Community-Based Point-of-Diagnosis Hepatitis C Treatment for Marginalized Populations

Key Points

Question

Can providing hepatitis C virus (HCV) treatment at the time of diagnosis in a nonclinical community setting improve direct-acting antiviral treatment uptake for medically underserved populations?

Findings

In this single-arm nonrandomized controlled trial of 89 people who inject drugs and are experiencing homelessness, point-of diagnosis direct-acting antiviral initiation was feasible, acceptable, and safe, with 87 patients starting HCV treatment at the time of diagnosis, 69 completing treatment, 58 attaining a sustained virologic response at posttreatment week 12 or later, and none experiencing adverse events.

Meaning

This study model can expand the toolkit of HCV test-and-treat models to reach a broader population of marginalized communities—a critical next step to achieving national HCV elimination goals.

This nonrandomized controlled trial evaluates the feasibility, acceptability, and safety of providing hepatitis C virus treatment at diagnosis within a medically underserved population.

Abstract

Importance

Disparities persist in testing and treatment for hepatitis C virus (HCV), leaving socially marginalized populations less likely to benefit from curative treatment. Linkage services are often insufficient to overcome barriers to navigating the medical system and contextual factors.

Objective

To determine the feasibility, acceptability, and safety of HCV treatment at the point of HCV infection diagnosis disclosure in a nonclinical community setting.

Design, Setting, and Participants

In this single-arm nonrandomized controlled trial conducted between July 1, 2020, and October 31, 2021, street-outreach recruitment targeted people experiencing homelessness and injecting drugs in an urban US community who were eligible for simplified HCV treatment.

Interventions

Study procedures were designed to reflect the community environment and services needed to provide HCV testing, disclosure, and treatment in a nonclinical site. The test-and-treat No One Waits (NOW) model of care provided a 2-week starter pack of 400 mg of sofosbuvir and 100 mg of velpatasvir at time of HCV RNA results disclosure. Participants were transitioned to insurance-provided sofosbuvir-velpatasvir when feasible to complete a 12-week treatment course.

Main Outcomes and Measures

The primary end point was sustained virologic response at posttreatment week 12 or later (SVR12). Acceptability end points were treatment initiation and completion. Safety end points were treatment discontinuation because of a late exclusion criterion and adverse events.

Results

Of the 492 people (median [IQR] age, 48 [37-58] years; 62 [71%] male) who underwent anti-HCV testing, 246 (50%) tested anti-HCV positive, and 111 (23%) tested HCV RNA positive and were eligible for simplified HCV treatment. Eighty-nine of the 111 eligible participants (80%) returned for confirmatory RNA results, and 87 (98%) accepted and initiated HCV treatment. Seventy (80%) were currently injecting drugs, 83 (97%) had an income below the poverty line, and 53 (61%) were currently unsheltered. Most had HCV genotype 1a (45 [52%]) or 3 (20 [23%]). Sixty-nine (79%) completed 12 weeks of sofosbuvir-velpatasvir treatment, 2 stopped treatment because of low adherence, and 16 were lost to follow-up. Of the 66 participants who completed treatment and had a successful blood draw, 61 (92%) had undetectable HCV RNA at treatment completion. Of the 87 treated patients, 58 achieved SVR12, leading to a treatment response of 67% (95% CI, 56%-76%) among the intention-to-treat group and 84% (95% CI, 73%-92%) among the per-protocol group. There were no adverse events, late exclusions, or deaths.

Conclusions and Relevance

In this nonrandomized controlled trial of HCV treatment at the point of diagnosis, the NOW model of care reduced steps between HCV testing and treatment initiation and resulted in high levels of treatment initiation, completion, and cure. The NOW model of care can expand the current HCV test-and-treat toolkit by reaching a broader population of marginalized communities and expediting curative therapy.

