University of Illinois at Chicago Institutional Review Board. A waiver of consent authorization was granted for this retrospective medical record review.

Data Collection

We reviewed electronic medical records and managed data using version 11.0 of the REDCap electronic data capture tool hosted at the University of Illinois at Chicago. Baseline data included data from the 12 months before treatment initiation, as available. Data collection included demographic characteristics, comorbidities, laboratory data, HCV genotype, HCV treatment history, fibrosis stage, and Child-Turcotte-Pugh class in patients with cirrhosis. Regimens were categorized as historic (no longer commercially available: simeprevir + sofosbuvir ± ribavirin, daclatasvir + sofosbuvir ± ribavirin, or paritaprevir/ritonavir/ombitasvir/dasabuvir ± ribavirin) versus all other currently available DAA regimens as of 2021 (elbasvir/grazoprevir ± ribavirin; glecaprevir/pibrentasvir; ledipasvir/sofosbuvir ± ribavirin; sofosbuvir/velpatasvir ± ribavirin; sofosbuvir/velpatasvir/voxilaprevir ± ribavirin; glecaprevir/pibrentasvir + sofosbuvir + ribavirin). We reviewed documentation from the clinical pharmacist’s treatment initiation notes and other clinicians’ notes for baseline patient-reported social history, including alcohol use, intranasal or inhaled illicit substance use, or intravenous drug use (IDU), and use of medications for opioid use disorder (MOUD). We defined sobriety as a composite of no current or recent alcohol use, inhaled or intranasal illicit drug use, or IDU in the 12 months prior to treatment initiation.

We collected HCV treatment–related data from the initial HCV evaluation documentation, through SVR assessment, treatment discontinuation, or determination of loss to follow-up as of April 9, 2021. Data included DAA regimen, length of therapy, and treatment discontinuation reason, if applicable. We collected HCV RNA results through ≥12 weeks after treatment completion to determine SVR. We defined loss to follow-up as no SVR data available at ≥12 weeks after the completion of therapy for any reason, including death, at the end of the study period. Patients who discontinued treatment were those who did not complete the initially prescribed length of treatment for any reason, including health care provider instruction, self-discontinuation, or death. Clinicians assessed treatment adherence and recorded any missed doses in the electronic medical record. We obtained adherence data from clinicians’ documentation of adherence during patient assessment (via patient’s self-reported adherence to the daily dosing of the DAA regimen) or per the dispensing pharmacy (which could have indicated missed doses due to late refills).

Statistical Analysis

We analyzed data using R version 4.0.2 (R Core Team) and described data with counts and percentages for categorical variables and means and SDs for continuous data. We compared differences in characteristics between pre–policy change and post–policy change groups by using the Pearson χ² test or Fisher exact test for categorical variables and the Wilcoxon rank–sum tests for continuous variables. We calculated standardized effect sizes with use of Cohen h for categorical variables or Cohen d for continuous variables. We considered P < .05 significant and reported 2-tailed P values. We reported SVR rates and corresponding 95% CIs in the pre–policy change versus post–policy change groups. We classified patients as “unknown” with respect to the outcome if they were lost to follow-up or had discontinued treatment.

We estimated the adjusted odds of achieving SVR among the pre– versus post–policy change groups using multivariate logistic regression to control for potential confounding variables. We used an iterative, purposeful selection process; clinically relevant variables that produced a 10% change in the parameter estimate for the impact of the policy change were candidates for inclusion. While inhaled or intranasal illicit drug use in the previous 12 months met these inclusion criteria, we did not include this variable in the model because it was directly affected by the policy change. The primary analysis included data only on patients for whom HCV RNA results were available at ≥12 weeks after treatment completion. To examine the impact of missing data, we performed sensitivity analyses that treated all unknown SVR results as treatment success versus treating all unknown SVR results as treatment failures. We performed 3 additional sensitivity analyses. First, we excluded patients from the pre–policy change group without sobriety in the 12 months prior to treatment initiation per the intent of the Illinois Medicaid sobriety policy, and we compared the SVR rates of the pre– and post–policy change groups. Second, we again excluded patients from the pre–policy change group without sobriety per the Medicaid policy intent, and we compared the SVR rates of these sober pre–policy change group patients with post–policy change group patients without documented sobriety, because one goal of the new policy was intention to allow treatment of patients with substance use. Third, we compared SVR rates after exclusion of pre–policy change group patients who received historic regimens. This approach accounted for historic regimen use in the pre–policy change group but not in the post–policy change group, because the lack of historic regimen use in the post–policy change group was a limitation in fitting the multivariate logistic regression model.

