Abstract
Background
The objective of this study was to evaluate the effectiveness of formal hepatitis C virus (HCV) education on engagement in therapy in persons experiencing homelessness in an on-site shelter-based model of care. As policies to eliminate Medicaid access restrictions to HCV treatment are expanded, patient education is paramount to achieving HCV elimination targets in difficult-to-engage populations including persons experiencing homelessness.
Methods
This prospective study was conducted at 4 shelters in San Francisco and Minneapolis from August 2018 to January 2021. Of the 162 HCV Ab–positive participants, 150 participated in a 30-minute HCV education session. Posteducation changes in knowledge, beliefs, barriers to care, and willingness to accept therapy scores were assessed.
Results
Following education, knowledge scores (mean change, 4.4 ± 4.4; P < .001) and willingness to accept therapy (70% to 86%; P = .0002) increased. Perceived barriers to HCV care decreased (mean change, –0.8 ± 5.2; P = .001). Higher baseline knowledge was associated with lesser gain in knowledge following education (coef., –0.7; P < .001). Posteducation knowledge (odds ratio, 1.2; P = .008) was associated with willingness to accept therapy.
Conclusions
An HCV educational intervention successfully increased willingness to engage in HCV therapy in persons experiencing homelessness in an on-site shelter-based HCV model of care.
Keywords: hepatitis C, direct-acting antiviral therapy, substance use, HCV education, homelessness
Hepatitis C virus (HCV) infection has undergone epidemiologic shifts in the last decade [1]. In 2016, the World Health Assembly pledged to eliminate HCV as a public health threat with a goal of 90% reduction in HCV incidence and 65% reduction in HCV mortality by 2030. As part of this pledge, core intervention strategies have focused on harm reduction for people who inject drugs through education campaigns [2]. In the United States, HCV and injection drug use disproportionately affect people experiencing homelessness (PEH) [3]. HCV prevalence among PEH ranges from 9.8% to 52%, vs 1.3% in the general population [3–5]. Experiencing homelessness is associated with known risk factors for HCV and transmission, such as substance use, injection drug use, syringe sharing, mental illness, and history of incarceration [6–9].
PEH are not only at increased risk for acquisition and transmission of HCV, but also encounter heightened barriers to care including perceived stigma and medical mistrust and competing priorities such as securing housing or employment [6, 10–12]. Barriers to care and lack of awareness of HCV have led to both underdiagnoses and undertreatment in this patient population [13, 14]. As policies to eliminate Medicaid access restrictions for HCV treatment across states are enacted, access to care for PEH is projected to be cost-effective and paramount to HCV elimination goals [15, 16]. Strategies to strengthen the hepatitis C cascade of care in PEH have included implementation of screening and linkage-to-care protocols as well as shelter-based direct antiviral therapy in this population [17–19]. As direct antiviral therapy becomes possible in a shelter-based model for PEH, identification of barriers to care is critical to successful outcomes.
Prior studies have identified low baseline HCV knowledge scores among PEH [20]. Studies conducted during the interferon era of HCV therapy have demonstrated the critical role of education in vulnerable populations by improving HCV knowledge, enhancing HCV care coordination, expediating HCV treatment, and improving antiviral response [21–23]. In the era of direct-acting anti-HCV therapy, key principles identified in the interferon era should be evaluated when creating new models of care for vulnerable populations including PEH. In 1 study where greater than one-third of the study participants were experiencing homelessness, a 15-minute HCV educational intervention improved knowledge scores regarding HCV transmission, disease progression, and expected treatment outcomes [24].
Our prior work in 4 geographically diverse shelters in San Francisco and Minneapolis supports the critical need for implementation of an HCV education intervention as part of a shelter-based model of care for HCV-infected PEH [12, 25]. When interviewing shelter clients, we identified misconceptions and limited HCV knowledge including themes surrounding screening, transmission, and treatment, as well as social stigma against PEH [12]. Prior theoretical models such as the health behavior framework (HBF) provide useful tools for guiding exploration of the complex interplay of factors that can help shape HCV disease understanding and management. The efficacy of the HBF as a guide for community-level intervention has been validated among multiple communities, including those at risk of viral hepatitis infection [26]. This framework is a multidimensional model that synthesizes multiple models of health behavior, social theory, and change to map factors that influence intentions [27]. The HBF assumes that individual, support system, and community cues influence health behavior, for example, willingness to accept HCV therapy. In this study, we used the established HBF constructs to assess HCV knowledge, beliefs, and perceived barriers to HCV care across geographically diverse shelter populations and evaluated the impact of a comprehensive formal HCV educational intervention on changes in these parameters and willingness to accept HCV treatment.
METHODS
Population
This prospective study was conducted by a multidisciplinary team at 4 large homeless shelters, 2 in San Francisco, California, and 2 in Minneapolis, Minnesota, from August 1, 2018, to January 30, 2021 [28]. Shelters for PEH included in this study provide support services on a daily basis to >600 residents in San Francisco and between 170 and 350 residents in Minneapolis. Following informed consent, adults ≥18 years of age seeking shelter services with the ability to read and write in English, and who were either treatment naïve or had not received prior HCV treatment within the prior 12 weeks, were enrolled. Patients with significant medical or psychiatric conditions that prevented participation in the study were excluded. For this analysis, participants who were hepatitis C antibody (HCV Ab) positive (based on point-of-care testing or serologic testing) were included.