Trial Registration

ClinicalTrials.gov Identifier: NCT03987503

Introduction

Worldwide hepatitis C virus (HCV) prevalence is high, with recent estimates ranging from 57 million to 71 million people living with HCV.^1,2,3 Marginalized populations, including people experiencing homelessness or housing instability, people in marginalized racial and ethnic groups, people with current or prior justice system involvement, and, particularly, people who inject drugs (PWID), are disproportionately affected by HCV.^4,5,6 In the US, more than 80% of HCV infections have been associated with injection drug use, and increases in HCV incidence have corresponded to the opioid epidemic in recent years.^6,7

Since the advent of direct-acting antivirals (DAAs) to treat HCV, sustained virologic response (SVR) or cure rates now approach 100%. Direct-acting antivirals have resulted in greater access to treatment, improved treatment completeness, and faster time from diagnosis to treatment, making HCV cure theoretically possible for everyone living with HCV.^8,9 Community-based efficacy trials have found SVR rates of 85% to 90% among historically undertreated populations.^10,11 Although studies^12,13,14,15 among PWID show a lower SVR range (20%-95%), US and international guidelines^12,16,17 recommend treatment for all adults living with chronic HCV infection, including people actively using substances.

Nevertheless, HCV treatment uptake among PWID in the US is low. A study¹¹ in Seattle, Washington, a city with high levels of HCV testing, found that far fewer PWID and people experiencing homelessness initiated HCV treatment, similar to other US studies.^18,19,20 This finding is unsurprising given that marginalized populations, including PWID, face a myriad of substantial and well-documented^2,11,21,22 systemic-, practitioner-, and individual-level barriers to HCV treatment in clinical care settings.²¹

Simplified HCV treatment algorithms have the potential to facilitate DAA treatment by non-HCV specialists (ie, general internal medicine physicians, clinical nurse practitioners, and pharmacists) and streamline the HCV care cascade. Several innovative approaches have been adopted to address these barriers and lower the threshold for DAA initiation among PWID. For example, colocating or coordinating HCV treatment within existing clinical care programs can be effective, but this treatment only reaches those already engaged in the health care system. Colocation within trust-established settings in the community can ease access barriers and reduce stigma and discrimination.^23,24 However, at each step along the traditional path from HCV diagnosis to treatment initiation, patients may be lost to follow-up. Providing HCV treatment in a trusted community setting at the time of diagnosis could facilitate improved treatment uptake and reduce loss to follow-up for difficult-to-link populations. The No One Waits (NOW) study aimed to evaluate the feasibility, acceptability, and safety of providing HCV treatment at the point of diagnosis in a nonmedical community setting to people experiencing social marginalization.

Methods

Patients signed written informed consent forms for this nonrandomized controlled trial. The study was approved by the University of California, San Francisco Institutional Review Board. A copy of the protocol can be found in Supplement 1. This study followed the Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) reporting guideline.

Study Design

The NOW study is a nonrandomized, single-arm, open-label, phase 4 study evaluating the feasibility, acceptability, and safety of a point-of-diagnosis HCV treatment model at a nonmedical community site in San Francisco. From July 1, 2020, to October 31, 2021, the study team screened participants for HCV infection using the OraQuick HCV Rapid Antibody (Ab) test (OraSure Technologies) and for HIV infection using the Alere Determine HIV-1/2 Ag/Ab Combo (Abbott Laboratories) with a fingerstick blood sample. Those who tested HIV positive without a prior diagnosis were referred to clinical care. For participants with a reactive HCV Ab result, on-site venipuncture was performed for HCV RNA quantification and reflex genotype testing as well as hepatitis B surface antigen (HBsAg) testing. Participants returned in 1 week for in-person HCV RNA results and were offered same-day initiation of HCV treatment if their HCV RNA was detectable and they met eligibility criteria. Rather than being randomized, participants self-selected to receive the intervention (starting HCV treatment at the HCV diagnosis visit). The study team provided medication through an investigator-initiated grant from Gilead Sciences and dispensed it to participants. Consented participants were given a 14-day starter pack of 400 mg of sofosbuvir and 100 mg of velpatasvir and took their first dose on the day of enrollment (Figure 1).

Figure 1. CONSORT Diagram.

HCV indicates hepatitis C virus; ITT, intention-to-treat; PP, per-protocol.

^aIncludes active hepatitis B virus (n = 1), cirrhosis and prior treatment failure (n = 1), previous HCV treatment longer than 2 weeks (n = 4), and other clinical exclusion (n = 3), including high-dose proton pump inhibitor use, treatment outside the study, and incomplete blood draw.