Results

We identified 496 Medicaid-insured patients for inclusion: 348 (70%) patients in the pre–policy change group and 148 (30%) patients in the post–policy change group. Baseline characteristics of each group were similar in age, sex, race and ethnicity, and previous HCV treatment experience (P > .05 for each; Table 1). When we compared the pre– and post–policy change groups, the standardized difference was >10% for the following baseline characteristics: sex, race and ethnicity, transplantation, substance use in the past 12 months, use of MOUD, genotype, cirrhosis, and regimen.

Table 1.

Characteristics of Medicaid-insured patients who completed HCV direct-acting antiviral treatment at an urban academic medical center, before and after removal of Illinois Medicaid sobriety restrictions, January 1, 2014, through October 21, 2020 (N = 496)

Characteristics	Pre–policy change group (n = 348), no. (%) a	Post–policy change group (n = 148), no. (%) a	Standardized difference b	P value c
Age, mean (SD), y	58.0 (8.0)	59.5 (9.3)	−0.05	.15
Born during 1945-1965	288 (82.8)	120 (81.1)	0.05	.65
Sex				.16
Female	134 (38.5)	67 (45.3)	−0.12
Male	214 (61.5)	81 (54.7)	0.12
Race and ethnicity				.73
Hispanic	51 (14.7)	19 (12.8)	0.06
Non-Hispanic African American/Black	211 (60.6)	98 (66.2)	−0.10
Non-Hispanic American Indian/Alaska Native	1 (0.3)	0	0.11
Non-Hispanic Asian/Pacific Islander	8 (2.3)	5 (3.4)	−0.07
Non-Hispanic White	75 (21.6)	26 (17.6)	0.10
Unknown	2 (0.6)	0	0.16
Alcohol use (past 12 mo)	93 (26.7)	67 (45.3)	−0.38	<.001
Intranasal or inhaled illicit drug use (past 12 mo)	47 (13.7) d	49 (33.1)	−0.46	<.001
Intravenous drug use (past 12 mo)	13 (3.9) e	3 (2.0)	0.11	.41
Concurrent medication for opioid use disorder	21 (6.0)	26 (17.6)	−0.38	<.001
Medication for opioid use disorder				.62
Buprenorphine/naloxone	1 (4.8)	3 (11.5)	−0.27
Methadone	20 (95.2)	23 (88.5)	0.26
HCV genotype				.02
1a	215 (61.8)	90 (60.8)	0.02
1b	91 (26.1)	31 (20.9)	0.12
2	11 (3.2)	11 (7.4)	−0.19
3	17 (4.9)	12 (8.1)	−0.13
4	3 (0.9)	4 (2.7)	−0.14
6	1 (0.3)	0	0.11
Other subtype or mixed genotype	10 (2.9)	0	0.34
Stage (noncirrhotic) f				.002
F0	17 (11.4)	7 (6.9)	0.11
F1	12 (8.1)	26 (25.7)	−0.49
F2	59 (39.6)	36 (35.6)	0.08
F3	61 (40.9)	32 (31.7)	0.19
Cirrhotic	194 (55.7)	45 (30.4)	0.53	<.001
Child-Turcotte-Pugh classification				.68
A	162 (83.5)	38 (84.4)	0
B	26 (13.4)	7 (15.6)	−0.09
C	6 (3.1)	0	0.35
History of hepatocellular carcinoma	11 (3.2)	4 (2.7)	0.03	>0.99
History of solid organ transplant	22 (6.3)	5 (3.4)	0.14	.19
HCV treatment history				.51
Experienced	55 (15.8)	20 (13.5)	0.06
Naïve	293 (84.2)	128 (86.5)	−0.06
HCV treatment regimen				<.001
Currently available regimen g	318 (91.4)	148 (100.0)	−0.61
Regimen no longer available h	30 (8.6)	0	0.60

Abbreviation: HCV, hepatitis C virus.