Study Intervention
Participants meeting study eligibility criteria completed a formal 30-minute HCV education session using a standardized Power Point slide format presented by a designated HCV coordinator (RN, PharmD, or advanced practice provider). Education was provided on-site and offered to participants on the same day as screening or at another convenient time and included up to 10 participants. The content of our education session consisted of information related to HCV epidemiology, transmission, diagnosis, natural history, HCV therapy, and vaccination for hepatitis A and hepatitis B, as well as resources for substance use disorder therapy (alcohol and injection drug use), syringe access services, and linkage to liver disease care. In addition, efforts were made to specifically address previously identified potential common perceived barriers and misconceptions about HCV infection and treatment [12]. A pre-education questionnaire and a posteducation questionnaire were completed to assess the participants’ demographic and substance use characteristics, risk factors for HCV, and internalized HCV stigma, as well as knowledge of, attitudes toward, and barriers to HCV care. Participants were given a $25 incentive for completion of the HCV education and questionnaires.
Patient Consent
Institutional review board approvals were obtained from the University of California San Francisco and Hennepin Healthcare Human Subjects Research Committee, and all participants provided written consent.
Questionnaire Design and Measures
The questionnaire instrument was developed using the Health Behavior Framework with input from expert hepatologists and behavioral scientists experienced in health behavior change research and information from published studies in patients with hepatitis and substance use disorders [12, 21, 27, 29–31]. In addition, the rigorous questionnaire development process also included engagement with shelter stakeholders and participants experiencing homelessness through qualitative interviews [12, 25]. The final questionnaire items were then organized into 3 domains: (1) HCV knowledge, (2) beliefs about HCV infection, and (3) barriers to HCV care (Supplementary Table 1). We then performed the Cronbach’s alpha test to assess for internal consistency and the scale reliability coefficient (alpha) of each of these domains in our questionnaire instrument. Cronbach’s alpha test score for the HCV knowledge domain was 0.9, for the beliefs domain it was 0.8, and for the barriers domain it was 0.95, indicating good to excellent scale reliability of our questionnaire.
Data Analysis
Descriptive analyses of patient characteristics were performed to obtain frequency (%) for categorical variables and median (interquartile range [IQR]) or mean (SD) for continuous variables. We then compared patient characteristics between patients who had detectable and undetectable HCV RNA. Categorical variables were compared using the chi-square (χ2) test or the Fisher exact test (if appropriate), while continuous variables were compared using the Mann-Whitney test.
Questionnaire responses were summarized using frequencies and percentages. Composite scores for each of the domains and subdomains were calculated from responses to questions designed to assess these factors as follows: (Domain 1) Knowledge score was computed as the number of correct responses to 21 questions (1 for correct, 0 for incorrect or do not know; max score 21); (Domain 2) Beliefs about HCV infection score were determined by summing numerical codes (1 or 0) assigned to the responses for the corresponding subdomains: perceived severity (max score of 5), stigma (max score of 5), treatment efficacy (max score of 1), perceived susceptibility to disease risk (max score of 3); (Domain 3) Barriers to HCV treatment (max score of 21) was assessed using a Likert Scale and coded as 0 for extremely confident, 1 for moderately confident, 2 for slightly confident, and 3 for not confident, with higher scores representing higher perceived barriers. Willingness to accept therapy was also assessed using a 5-point Likert scale and coded as 1 for extremely likely and 0 for not at all. Mean scores between pre-education and posteducation were examined using the Wilcoxon signed-rank test.
Univariate and multivariate forward linear regression (for linear scale) and logistic regression (for dichotomous scale) modeling were used to assess factors associated with 2 outcomes following educational intervention: (1) change in knowledge (linear scale) and (2) willingness to accept hepatitis C treatment (dichotomous scale: 1 = extremely likely, 0 = all other categories). Predictors included those selected a priori. The multivariable model for the outcome of change in knowledge was adjusted for age, sex, race, and education and included baseline predictors with a P value <.05 on forward regression. The multivariable model for the outcome of willingness to accept hepatitis C treatment was adjusted for age, sex, and race and included both baseline predictors along with posteducation predictors of knowledge, beliefs, and barriers with a P value <.05 on forward regression. All analyses were performed using Stata 15 statistical software (Stata Corp LP, College Station, TX, USA).
RESULTS
Participants
Availability of HCV testing sessions was advertised in the shelters, and in addition 1199 shelter clients were directly approached for HCV Ab testing. A total of 772 clients agreed to participate in the study, of whom 766 were deemed eligible. Of the 766 participants, a total of 162 were HCV antibody positive and eligible to receive HCV education. Of these,155 agreed to receive education and completed the pre-education questionnaire (99 in San Francisco and 56 in Minneapolis), and 150 received education and completed the posteducation questionnaire. One participant who did not complete a pre-education questionnaire completed a posteducation questionnaire. Table 1 summarizes patient characteristics (n = 155). The median age of the participants was 56, and 75% were men. Nearly two-thirds of the study population identified as racial/ethnic minorities. The education level among the study population varied, with 25% completing less than a high school education. The overall median length of homelessness was 18 months. With respect to substance use, >80% of the participants had used illicit drugs in the past year, one-third had heavy alcohol consumption, and approximately two-thirds had a history of injection drug use or had received substance use disorder therapy in the past. Nearly 60% had a history of psychiatric illness, most participants reported having a health care provider (78.7%), and nearly 90% were publicly insured. Overall, two-thirds of the participants had a detectable HCV RNA. There was a significant difference in the proportion of participants with detectable HCV RNA by shelter location: 57.3% were from San Francisco, and 42.7% were from Minneapolis. A higher proportion of those with undetectable HCV RNA had a history of psychiatric illness compared with those who had detectable HCV RNA (70.0% vs 52.9%), but this did not quite reach statistical significance (P = .054). There were no other significant differences in other participant characteristics by HCV RNA status (Table 1).