Study Participants

Participants were identified through (1) flyers and study cards posted in neighborhoods and venues frequented by PWID, (2) referrals from service organizations, (3) street outreach to PWID and people experiencing homelessness, and (4) lists of participants in the investigators’ previous research studies who gave contact permission for future research. Eligibility was assessed and consent obtained at 2 time points: (1) HCV screening and (2) HCV treatment (Figure 2). For HCV screening, eligible participants were 18 years or older and reported either injecting drugs in their lifetime or having a blood transfusion in or before 1992. Participants with confirmed HCV viremia were assessed for eligibility at their HCV RNA results disclosure. Clinical exclusion criteria included the following: positive HBsAg result, untreated HIV, history of hepatic decompensation, prior treatment with an nonstructural 5A (NS5A)–inclusive DAA regimen for more than 2 weeks (unless there was evidence of prior SVR12 with HCV reinfection), pregnancy or breastfeeding, current medication use incompatible with sofosbuvir-velpatasvir or sofosbuvir-velpatasvir-voxilaprevir, or a life expectancy of less than 12 months as assessed by the study clinician.

Figure 2. Study Flow.

Ab indicates antibody; Ag, antigen; EMR, electronic medical record; HBsAg, hepatitis B surface antigen; HCV, hepatitis C virus; NOW, No One Waits; and PA, prior authorization.

^aExclusion criteria were direct-acting antiviral experience, untreated HIV, HBsAg positivity, or decompensated cirrhosis.

^bPretreatment laboratory tests included complete blood cell count, comprehensive metabolic panel, international normalized ratio, anti–hepatitis B core Ab, and anti–hepatitis B surface Ab; a pretreatment sample was collected for later nonstructural 5A RAS testing if visit 1 and 2 testing revealed fibrosis-4 index greater than 3.25 and genotype 3.

^cExclusion criteria were as follows: albumin, less than 3.0 g/dL (to convert to grams per liter, multiply by 10); hemoglobin, less than 8.0 g/dL (women) or less than 9.0 g/dL (men) (to convert to grams per liter, multiply by 10); platelets, less than 50 × 10³/μL (to convert to ×10⁹/L, multiply by 1); creatinine clearance (estimated by Cockcroft-Gault equation), less than 30 mL/min/1.73 m² (to convert to milliliters per second per meters squared, multiply by 0.0167); aspartate aminotransferase or alanine aminotransferase, more than 10 times the upper limit of normal (ULN); total bilirubin, more than 1.5 times the ULN; or international normalized ratio, more than 1.5 times the ULN.

^dEvery 2 weeks during treatment, week 4 after treatment, and week 12 after treatment.

Late study exclusion criteria were considered based on pretreatment laboratory test results (drawn at study entry): albumin, less than 3.0 g/dL (to convert to grams per liter, multiply by 10); hemoglobin, less than 8.0 g/dL (women) or less than 9.0 g/dL (men) (to convert to grams per liter, multiply by 10); platelets, less than 50 × 10³/μL (to convert to ×10⁹/L, multiply by 1); creatinine clearance (estimated by Cockcroft-Gault equation), less than 30 mL/min/1.73 m² (to convert to milliliters per second per meters squared, multiply by 0.0167); aspartate aminotransferase or alanine aminotransferase, more than 10 times the upper limit of normal (ULN); total bilirubin, more than 1.5 times the ULN; or international normalized ratio, more than 1.5 times the ULN. For participants with genotype 3 and suspected cirrhosis based on a ifbrosis-4 index greater than 3.25, NS5A resistance–associated substitution testing was performed using a pretreatment sample to assess for the presence of the Y93H mutation; if detected, patients were switched to sofosbuvir-velpatasvir-voxilaprevir. Persons not meeting eligibility criteria or who declined enrollment received harm-reduction counseling and referrals to community-based medical clinics for HCV and medical care needs. Data on self-reported racial and ethnic identity were collected as part of broader sociodemographic characteristics.

Study Enrollment and Follow-Up Procedures

After providing written informed consent, participants completed a research questionnaire, underwent a blood draw for pretreatment laboratory testing (complete blood cell count, comprehensive metabolic panel, and prothrombin time; hepatitis B surface antibody and hepatitis B core antibody if no prior clinical documentation), received clinical assessment via staff-assisted telemedicine, received a 14-day sofosbuvir-velpatasvir starter pack, and were observed taking the first sofosbuvir-velpatasvir dose by the staff. The study pharmacy team submitted documentation to authorize participants’ transition to insurance-provided sofosbuvir-velpatasvir treatment. If insurance-provided medication was delayed or not authorized, participants received study-issued sofosbuvir-velpatasvir until insurance provided sofosbuvir-velpatasvir was attained or for the duration of treatment.