Source: Authors’ data from retrospective cohort study.

^aUntil November 1, 2018, Illinois Medicaid restricted coverage of HCV medication to patients with sobriety from alcohol and illicit substances for ≥12 months.

^bStatistical tests performed: Cohen h for categorical variables and Cohen d for continuous variables.

^cStatistical tests performed: Wilcoxon rank–sum test, Pearson χ² test, Fisher exact test, with P < .05 considered significant.

^dn = 344 for the intranasal or inhaled illicit drug use (past 12 months) pre–policy change group; 4 patients did not have data available.

^en = 337 for the intravenous drug use (past 12 months) pre–policy change group; 11 patients who were noncirrhotic but not otherwise staged were excluded from the statistical analysis.

^fn = 149 for the noncirrhotic pre–policy change group; 5 patients who were noncirrhotic but not otherwise staged were excluded from the statistical analysis. n = 101 for the noncirrhotic post–policy change group; 2 patients who were noncirrhotic but not otherwise staged were excluded from the statistical analysis.

^gElbasvir/grazoprevir ± ribavirin; glecaprevir/pibrentasvir; ledipasvir/sofosbuvir ± ribavirin; sofosbuvir/velpatasvir ± ribavirin; sofosbuvir/velpatasvir/voxilaprevir ± ribavirin; glecaprevir/pibrentasvir + sofosbuvir + ribavirin.

^hDaclatasvir + sofosbuvir ± ribavirin; paritaprevir/ritonavir/ombitasvir/dasabuvir ± ribavirin; sofosbuvir + simeprevir ± ribavirin.

A larger proportion of patients in the post–policy change group than in the pre–policy change group self-reported use of alcohol (45.3% vs 26.7%; P < .001) and intranasal or inhaled illicit drugs (33.1% vs 13.7%; P < .001) in the past 12 months; however, IDU in the past 12 months was not significantly different between the post–policy change group (3.9%) and the pre–policy change group (2.0%; P = .41). Patients in the post–policy change group were significantly more likely to use MOUD than those in the pre–policy change group (17.6% vs 6.0%; P < .001); methadone was the most used agent by both groups. Patients in the post–policy change group had lower rates of cirrhosis than did patients in the pre–policy change group (30.4% vs 55.7%; P < .001). The pre–policy change group included the use of historic regimens, whereas the post–policy change group did not (8.6% vs 0%; P < .001). Few treatment-experienced patients (13 [3.7%] in the pre–policy change group and 10 [6.8%] in the post–policy change group) had been previously treated with DAAs; the remainder of the treatment-experienced patients were treated with interferon-based regimens. In the pre–policy change group, 222 (63.8%) patients reported full sobriety in the 12 months before HCV treatment initiation, and in the post–policy change group, 65 (43.9%) patients reported sobriety. In the pre–policy change group, 126 (36.2%) patients reported substance use in the 12 months prior to HCV treatment initiation, and in the post–policy change group, 83 (56.1%) patients reported substance use.

Overall, the pre– and post–policy change groups did not differ by loss to follow-up (4.6% vs 5.4%; P = .70) or early treatment discontinuation rates (2.3% vs 1.4%; P = .73). Groups did differ by medication adherence; of the 325 patients with adherence data available in the pre–policy change group and of the 143 patients with adherence data available in the post–policy change group, 17.5% and 25.9%, respectively, reported missing ≥1 dose of HCV treatment (P = .038).