Table 1.
Demographics of Participants Completing Pre-education Questionnaire
| Total (n = 155) |
Detectable HCV RNA (n = 103) |
Undetectable HCV RNA (n = 52) |
P Valuea | |
|---|---|---|---|---|
| Age, median, y | 56.1 | 55.7 | 56.6 | .6 |
| [IQR], y | [49.7–63] | [48.8–62.7] | [51–63.2] | |
| (range), y | (21.2–82.1) | (21.2–82.1) | (31.7–75.4) | |
| Male, No. (%) | 116 (74.8) | 80 (77.7) | 36 (69.2) | .3 |
| Race, No. (%) | n = 154 | n = 51 | .1 | |
| White, non-Hispanic | 60 (39.0) | 46 (44.7) | 14 (27.5) | |
| Black/African American | 64 (41.6) | 41 (39.8) | 23 (45.1) | |
| Hispanic | 14 (9.1) | 8 (7.8) | 6 (11.8) | |
| Other | 16 (10.4) | 8 (7.8) | 8 (15.7) | |
| Education level, No. (%) | n = 154 | n = 102 | .2 | |
| Less than HS | 38 (24.7) | 22 (21.6) | 16 (30.8) | |
| High school | 60 (39.0) | 45 (44.1) | 15 (28.9) | |
| More than high school/GED | 56 (36.4) | 35 (34.3) | 21 (40.4) | |
| Shelter location, No. (%) | .02 | |||
| San Francisco | 99 (63.9) | 59 (57.3) | 40 (76.9) | |
| Minneapolis | 56 (36.1) | 44 (42.7) | 12 (23.1) | |
| Length of homelessness | n = 149 | n = 99 | n = 50 | .4 |
| Median, mo | 18.0 | 17.2 | 24.0 | |
| [IQR], mo | [5.0–60.0] | [5.0–60.0] | [8.0–60.0] | |
| Illicit drug use within the past year, No. (%) | n = 154 130 (84.4) |
n = 102 84 (82.4) |
46 (88.5) | .4 |
| History of injection drug use ever, No. (%) | n = 152 101 (66.5) |
n = 102 67 (65.7) |
34 (68) | .9 |
| History of substance use disorder therapy, No. (%) | n = 150 95 (63.3) |
n = 99 63 (63.6) |
n = 51 32 (62.8) |
1.0 |
| Alcohol use within the past year, No. (%) | n = 153 | n = 101 | n = 52 | 1.0 |
| None/minimal | 65 (42.5) | 42 (41.6) | 23 (44.2) | |
| Moderate | 37 (24.2) | 25 (24.8) | 12 (23.1) | |
| Heavy | 51 (33.3) | 34 (33.7) | 17 (32.7) | |
| HIV co-infection, No. (%) | n = 147 6 (4.1) |
n = 101 5 (5.0) |
n = 46 1 (2.2) |
.7 |
| History of psychiatric illness, No. (%) | n = 152 89 (58.6) |
n = 102 54 (52.9) |
n = 50 35 (70.0) |
.054 |
| Has a health care provider, No. (%) | n = 141 111 (78.7) |
n = 98 80 (81.6) |
n = 43 31 (72.1) |
.3 |
| Insurance type, No. (%) | n = 145 | n = 96 | n = 49 | 1.0 |
| Public | 129 (89.0) | 85 (88.5) | 44 (89.8) | |
| Private | 6 (4.1) | 4 (4.2) | 2 (4.1) | |
| Uninsured | 10 (6.9) | 7 (7.3) | 3 (6.1) |
Categorical variables were compared using the chi-square (χ2) test or Fisher exact test (if appropriate), while continuous variables were compared using the Mann-Whitney test (P < .05 considered significant).
Bolded values in the table reflect statistically significant values where P < .05.
Abbreviations: HCV, hepatitis C virus; IQR, interquartile range.
Change in HCV Knowledge, Beliefs, and Perceived Barriers to HCV Care and Willingness to Accept Therapy Following Education
The mean baseline knowledge score before education was 12.4. The most common (>50% of respondents) knowledge deficit was related to HCV treatment, followed by modes of transmission of HCV. Following education, there was an increase in knowledge score to 16.9, with the mean change in score following education being 4.4 points (P < .001) (Table 2). When evaluating knowledge scores across shelter sites, San Francisco participants had higher baseline HCV knowledge scores compared with Minneapolis participants (mean score ± SD, 13.3 ± 4.3 vs 10.9 ± 4.9; P < .001), and the mean change in knowledge score following education was higher among Minneapolis participants compared with San Francisco participants (mean score ± SD, 5.4 ± 4.2 vs 3.9 ± 4.4; P = .04). With respect to beliefs about HCV and barriers to HCV care, the most common (>50% of respondents) belief was related to feelings of self-blame or blame by others for having HCV infection, and the most commonly perceived barriers to HCV care was keeping doctors’ appointments and adherence to HCV medication when actively using substances (alcohol or drug use). For all participants, in beliefs about HCV infection there was an increase in perceived severity (mean score, 3.7 to 4.6 out of max score 5; P < .001), perceived susceptibility to disease risk (mean score, 2.6 to 2.9 out of max score 3; P < .001), and treatment efficacy (0.7 to 0.9 out of max score 1; P < .001) following education. There was a reduction in perceived barriers to HCV treatment following education (pre-education mean score, 4.6; posteducation mean score, 3.6 out of max score 21; P = .001). The reduction in stigma following education did not reach statistical significance (2.9 to 2.6 out of max score 5; P = 1.0) (Table 3). Before education, the proportion of participants who were extremely likely to accept HCV therapy if needed was 70.8%. Following education, the proportion was 86% (P = .0002).