The study team scheduled follow-up visits every 2 weeks during the 12 weeks of treatment and at 4 and 12 weeks after treatment. At each follow-up study visit, participants picked up medication and completed research questionnaires. For patients who missed follow-up study visits, medication pickup or delivery was arranged. Clinical assessments were conducted via staff-assisted telemedicine at the end of treatment. Additionally, participants were offered clinical consultations with a clinician at treatment weeks 4 and 8 and 12 weeks after treatment. For all telemedicine visits, staff set up the telemedicine technology on a computer provided by the study and introduced the clinician. The telemedicine visit was led by a physician or clinical nurse practitioner.

Enrollment and follow-up questionnaires collected demographic characteristics, physical and emotional health, current living context, health care access, sexual and drug use behaviors, and medication adherence. Study blood draws were obtained at the treatment midpoint and completion and at least 12 weeks after treatment. We approach research participation as a form of specialized work and consulted our community advisory board to determine suitable compensation.^25,26 Participants received reimbursement ranging from $20 to $60 per visit, with a total potential compensation of up to $325 for the 10 study visits.

Research Site

The study was conducted at (1) a fixed community space located in a neighborhood close to public transit and social service organizations, in an area where PWID were known to congregate, and (2) a mobile medical van parked in the neighborhood with the highest percentage of African American and Black people in San Francisco and few city-supported social or medical service organizations. In 2018 and 2019, African American and Black individuals made up approximately one-quarter of HCV cases reported in San Francisco despite representing only approximately 5% of the overall population.²⁷ Study procedures at the mobile site were discontinued in January 2021 because of COVID-19, and participants were rolled over to the fixed site. Both sites had private spaces for interviews and clinical research activities. Wrap-around services, including food, harm-reduction supplies, and COVID-19 information and vaccination, were also available. Site staff were trained in HIV and HCV testing and disclosure counseling, harm reduction, ethical human research practices, and deescalation methods. Study materials were refined in partnership with local advocacy organizations and members of the target population.

End Point Measures

Acceptability, feasibility, and safety of HCV treatment were assessed for all participants who started HCV treatment (defined as receiving at least 1 dose of study drug) at the HCV status disclosure visit. The primary treatment feasibility outcome measure was treatment response, defined as undetectable HCV RNA at least 12 weeks after treatment (SVR12). We calculated SVR12 by intention-to-treat (ITT) group, defined as those who initiated treatment, and per-protocol (PP) group, defined as those who completed a 12-week course of DAA therapy. Participants with undetected HCV RNA after treatment completion who subsequently developed HCV viremia with an HCV genotype different from their pretreatment genotype within 12 weeks of DAA completion were considered to have posttreatment reinfection within 12 weeks and were classified as not attaining SVR12. Secondary treatment feasibility end points were undetectable HCV RNA at treatment completion. Primary acceptability end points included treatment initiation at time of HCV RNA result disclosure and treatment completion. Safety end points were treatment discontinuation because of late exclusion criterion or adverse event.

Statistical Analysis

All acceptability and safety results and secondary outcomes were summarized with descriptive statistics. Treatment initiation was evaluated among all participants with current HCV infection who met the eligibility criteria and returned for their HCV RNA results disclosure visit. Treatment response was evaluated among those who initiated treatment (ITT group, n = 87) and those who completed DAA therapy (PP group, n = 69). Point estimates and 2-sided 95% CIs were also calculated for primary end points. Significance assessments were conducted using appropriate statistical tests, with P values calculated using either the Wilcoxon rank-sum test for continuous or ordinal data or χ² tests and Fisher exact test for categorical data, based on the nature of the variables and data distribution. A 2-sided P < .05 was considered statistically significant. All analyses were performed using Stata software, version 15 (StataCorp).²⁸

Results

Study Sample Profile and Trial Flow

Of the 492 people who underwent anti-HCV testing, 246 (50%) tested anti-HCV positive and 111 (23%) were eligible to enroll. Of the 381 ineligible individuals, most were HCV uninfected (anti-HCV negative [n = 246] or HCV RNA negative [n = 102]), with fewer exclusions attributable to untreated HIV (n = 3), other clinical criteria (n = 12), and incomplete blood draw (n = 18). Eighteen individuals with anti-HCV Ab positivity were unable to get HCV RNA confirmatory testing, and 3 participants could not receive end-of-treatment HCV RNA testing because of blood draw challenges. Eighty-nine of the 111 eligible individuals (80%) returned for HCV RNA results disclosure. Among these, 2 (2%) declined treatment via a research study and sought care at a community clinic, 87 (98%) accepted and initiated treatment at point of diagnosis, 82 (94%) had lifetime injection drug use, and 5 (6%) had blood transfusion before 1992 (Figure 1). The median time from anti-HCV testing to treatment start was 7 days (IQR, 6, 14 days).