After excluding 16 patients who were lost to follow-up and 8 patients with early discontinuation, the pre–policy change group included 324 patients with an SVR of 95.4%; after excluding 8 patients who were lost to follow-up and 2 patients with early discontinuation, the post–policy change group included 138 patients with an SVR rate of 97.1%, for a crude odds ratio (OR) of 1.63 (95% CI, 0.58-5.79; Table 2). The odds of achieving SVR did not differ between the post–policy change group and the pre–policy change group after adjusting for cirrhosis and use of historic HCV regimens in the multivariate logistic regression model (adjusted OR = 0.98; 95% CI, 0.32-3.67). In all sensitivity analyses, SVR rates were not significantly different in crude or adjusted models.

Table 2.

Sustained virologic response (SVR) results and sensitivity analyses among Medicaid-insured patients who completed HCV direct-acting antiviral treatment at an urban academic medical center, before and after removal of Illinois Medicaid sobriety restrictions, January 1, 2014, through October 21, 2020 (N = 496) ^a

Variable	No. of patients		SVR rates, %		Logistic regression outputb,c
	Pre–policy change group	Post–policy change group	Pre–policy change group	Post–policy change group	Crude odds ratio (95% CI)	Adjusted odds ratio (95% CI)
Patients with known SVR results only d	324	138	95.4	97.1	1.63 (0.58-5.79)	0.98 (0.32-3.67)
Missing SVR data treated as success (SVR achieved)	348	148	95.7	97.3	1.62 (0.58-5.76)	0.97 (0.31-3.63)
Missing SVR data treated as failure (SVR not achieved)	348	148	88.8	90.5	1.21 (0.65-2.37)	1.05 (0.55-2.12)
Restriction of pre–policy change group to patients with sobriety e in past 12 mo; known SVR results only d	205	138	95.6	97.1	1.54 (0.49-5.77)	1.08 (0.30-4.33)
Missing SVR data treated as success (SVR achieved)	222	148	95.9	97.3	1.52 (0.49-5.70)	1.06 (0.30-4.26)
Missing SVR data treated as failure (SVR not achieved)	222	148	88.3	90.5	1.27 (0.65-2.58)	1.14 (0.56-2.40)
Restriction of pre–policy change group to patients with sobriety e in past 12 mo; restriction of post–policy change group to patients without sobriety e in past 12 mo; known SVR results only d	205	83	95.6	97.6	1.86 (0.47-12.40)	1.17 (0.26-8.17)
Missing SVR data treated as success (SVR achieved)	222	92	95.9	97.8	1.90 (0.48-12.60)	1.19 (0.27-8.33)
Missing SVR data treated as failure (SVR not achieved)	222	92	88.3	88.0	0.98 (0.47-2.15)	0.84 (0.39-1.91)
Restriction of pre–policy change group to patients using currently available regimens only; known SVR results only f	294	138	96.3	97.1	1.30 (0.44-4.77)	1.06 (0.34-3.99)
Missing SVR data treated as success (SVR achieved)	318	148	96.5	97.3	1.29 (0.43-4.72)	1.05 (0.34-3.95)
Missing SVR data treated as failure (SVR not achieved)	318	148	89.0	90.5	1.18 (0.63-2.34)	1.09 (0.57-2.20)

Abbreviation: HCV, hepatitis C virus.

Source: Authors’ data from retrospective cohort study.

^aUntil November 1, 2018, Illinois Medicaid restricted coverage of HCV medication to patients with sobriety from alcohol and illicit substances for ≥12 months.

^bThe pre–policy change group is the reference group in all models.

^cOdds of achieving SVR 12 weeks after treatment completion.

^dAdjusted for presence of cirrhosis (yes vs no) and therapeutic regimen (no longer available vs currently available).

^eSobriety refers to patients who denied current and recent use of alcohol, inhaled or intranasal illicit drug use, or intravenous drug use in the 12 months before treatment initiation.

^fModels adjusted for presence of cirrhosis (yes vs no).