Table 2.
Knowledge, Beliefs, and Barrier Scores Before and After Education
| Pre-education | Posteducation | Mean Change | P Value | ||||
|---|---|---|---|---|---|---|---|
| No. | Mean ± SD | No. | Mean ± SD | No. | Mean ± SD |
||
| Domain 1: Knowledge | 150 | 12.4 ± 4.7 | 146 | 16.9 ± 3.4 | 141 | 4.4 ± 4.4 | <.001 |
| Domain 2: Beliefs about HCV infection | |||||||
| (a) Perceived severity | 152 | 3.7 ± 1.3 | 134 | 4.6 ± 0.6 | 130 | 0.9 ± 1.3 | <.001 |
| (b) Stigma | 152 | 2.9 ± 1.5 | 148 | 2.6 ± 1.5 | 145 | 0.0 ± 1.3 | 1.0 |
| (c) Treatment efficacy | 154 | 0.7 ± 0.5 | 150 | 0.9 ± 0.3 | 148 | 0.2 ± 0.5 | <.001 |
| (d) Perceived susceptibility to disease risk | 154 | 2.6 ± 0.8 | 145 | 2.9 ± 0.3 | 143 | 0.3 ± 0.8 | <.001 |
| Domain 3: Barriers to HCV treatment | 151 | 4.6 ± 6.3 | 146 | 3.6 ± 5.7 | 142 | –0.8 ± 5.2 | .001 |
Bolded values in the table reflect statistically significant values where P < .05.
Abbreviation: HCV, hepatitis C virus.
Table 3.
Univariable and Multivariable Analysis of Factors Associated With Change in Knowledge Score Following Education
| Univariable Analysis | Multivariable Analysis (n = 137) | ||||||
|---|---|---|---|---|---|---|---|
| No. | Coef. | 95% CI | P Value | Coef. | 95% CI | P Valuea | |
| Age | 141 | 0.02 | –0.05 to 0.08 | .6 | 0.01 | –0.5 to 0.5 | 1.0 |
| Female sex | 141 | –1.5 | –3.2 to 0.12 | .07 | –0.5 | –1.8 to 0.7 | .4 |
| Race, No. (%) | 140 | ||||||
| White, non-Hispanic | Ref. | ||||||
| Black/African American | 0.4 | –1.3 to 2.0 | .7 | –1.1 | –2.4 to 0.10 | .1 | |
| Hispanic | 2.6 | –0.3 to 5.4 | .08 | 0.1 | –1.9 to 2.1 | .9 | |
| Other | –0.7 | –3.2 to 1.8 | .6 | –0.8 | –2.6 to 1.0 | .4 | |
| Education level, No. (%) | 140 | ||||||
| Less than HS | Ref. | ||||||
| High school | 0.5 | –1.4 to 2.4 | .6 | 0.3 | –0.9 to 1.5 | .6 | |
| More than high school/GED | 0.2 | –1.8 to 2.1 | .9 | ||||
| Shelter location, No. (%) | 141 | — | — | — | |||
| San Francisco | Ref. | ||||||
| Minneapolis | 1.5 | 0.01 to 3.0 | .048 | ||||
| Length of homelessness, mo | 136 | –0.001 | –0.01 to 0.01 | .8 | — | — | — |
| Illicit drug use within the past year, No. (%) | 140 | –0.9 | –2.9 to 1.2 | .4 | — | — | — |
| History of injection drug use ever, No. (%) | 140 | –0.9 | –2.9 to 1.2 | .4 | — | — | — |
| History of substance use disorder therapy, No. (%) | 138 | –0.1 | –1.6 to 1.4 | .9 | 1.4 | 0.3 to 2.5 | .01 |
| Alcohol use within the past year, No. (%) | 139 | — | — | — | |||
| None/minimal/moderate | Ref. | — | — | — | |||
| Heavy | 1.1 | –0.7 to 3.0 | .2 | ||||
| 2.0 | 0.4 to 3.7 | .018 | |||||
| HIV co-infection, No. (%) | 135 | –2.9 | –6.7 to 1.0 | .1 | — | — | — |
| History of psychiatric illness, No. (%) | 139 | –1.1 | –2.6 to 0.4 | .2 | — | — | — |
| Has a health care provider, No. (%) | 128 | –0.05 | –2.0 to 1.9 | 1.0 | — | — | — |
| Domain 1: Baseline knowledge | 141 | –0.7 | –0.8 to –0.6 | <.001 | –0.7 | –0.8 to –0.6 | <.001 |
| Domain 2: Baseline | |||||||
| Beliefs About HCV infection | |||||||
| (a) Perceived severity | 139 | –1.2 | –1.6 to –0.7 | <.001 | — | — | — |
| (b) Stigma | 138 | 0.2 | –0.3 to 0.7 | .4 | — | — | — |
| (c) Treatment efficacy | 141 | –2.5 | –4.1 to –1.0 | .001 | — | — | — |
| (d) Perceived susceptibility to disease risk | 141 | –2.6 | –3.4 to –1.8 | <.001 | — | — | — |
| Domain 3: Baseline | |||||||
| Barriers to HCV treatment | 138 | 0.004 | –0.1 to 0.1 | .9 | — | — | — |
Abbreviation: HCV, hepatitis C virus.