Baseline demographic characteristics of those enrolled were statistically similar to those screened (Table 1). The median age was 48 years (IQR, 37-58 years); 62 (71%) were male and 25 (29%) female; 22 (25%) were African American or Black; 5 (6%) were Asian, Native Hawaiian, Pacific Islander, or multiracial; 9 (10%) were Latino/a/e/x; 49 (56%) were White; and 2 (2%) had no race or ethnicity specified. Fifty-three participants (61%) reported sleeping on the street (homelessness) in the past year, and 83 (97%) had an income below the national poverty line. Thirty-six (41%) were unemployed, 6 (7%) were looking for work, and 34 (39%) were unable to work because of disability. Seventy (80%) reported current injection drug use, and approximately half reported food insecurity in the past 3 months. Forty-five participants (52%) could travel to the study site in less than 15 minutes, and 70 (80%) in 30 minutes or less; 46 (53%) could walk, and 15 (17%) took public transit. Seventy-eight participants (94%) had health insurance at enrollment, of whom 74 (95%) had government-provided public insurance (Table 2). Forty-one (48%) reported having a primary care physician. Ten (11%) had previously received HCV treatment, resulting in SVR; of these, 8 (80%) were treated with all oral DAAs (all with evidence of reinfection) and 2 (20%) with an interferon-based regimen.

Table 1. Sociodemographic Characteristics Overall and by SVR12 Response Group^a.

Characteristic	Overall (N = 87)	Attained SVR12 (n = 58)	Did not attain SVR12 (n = 29)	P valueb
Gender
Female	25 (29)	18 (31)	7 (24)	.62
Male	62 (71)	40 (69)	22 (76)
Race and ethnicity
African American and Black	22 (25)	15 (26)	7 (24)	.64
Asian, Native Hawaiian, Pacific Islander, or multiracial	5 (6)	2 (3)	3 (34)
Latino/a/e/x	9 (10)	7 (12)	2 (7)
White	49 (56)	32 (55)	17 (58)
Not specified	2 (2)	2 (3)	0
Age, median (IQR), y	48 (37-58)	50 (41-59)	45 (34-57)	.13
Slept outside or in vehicle in past 12 mo
Yes	53 (61)	29 (50)	24 (81)	.003
No	34 (39)	29 (50)	5 (9)
Current housing circumstance
Outdoors or vehicle	37 (43)	21 (36)	16 (55)	.02
Shelter	8 (9)	3 (5)	5 (17)
SRO or hotel	15 (17)	14 (24)	1 (3)
With friend or family	3 (3)	3 (5)	0
Treatment or transitional	9 (10)	5 (9)	4 (14)
Rent or own	15 (17)	12 (21)	3 (10)
Income below the national poverty line
Yes	83 (97)	50 (86)	26 (96)	.71
No	3 (3)	7 (14)	2 (4)
Current main daily activities or responsibilities
Full-time work	4 (5)	4 (7)	0	.28
Part-time work	6 (7)	5 (9)	1 (3)
Unemployed or laid off	36 (41)	20 (34)	16 (55)
Looking for work	6 (7)	4 (7)	2 (7)
Unable to work because of disability	34 (39)	25 (43)	9 (31)
Any injection drug use in past 3 mo
Yes	70 (80)	48 (83)	22 (76)	.44
No	17 (20)	10 (17)	7 (24)
Worried food would run out before money for more (past 3 mo)
Always true	13 (15)	8 (14)	5 (17)	.79
Sometimes true	31 (37)	22 (38)	9 (31)
Never true	40 (48)	26 (45)	14 (48)
The food I bought just wouldn’t last and I didn’t have more money to buy more (3 mo)
Always true	11 (13)	7 (13)	4 (15)	.94
Sometimes true	32 (38)	22 (39)	10 (36)
Never true	41 (49)	27 (48)	14 (49)
Time it took to get to study location, min
<15 min	45 (52)	30 (52)	15 (54)	.34
15-30	25 (29)	19 (33)	6 (21)
>30	15 (17)	8 (15)	7 (25)
Mode of travel to study location
Walking	46 (53)	NA	NA	NA
Public transit	15 (17)	NA	NA
Personal, family, or friend’s vehicle	0	NA	NA
Taxi or ride share	2 (2)	NA	NA
Bicycle	4 (5)	NA	NA
Other	0	NA	NA

Abbreviations: NA, not applicable; SRO, single-room occupancy; SVR12, sustained virologic response at posttreatment week 12 or later.