Discussion

The objective of our study was to compare SVR rates between patients treated before and after the change in Illinois Medicaid policy in 2018 that removed the 12-month sobriety restriction for HCV treatment access. After the policy change, we found an increase in the proportion of HCV-treated patients who reported substance use within 12 months prior to HCV treatment initiation. Despite the differences in the populations treated, the SVR rates for HCV-treated patients were similar before and after the policy change, and the SVR rates were high. Collectively, these results demonstrate that the removal of sobriety restrictions did not affect the SVR rate for a representative sample of nearly 500 Medicaid-insured patients.

Loss to follow-up and early treatment discontinuation led to missing SVR outcome data for 34 (6.9%) patients. Reasons for loss to follow-up were not available for all patients, and the impact of potential differences in characteristics between these patients and patients with SVR data available may have led to an overestimation of SVR rates. However, the lost-to-follow-up group was small (4.6% of the pre–policy change group; 5.4% of the post–policy change group), and the sensitivity analyses testing the impact of not including this population were consistent with the results of our primary analysis.

Treatment adherence in our study may have been exaggerated because data were patient-reported. While our strict definition of adherence (no missed doses) did differ significantly between the pre– and post–policy change groups in our study, this difference may be the result of improved documentation across the years of the study. Historically, payor restrictions on sobriety were tied to assumed nonadherence among patients with substance use. ²¹ However, high SVR rates are documented with current DAAs even in patients with imperfect or low adherence. ²²

Restrictions for HCV treatment access may provide short-term cost savings, yet HCV elimination is estimated to provide even greater cost savings. Strict behavioral criteria, including abstinence from illicit substances and alcohol prior to treatment coverage, was associated with less Medicaid spending on sofosbuvir-based HCV treatment. ²³ However, treatment of all patients was estimated to provide >$12 billion in cumulative Medicaid savings. ²⁴ Without treatment of all HCV-infected patients in the United States, total HCV-related health care costs were projected to increase from $6.5 billion in 2011 to $9.1 billion annually by 2024. ²⁵ In addition to cost savings, HCV DAA treatment in patients with substance use disorders was shown to reduce the risk of decompensated cirrhosis in noncirrhotic patients and reduce the risk of hepatocellular carcinoma in patients with cirrhosis. ²⁶

More than half of states’ Medicaid plans had implemented sobriety restrictions on HCV treatment coverage since the approval of all-oral DAAs in 2014; 13 had remaining sobriety restrictions in August 2020 that persisted and were documented in the May 2021 report, and 8 maintained restrictions in January 2022.^18,19 The total number of Medicaid patients in the United States who have been denied access to HCV treatment is unknown. Yet, these treatment restrictions have undoubtedly slowed the nation’s progress toward HCV elimination and perpetuated the spread of HCV infection. Medicaid coverage for the 8 states with sobriety restrictions for HCV treatment is inconsistent with national strategic planning for HCV elimination.

The AASLD-IDSA joint HCV guidance recommends HCV treatment for nearly all patients, and specifically for PWID. ¹⁴ In addition, 4 objectives of the Viral Hepatitis National Strategic Plan (1.4, 2.2, 3.2, 5.1) explicitly highlight the need for HCV prevention, diagnosis, and treatment for patients with substance use disorder. ³ Both the National Academies of Sciences and the World Health Organization address the need for harm reduction and HCV treatment among patients with substance use.^1,2