Bolded values in the table reflect statistically significant values where P < .05.
Mean scores were compared using Wilcoxon signed-rank test (P < .05 considered significant).
Baseline Factors Associated With Change in Knowledge
On univariate analysis, shelter site (Minneapolis vs San Francisco; coef., 1.5; 95% CI, 0.01 to 2.9; P = .048) and heavy alcohol use within the past year (coef., 2.0; 95% CI, 0.4 to 3.7; P = .02) were associated with higher change in knowledge. Baseline knowledge score (coef., –0.7; 95% CI, –0.78 to –0.56; P < .0001), baseline scores of perceived severity (coef., –1.2; 95% CI, –1.6 to –0.7; P < .001), treatment efficacy (coef., –2.6; 95% CI, –4.1 to –1.0; P = .001), and perceived susceptibility to disease risk (coef., –2.6; 95% CI, –3.4 to –1.8; P < .001) were associated with lesser change in knowledge (Table 3).
On multivariate analysis, when adjusting for age, sex, race/ethnicity, and education, higher baseline knowledge score was associated with lesser change in knowledge (coef., –0.7; 95% CI, –0.8 to –0.6; P < .0001), and a history of substance use disorder therapy was associated with greater change in knowledge (coef., 1.4; 95% CI, 0.3 to 2.5; P = .01) (Table 3). To address any possible differential influence of HCV RNA–positive and –negative status on change in knowledge, a sensitivity analysis was performed that limited the analysis to HCV RNA–positive participants and showed that baseline HCV knowledge was the only independent predictor of change in knowledge (coef., –0.74; 95% CI, –0.96 to –0.51; P < .0001) following education on adjusted analysis.
Factors Associated With Willingness to Accept HCV Treatment
On univariate analysis, posteducation knowledge score (OR, 1.2; 95% CI, 1.0 to 1.3; P = .008) was positively associated with a willingness to accept HCV treatment, and posteducation stigma score (OR, 0.7; 95% CI, 0.5 to 0.95; P = .03) and posteducation barrier score (OR, 0.9; 95% CI, 0.8 to 0.97; P = .006) were negatively associated with a willingness to accept HCV treatment (Table 4).
Table 4.
Univariable and Multivariable Analysis of Factors Associated With Willingness to Accept Therapy Following Education: San Francisco and Minneapolis, Aug 2018–Jan 2021
| Univariable Analysis | Multivariable Analysis (n = 144) | ||||||
|---|---|---|---|---|---|---|---|
| No. | OR | 95% CI | P Value | OR | 95% CI | P Value | |
| Age | 150 | 0.8 | 0.5 to 1.2 | .3 | 1.0 | 0.5 to 1.6 | .9 |
| Female sex | 150 | 2.2 | 0.5 to 1.2 | .2 | 2.1 | 0.5 to 9.2 | .3 |
| Race, No. (%) | 149 | ||||||
| White, non-Hispanic | Ref. | ||||||
| Black/African American | 0.5 | 0.2 to 1.5 | .2 | 0.7 | 0.2 to 2.4 | .6 | |
| Hispanic | 1.3 | 0.1 to 12.3 | .7 | 1.7 | 0.2 to 18.2 | .6 | |
| Other | 0.5 | 0.1 to 2.2 | .4 | 0.5 | 0.1 to 2.8 | .5 | |
| Education level, No. (%) | 149 | ||||||
| Less than HS | Ref. | ||||||
| High school | 2.0 | 0.6 to 7.3 | .3 | — | — | — | |
| More than high school/GED | 1.0 | 0.3 to 3.0 | 1.0 | — | — | — | |
| Shelter location, No. (%) | 150 | ||||||
| San Francisco | Ref. | ||||||
| Minneapolis | 0.9 | 0.3 to 2.4 | .9 | — | — | — | |
| Length of homelessness, mo | 144 | 1.0 | 0.99 to 1.0 | .5 | — | — | — |
| Illicit drug use within the past year, No. (%) | 149 | 0.5 | 0.1 to 2.5 | .4 | — | — | — |
| History of injection drug use ever, No. (%) | 147 | 1.2 | 0.4 to 3.2 | .7 | — | — | — |
| History of substance use disorder therapy, No. (%) | 145 | 1.8 | 0.7 to 4.8 | .2 | — | — | — |
| Alcohol use within the past year, No. (%) | 148 | ||||||
| None/minimal | Ref. | ||||||
| Moderate | 2.1 | 0.5 to 8.1 | .3 | — | — | — | |
| Heavy | 1.1 | 0.4 to 3.2 | .8 | — | — | — | |
| HIV co-infection, No. (%) | 143 | 0.7 | 0.1 to 6.3 | .7 | — | — | — |
| History of psychiatric illness, No. (%) | 148 | 0.8 | 0.3 to 2.1 | — | — | — | |
| Has a health care provider, No. (%) | 137 | 1.5 | 0.5 to 4.7 | .5 | — | — | — |
| Domain 1: Posteducation knowledge score | 147 | 1.2 | 1.04 to 1.3 | .008 | 1.2 | 1.0 to 1.3 | .02 |
| Domain 2: Posteducation | |||||||
| Beliefs About HCV infection score | |||||||
| (a) Perceived severity | 134 | 1.8 | 0.9 to 3.8 | .1 | — | — | — |
| (b) Stigma | 148 | 0.7 | 0.5 to 0.95 | .03 | 0.7 | 0.5 to 0.96 | .03 |
| (c) Treatment efficacy | 150 | 0.9 | 0.2 to 4.4 | .9 | — | — | — |
| (d) Perceived susceptibility to disease risk | 145 | 1.6 | 0.3 to 6.7 | .5 | — | — | — |
| Domain 3: Posteducation barriers to HCV treatment | 146 | 0.9 | 0.8 to 0.97 | .006 | — | — | — |
Bolded values in the table reflect statistically significant values where P < .05.