^a Data are presented as number (percentage) of study participants unless otherwise indicated. Variations in cell numbers reflect nonresponse to specific questions; missing data were excluded from the analysis. Percentages reported are based on the total responses reported.

^b P values were calculated using the Wilcoxon rank-sum test for continuous or ordinal data or the χ² test or Fisher exact test for categorical data.

Table 2. Medical Service Engagement and HCV Characteristics Overall and by SVR12 Response Group.

Characteristic	No. (%) of study participantsa			P value
	Overall (N = 87)	Attained SVR12 (n = 58)	Did not attain SVR12 (n = 29)
Insurance status at time of enrollment
Yes	78 (94)	53 (91)	25 (98)	.14
No	5 (6)	4 (7)	1 (2)
Type of insurance
Medicare, Medi-Cal, or SFHP	74 (95)	50 (94)	24 (98)	.59
Veterans Administration	1 (1)	0	1 (2)
Blue Cross	2 (3)	2 (3)	0
Kaiser	1 (1)	1 (2)	0
Primary care practitioner
Yes	41 (48)	31 (54)	10 (36)	.12
No	45 (52)	27 (46)	18 (64)
Currently taking prescribed medication for other physical or mental or emotional health needs
Yes	47 (54)	11 (19)	1 (4)	.054
No	40 (46)	47 (81)	27 (96)
No. of ED visits in past 3 mo
0	61 (71)	41 (76)	15 (57)	.14
1	17 (20)	10 (18)	10 (39)
≥2	8 (8)	3 (6)	1 (4)
Previous HCV treatment
No	77 (89)	51 (88)	26 (90)	>.99
Yes	10 (11)	7 (12)	3 (10)
HCV infection characteristics and treatment responses
HCV genotype
1	1 (1)	0	1 (3)	.58
1a	45 (52)	29 (50)	16 (57)
1b	6 (7)	4 (7)	2 (6)
2	6 (7)	5 (9)	1 (3)
3	20 (23)	15 (26)	5 (17)
Hybridization consistent with multiple genotypes, or unable to genotype	10 (9)	5 (9)	4 (14)
Response at treatment week 4
Yes (HCV RNA not detected)	38 (44)	34 (58)	4 (14)	<.001
No (HCV RNA detected)	35 (40)	19 (33)	16 (55)
Unknownb	14 (16)	5 (9)	9 (31)
Response at treatment completion
Yes (HCV RNA not detected)	61 (70)	57 (98)	4 (14)	<.001
No (HCV RNA detected)	5 (6)	0	5 (17)
Unknownc	21 (24)	1 (2)	20 (69)

Abbreviations: ED, emergency department; HCV, hepatitis C virus; SFHP, San Francisco Health Plan; SVR12, sustained virologic response at posttreatment week 12 or later.

^a Variations in cell numbers reflect nonresponse to specific questions; missing data were excluded from the analysis. Percentages reported are based on the total responses reported.

^b P values were calculated using the Wilcoxon rank-sum test for continuous or ordinal data or the χ² test or Fisher exact test for categorical data.

^c Because of loss to follow-up or incomplete blood draw.

HCV Cascade of Care

Among the 126 individuals with HCV viremia, 111 (88%) were eligible and 89 (80%) returned for their HCV RNA positive disclosure visit (Figure 3). Eighty-seven eligible participants (98%) initiated treatment, of whom 69 (79%) completed treatment. Two participants withdrew because of poor adherence, and 16 were lost to follow-up during the treatment period. Thirty-eight of the 73 participants (52%) with RNA tests at week 4 had undetectable HCV RNA. Of the 66 with complete blood draw at the end of treatment, 61 (92%) had undetectable HCV RNA at treatment completion and 58 (95%) attained SVR12. Of note, among the 4 with undetectable HCV RNA at treatment completion and quantifiable HCV RNA at 12 weeks after treatment completion, 2 were assessed to have reinfection because of achievement of SVR4 and different posttreatment genotypes. Four participants were lost to follow-up between end of treatment and SVR12.