Despite the Medicaid sobriety restriction, 3.9% of pre–policy change group patients had documented IDU, 14% used intranasal or inhaled illicit drugs, and 27% reported alcohol use within 12 months prior to HCV treatment initiation. Initiation of HCV treatment despite substance use was the result of patients’ disclosure of substance use that either occurred or was revealed after DAA approval. The use of MOUD at the time of HCV treatment initiation was significantly higher in the post–policy change group than in the pre–policy change group. The post–policy change group included patients whose substance use was not a barrier for HCV treatment access, and social history of patients taking MOUD may include substance use relapse or recent (<12 months) sobriety. The proportion of PWID was perhaps surprisingly small in the post–policy change group despite the ongoing opioid epidemic. Up to 2.6% of adults in the United States have injected drugs in their lifetime, and the estimated prevalence of HCV among PWID is 53% (ranging from 38% to 68% among states).^27-29 A study that analyzed the effect of HCV treatment as prevention among HCV-infected PWID found that with a PWID HCV prevalence of ≤75%, treatment of 24% of patients annually would be needed to eliminate HCV within 20 years. ²⁵ In the absence of complete HCV treatment coverage for the patient population with the highest risk of HCV transmission in the United States, stemming the HCV epidemic will be impossible.

Health care providers are encouraged to link all patients with substance use disorder to appropriate resources to assist with treatment, in addition to treating their HCV infection, and co-localization of care for HCV and substance use disorder is a strategy for HCV elimination. ¹⁴ Published data from real-world and clinical trials document SVR rates at ≥90% in patients with past, recent, or active alcohol and drug use. In a Veterans Affairs study of 13 742 veteran patients, high SVR rates were achieved regardless of alcohol use. ³⁰ A meta-analysis of >1700 patients found no significant difference in SVR rates of patients taking opioid substitution therapy compared with controls or in PWID versus controls. ¹⁴ Similarly, high SVR rates were documented in prospective clinical trials including patients with recent or ongoing drug use and patients taking MOUD.^6-9 Removal of all HCV treatment restrictions in these patients is necessary to work toward the goal of HCV elimination.

Limitations

This study had several limitations, including the scope of our research. Our study was performed at a single institution with a pharmacist-driven HCV treatment model and included only patients with Illinois Medicaid coverage. Our pre–policy change group included a larger number of patients than the post–policy change group because of the longer time frame for treatment initiation in that group (58 months vs 23 months). The PWID proportion is perhaps smaller than that of other institutions, which limits the applicability of our study results in settings with a high proportion of PWID, especially in the midst of the HCV–opioid syndemic. Our results may not be generalizable to those of other institutions, nor is our patient population necessarily representative of other state Medicaid plan enrollees.

For the multivariate logistic regression model stability, the number of variables included in the final model was limited by the small number of patients who did not achieve SVR. Accordingly, the potential confounders may not have been accounted for in the multivariate logistic regression model.

The pre–policy change group had a higher percentage of patients with cirrhosis than the post–policy change group because Illinois Medicaid restricted HCV treatment to patients with cirrhosis in 2014 after the approval of all-oral DAAs. Treatment access was later opened to include stage 3 fibrosis (F3) patients in addition to stage 4 (F4, cirrhotic) patients. The removal of treatment restriction based on stage of fibrosis occurred on the same date as the Medicaid sobriety restriction removal on November 1, 2018, so patients with stages F0-F4 were eligible for HCV treatment in the post–policy change group. Cirrhosis was associated with slightly lower SVR rates than those of patients without cirrhosis, specifically with use of DAA regimens that are no longer available.^31,32 In addition, patients with decompensated cirrhosis have poorer SVR rates than patients with compensated cirrhosis or those without cirrhosis. ³⁰ Six of the 15 treatment failures in the pre–policy change group had decompensated cirrhosis (Child-Turcotte-Pugh class B). The shift in regimen use varied with Medicaid formulary coverage, which was affected by cost and availability of the DAAs. Early DAA regimens were effective, yet current regimens deliver even consistently higher SVR rates across patient populations whose HCV was previously difficult to cure. ³² SVR results remained similar after sensitivity analysis controlling for historic regimen use, which could be the result of use of currently available regimens in most (91.4%) pre–policy change group patients.

Conclusion

Medicaid-insured HCV-treated patients at our institution had similar SVR rates before and after the Illinois Medicaid sobriety restriction removal. These real-world clinical results may be used as support for removal of sobriety restrictions for HCV treatment and future policy change. State Medicaid plans must cover treatment of patients with substance use to align with national HCV elimination goals.