Abbreviations: HCV, hepatitis C virus; OR, odds ratio.
On multivariate analysis, when adjusting for age, sex, and race/ethnicity, the posteducation knowledge score was associated with 1.2 times greater odds of willingness to accept therapy (OR, 1.2; 95% CI, 1.03 to 1.3; P = .02), and higher posteducation stigma score was associated with lower odds of willingness to accept therapy (OR, 0.7; 95% CI, 0.5 to 0.95; P = .03) (Table 4). When including the posteducation barrier score in the final model, while the direction of the estimated effect of both posteducation knowledge and stigma scores remained the same, the estimated effect size of the posteducation knowledge score decreased slightly, and the P value was no longer statistically significant (OR, 1.1; 95% CI, 0.98 to 1.3; P = .1). In a sensitivity analysis that was limited to HCV RNA–positive participants, posteducation HCV knowledge was the only independent predictor of willingness to accept therapy, with the same effect size as the whole cohort (OR, 1.21; 95% CI, 1.00 to 1.46; P = .049) following education on adjusted analysis.
DISCUSSION
This is the first study to evaluate the impact of formal HCV education on willingness to accept therapy in an on-site shelter-based HCV model of care for PEH. In this study, we conducted a targeted HCV educational intervention for this high-risk population. We demonstrated the effectiveness of an educational intervention in scores across all questionnaire domains of knowledge, beliefs about HCV infection, and barriers to HCV treatment. We showed that while HCV knowledge increased and barriers to care decreased following education, the presence of stigma regarding HCV did not change with education and was associated with decreased willingness to accept therapy. Conversely, posteducation knowledge score was the only predictor of increased willingness to accept therapy that was independent of perceived barriers to HCV care and beliefs about HCV, including perceived HCV stigma.
Our education session was specifically designed to encompass all aspects of HCV disease including prevention and management given our prior evaluation of knowledge gaps in shelter stakeholder and client perspectives [12, 25], as well as prior patient HCV interventions in at-risk populations [20–24]. Baseline knowledge scores across our participant population were low, which is consistent with prior literature that assessed vulnerable populations and included participants experiencing homelessness [20, 22, 24]. These gaps in knowledge highlight the imperative of a patient-centered education pillar as part of a comprehensive HCV shelter-based model of care. Indeed, our study population was high-risk for active HCV infection and part of marginalized groups associated with lower linkage to HCV care due to barriers in accessing care. The majority of our study participants had a history of illicit drug use, active heavy alcohol use, and a history of psychiatric illness [9, 32]. Despite this, HCV knowledge increased across all groups following our formal HCV education.
In addition, our study examined 2 geographically distinct regions. Though there were no differences in participant education levels across shelter sites, participants from Minneapolis had lower baseline HCV knowledge than participants from San Francisco. The baseline differences in knowledge may reflect public health efforts in the respective cities. For example, San Francisco launched the EndHepC SF campaign in 2016, allocating public funds for testing, treatment, and education regarding HCV infection [33]. Participants in both sites, however, had an increase in mean knowledge score following education.
We identified specific baseline factors that were independently associated with change in knowledge. As expected, having a higher baseline knowledge score was associated with lesser gain in knowledge post education. This likely reflects that education has a greater influence in enhancing knowledge among those with lower baseline knowledge scores. Interestingly, on univariate analysis heavy alcohol use was associated with higher change in knowledge. This subpopulation is uniquely vulnerable given the known heightened risks of alcohol consumption in cirrhosis [34]. In addition, prior literature highlights a knowledge gap in patients with heavy alcohol consumption and disease risk and may demonstrate potential for greater change in knowledge following education in this population [35]. While heavy alcohol use and other substance use were not independent predictors of change in knowledge in our multivariable analysis, a history of substance therapy was associated with a higher posteducation knowledge score. This suggests that these clients may significantly benefit from integration of formal HCV education in substance use programs and thus represents a patient-centered opportunity to enhance HCV care in this population.