Figure 3. Hepatitis C Virus (HCV) Cascade of Care.

The proportions of people in each step of the cascade from the people in the preceding step are shown next to the arrows between each bar. SVR12 indicates sustained virologic response at posttreatment week 12 or later.

^aThree patients with untreated HIV, 12 with other clinical criteria, 18 with unsuccessful blood draws for RNA were excluded.

^bTwo people declined enrollment into study.

^cFour people with incomplete blood draws at end of treatment.

^dTwo people with presumed reinfection after treatment.

Primary and Secondary End Points

Of the 87 treated patients, 58 achieved SVR12, leading to a treatment response of 67% (95% CI, 56%-76%) among the ITT group and 84% among the PP group (58 of 69; 95% CI, 73%-92%). A significantly greater number of people who experienced homelessness in the past 12 months did not attain SVR12 compared with those who did not experience homelessness (24 [83%] vs 5 [17%], P = .003). A significantly greater proportion of those who had undetectable HCV RNA results at 4 weeks (34 [64%] vs 4 [20%], P < .001) and at the end of treatment points (57 [100%] vs 4 [44%], P < .001) attained SVR12 compared with those who did not attain SVR12 (Table 2).

A total of 87 eligible participants started treatment at diagnosis (98%; 95% CI, 96%-99.5%); of those who started treatment, 69 (79%; 95% CI, 69%-87%) completed treatment and 61 (70%; 95% CI, 59%-79%) had undetectable HCV RNA at the end of treatment. There were no adverse events, including death, to cause participants to prematurely discontinue treatment. Notably, no participants’ treatment course was stopped or altered because of late study exclusion criteria based on pretreatment laboratory test results.

Discussion

The NOW study demonstrates that offering HCV treatment at the point of diagnosis in a nonclinical community setting is feasible and acceptable among a highly marginalized population that included a majority of participants who were current PWID, were recently experiencing homelessness, and had incomes below the federal poverty level. Nevertheless, all but 2 people started HCV treatment on the same day as their HCV viremia disclosure. Treatment at the point of diagnosis reduces the likelihood of loss to follow-up by not requiring another visit and may reduce forward transmission of HCV by accelerating the time to attaining cure. The NOW model of care can expand the current toolkit of test-and-treat models to reach a broader population of marginalized communities—a critical next step to achieving the national HCV elimination goals.²⁹

Documented barriers to health care and clinical trial participation among socially marginalized populations include financial, transportation, social isolation issues, fear of stigma and discrimination from practitioners, lack of social support, insufficient insurance, limited understanding of health care systems, and concerns about adherence and drug use among participants.^30,31 We hypothesized that providing HCV treatment in a nonclinical community setting at the point of diagnosis would alleviate many of these barriers and improve HCV treatment uptake. Indeed, 98% of eligible participants initiated HCV treatment at the point of diagnosis. The 2 who did not enroll opted instead to start treatment via a nonresearch mobile clinical program after an engaged referral from study staff. Our findings demonstrate that the crucial step between linkage from diagnosis to treatment can be bridged with the NOW model.

The NOW model of care shortened the steps between testing and treatment initiation. The American Association for the Study of Liver Diseases/Infectious Diseases Society of America simplified HCV treatment algorithm requires 4 to 5 steps between testing and treatment initiation.³² The NOW model of care truncated the process to 2 steps.³² We demonstrated several important points regarding this approach. First, of the 126 individuals with HCV viremia identified at the screening stage, 111 (88%) were eligible for the simplified HCV treatment algorithm. Second, pretreatment laboratory test samples were drawn on the same day as DAA initiation, allowing for late exclusion based on these laboratory test results, yet no participants required treatment adjustments because of exclusionary safety laboratory test results or the presence of genotype 3 with cirrhosis. Thus, although pretreatment laboratory tests play an important role in the clinical assessment of patients with HCV, especially for estimating fibrosis stage and determining post-SVR hepatocellular carcinoma surveillance needs, these assessments can coincide with DAA initiation. The likelihood of selecting for or enriching RAS variants if a patient stops treatment or treatment fails is higher the longer the patient is exposed to the NS5A inhibitor. The risk of RASs with 2 weeks or less of sofosbuvir-velpatasvir is presumed to be lower than in patients who relapse after a full treatment course. Third, despite a streamlined care cascade, 22 (20%) of those eligible for point-of-diagnosis treatment did not return for their HCV RNA results. This finding highlights the critical need for a commercially available point-of-care HCV RNA test. Although we cannot know whether these individuals would have opted for same-day DAA treatment, other studies^33,34 suggest that point-of-care HCV RNA testing embedded in a simplified care model improves DAA uptake. Finally, 18 individuals with anti-HCV Ab positivity were unable to get HCV RNA confirmatory testing, and 3 participants could not receive end-of-treatment HCV RNA testing because of blood draw challenges. This underscores the need for a fingerstick viral load platform for diagnosis, confirmation of cure, and screening for reinfection.