We have shown that higher posteducation knowledge is associated with greater odds of willingness to accept HCV therapy, emphasizing the importance of formal education in linkage to HCV care, and consequently HCV elimination efforts in PEH. Stigma represents an additional component in health behavior frameworks where perceived stigma is a determinant of health behaviors [36]. In our study, a higher stigma score despite education was negatively associated with willingness to accept HCV therapy. Public health interventions in this area aimed at positive outcomes of treatment should directly address patients’ perceived and self-stigma. This may enhance willingness to accept therapy and prevent liver disease progression. Willingness to undergo therapy may be enhanced by integration of HCV care on-site at shelters. Previous studies have identified that the likelihood of linkage to care increases for PEH when they have an assigned shelter bed for the evening before an appointment and that PEH also cite distance and cost of traveling to the clinic as predominant barriers to HCV care [18]. Although we also observed that higher posteducation perceived barrier scores were associated with lower odds of willingness to accept HCV therapy, this was no longer statistically significant when accounting for other factors including posteducation HCV knowledge and stigma scores.
Our study has several limitations. Despite inclusion of 2 geographically distinct regions, our findings may not be generalizable to other populations of PEH. About three-quarters of our population had completed at least a high school education, and HCV education was conducted in English. In addition, participants who engaged in HCV education may have been more motivated than those who did not receive HCV education, which may have influenced their willingness to subsequently engage in HCV therapy. While the assessment of the internal consistency of the domains in our questionnaire instrument showed good to excellent consistency, further studies are needed to validate the use of this questionnaire in other settings. Nonetheless, our study demonstrated the efficacy of an educational intervention across participants of diverse geographic and racial/ethnic backgrounds.
In summary, we provide evidence for the effectiveness of formal HCV education to enhance willingness to engage in HCV therapy as part of an HCV shelter model of care for PEH. Formal HCV education is a key component of the comprehensive HCV model of care delivery based on the health behavior framework [27]. Here, we demonstrate that formal HCV education is critical in engaging a particularly vulnerable and high-risk population of patients in HCV therapy as part of a novel, comprehensive, and on-site shelter-based HCV model of care for PEH.
Supplementary Material
Acknowledgments
The investigators would like to thank participating shelter staff, providers, and leadership, as well as GLIDE’s HIV/Hep C and Harm Reduction Services program (San Francisco), the San Francisco Department of Public Health, Ms. Katie Burke, MPH, from the San Francisco Department of Public Health Population Health Division, and Minneapolis Healthcare for the Homeless.
Financial support. This work was supported by an investigator-initiated grant through Gilead Sciences Inc. (IN-US-342-4531; to M.K.), National Institutes of Health K24AA022523 (to M.K.), University of California San Francisco Liver Center P30 DK026743, and National Institutes of Health T32DA007250 (to J.K.F.).
Potential conflicts of interest. J.P. is a recipient of a research grant from Gilead Sciences Inc., and he has served on an advisory board for Gilead Sciences Inc. M.K. is a recipient of a research grant (to her institution) from Gilead Sciences Inc. and Intercept Pharmaceuticals, and she has served as consultant for Gilead Sciences Inc. All other authors declare no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Author contributions. D.P. participated in study design, data analysis, and interpretation, wrote the manuscript, and approved the final submission. J.P., M.R., J.N., C.Mag., B.Z., C.Mas., and J.K.F. contributed to study design, collected data, provided material support, edited the manuscript, and approved the final submission. D.G. contributed to data clean-up and data analysis, contributed to the manuscript write-up, edited the manuscript, and approved the final submission. M.K. designed the study, provided material support, reviewed the statistical analyses and data interpretation, edited the manuscript, and approved the final submission.
Prior presentation. The 70th Annual American Association for the Study of Liver Diseases The Liver Meeting 2019, Hynes Convention Center, Boston, MA; Poster Number: A735.
References
- 1. Suryaprasad AG, White JZ, Xu F, et al. Emerging epidemic of hepatitis C virus infections among young nonurban persons who inject drugs in the United States, 2006-2012. Clin Infect Dis 2014; 59:1411–9. [DOI] [PubMed] [Google Scholar]
- 2. World Health Organization. Global health sector strategy on viral hepatitis 2016-2021. Available at: https://www.who.int/publications/i/item/WHO-HIV-2016.06. Accessed 2 November 2020.