Sixty-seven percent of the ITT group participants and 84% of the PP group participants achieved SVR12. These proportions are consistent with other treatment trials,^35,36,37 demonstrating that once enrolled, PWID can successfully complete treatment and achieve SVR. In the HERO (Patient-Centered Models of HCV Care for People Who Inject Drugs) study, 755 PWID recruited at opioid treatment programs and community health centers were randomized to receive HCV treatment with patient navigation vs modified directly observed therapy.³⁵ Of the 87 treated patients, 58 achieved SVR12, leading to a treatment response of 67% among the ITT group and 84% among the PP group. We achieved comparable SVR12 rates in a nonclinical setting, demonstrating that our model is feasible in a community space that is not colocated with clinical services. Indeed, treatment acceptability and adherence were also high as measured by the proportion of eligible participants who started and completed treatment. Treatment initiation was more common in the NOW study model,^36,37 and point-of-diagnosis treatment was tolerated and safe, with no deaths or other serious adverse events.

Several facilitators to our study’s success were identified and could be replicated in other contexts. The community-based nonclinical location was selected because it was close to public transit and familiar for study participants. Placing study sites in areas where participants were known to live and/or congregate likely contributed to successful recruitment and retention. Because a large proportion of our participants reported both food insecurity and continued injection drug use, providing wrap-around services, including food support and harm-reduction supplies, also likely facilitated participation. Retention efforts included street outreach visit reminders. Recruitment materials were carefully vetted and designed for this particular population. Providing 2-week starter packs of medication at the start of treatment helped address insurance-related treatment start delays.

Limitations

Our study has several limitations. First, it was conducted in a city with a strong focus on harm-reduction health care services. Participants were compensated monetarily for attending study visits, which may limit the generalizability of our findings to nonresearch settings. Additionally, the study is situated in a state with Medicaid expansion, ensuring that nearly all participants had health benefits. This situation contrasts with what might be observed in a marginally housed population in a nonexpansion state. As such, participation and recruitment success and feasibility results may not be generalizable to other geographic locations. In addition, the study took place during the COVID-19 pandemic, and study procedures had to be adjusted to accommodate COVID-19 safety guidelines and shelter-in-place mandates, which may have affected participants’ behavior, willingness to participate, and follow-up. It was recognized in the early months of the epidemic that marginalized populations, including those experiencing homelessness and PWID, were highly affected and vulnerable. Additional attention and COVID-19–specific resources were directed to serve these populations³⁸ and may have also influenced acceptability and study outcomes during this unique time. Also of note, the NOW study was launched when HCV testing and treatment were halted globally because of COVID-19.³⁹ Nevertheless, in 15 months we conducted HCV testing in nearly 500 at-risk individuals and initiated treatment in 87. With additional resources dedicated to improving HCV diagnosis, access to DAA treatment, and public health capacity to address HCV, as outlined in the US National Hepatitis C Elimination Program,²⁹ our model can be implemented in community-based settings and other areas of colocalized services in other cities.

Conclusions

In this nonrandomized controlled trial, we found that point-of-diagnosis HCV treatment initiation among PWID and people experiencing homelessness was feasible and acceptable and yielded high SVR12 rates. Our trial was conducted in a nonclinical community space with procedures specifically designed to reflect the environment and services needed to provide HCV testing, disclosure, and treatment to a socially marginalized population. A similar approach could be used in other nonclinical settings and same-day HCV test-and-treat models.

Supplement 1.

Trial Protocol

Supplement 2.

Data Sharing Statement