- 3. Bradley H, Hall EW, Rosenthal EM, Sullivan PS, Ryerson AB, Rosenberg ES.. Hepatitis C virus prevalence in 50 U.S. states and D.C. by sex, birth cohort, and race: 2013-2016. Hepatol Commun 2020; 4:355–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hofmeister MG, Rosenthal EM, Barker LK, et al. Estimating prevalence of hepatitis C virus infection in the United States, 2013-2016. Hepatology 2019; 69:1020–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Beijer U, Wolf A, Fazel S.. Prevalence of tuberculosis, hepatitis C virus, and HIV in homeless people: a systematic review and meta-analysis. Lancet Infect Dis 2012; 12:859–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Strehlow AJ, Robertson MJ, Zerger S, et al. Hepatitis C among clients of health care for the homeless primary care clinics. J Health Care Poor Underserved 2012; 23:811–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Aisyah DN, Shallcross L, Hayward A, et al. Hepatitis C among vulnerable populations: a seroprevalence study of homeless, people who inject drugs and prisoners in London. J Viral Hepat 2018; 25:1260–9. [DOI] [PubMed] [Google Scholar]
- 8. Arum C, Fraser H, Artenie AA, et al. Homelessness, unstable housing, and risk of HIV and hepatitis C virus acquisition among people who inject drugs: a systematic review and meta-analysis. Lancet Public Health 2021; 6:e309–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Nyamathi AM, Dixon EL, Robbins W, et al. Risk factors for hepatitis C virus infection among homeless adults. J Gen Intern Med 2002; 17:134–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Gelberg L, Gallagher TC, Andersen RM, Koegel P.. Competing priorities as a barrier to medical care among homeless adults in Los Angeles. Am J Public Health 1997; 87:217–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sublette VA, Smith SK, George J, McCaffery K, Douglas MW.. Listening to both sides: a qualitative comparison between patients with hepatitis C and their healthcare professionals’ perceptions of the facilitators and barriers to hepatitis C treatment adherence and completion. J Health Psychol 2017; 22:1300–11. [DOI] [PubMed] [Google Scholar]
- 12. Masson CL, Fokuo JK, Anderson A, et al. Clients’ perceptions of barriers and facilitators to implementing hepatitis C virus care in homeless shelters. BMC Infect Dis 2020; 20:386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Noska AJ, Belperio PS, Loomis TP, O’Toole TP, Backus LI.. Engagement in the hepatitis C care cascade among homeless veterans, 2015. Public Health Rep 2017; 132:136–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Chhatwal J, Chen Q, Bethea ED, Hur C, Spaulding AC, Kanwal F.. The impact of direct-acting anti-virals on the hepatitis C care cascade: identifying progress and gaps towards hepatitis C elimination in the United States. Aliment Pharm Ther 2019; 50:66–74. [DOI] [PubMed] [Google Scholar]
- 15. Schackman BR, Gutkind S, Morgan JR, et al. Cost-effectiveness of hepatitis C screening and treatment linkage intervention in US methadone maintenance treatment programs. Drug Alcohol Depend 2018; 185:411–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Greenwald R, Waters P, Cayer S.. Enforcement of legal remedies to secure hepatitis C virus treatment with direct-acting antiviral therapies in correctional facilities and Medicaid programs. Public Health Rep 2020; 135:44S–9S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Benitez TM, Fernando SM, Amini C, Saab S.. Geographically focused collocated hepatitis C screening and treatment in Los Angeles’s skid row. Dig Dis Sci 2020; 65:3023–31. [DOI] [PubMed] [Google Scholar]
- 18. Fuster D, Gelberg L.. Community screening, identification, and referral to primary care, for hepatitis C, B, and HIV among homeless persons in Los Angeles. J Community Health 2019; 44:1044–54. [DOI] [PubMed] [Google Scholar]
- 19. Gelberg L, Robertson MJ, Arangua L, et al. Prevalence, distribution, and correlates of hepatitis C virus infection among homeless adults in Los Angeles. Public Health Rep 2012; 127:407–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Tyler D, Nyamathi A, Stein JA, Koniak-Griffin D, Hodge F, Gelberg L.. Increasing hepatitis C knowledge among homeless adults: results of a community-based, interdisciplinary intervention. J Behav Health Serv Res 2014; 41:37–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Surjadi M, Torruellas C, Ayala C, Yee HF, Khalili M.. Formal patient education improves patient knowledge of hepatitis C in vulnerable populations. Dig Dis Sci 2011; 56:213–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lubega S, Agbim U, Surjadi M, Mahoney M, Khalili M.. Formal hepatitis C education enhances HCV care coordination, expedites HCV treatment and improves antiviral response. Liver Int 2013; 33:999–1007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Gupta K, Romney D, Briggs M, Benker K.. Effects of a brief educational program on knowledge and willingness to accept treatment among patients with hepatitis C at inner-city hospitals. J Community Health 2007; 32:221–30. [DOI] [PubMed] [Google Scholar]
- 24. Norton BL, Voils CI, Timberlake SH, et al. Community-based HCV screening: knowledge and attitudes in a high risk urban population. BMC Infect Dis 2014; 14:74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Fokuo JK, Masson CL, Anderson A, et al. Recommendations for implementing hepatitis C virus care in homeless shelters: the stakeholder perspective. Hepatol Commun 2020; 4:646–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Maxwell AE, Bastani R, Glenn BA, et al. Developing theoretically based and culturally appropriate interventions to promote hepatitis B testing in 4 Asian American populations, 2006–2011. Prev Chronic Dis 2014; 11:130245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Bastani R, Glenn BA, Taylor VM, et al. Integrating theory into community interventions to reduce liver cancer disparities: the health behavior framework. Prev Med 2010; 50:63–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Khalili M, Powell J, Park HH, et al. Shelter-based integrated model is effective in scaling up Hepatitis C testing and treatment in persons experiencing homelessness. Hepatol Commun 2022; 6:50–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Mukhtar NA, Evon DM, Yim C, et al. Patient knowledge, beliefs and barriers to hepatitis B Care: results of a multicenter, multiethnic patient survey. Dig Dis Sci 2021; 66:434–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Balfour L, Kowal J, Corace KM, et al. Increasing public awareness about hepatitis C: development and validation of the Brief Hepatitis C Knowledge Scale. Scand J Caring Sci 2009; 23:801–8. [DOI] [PubMed] [Google Scholar]
- 31. Fife BL, Wright ER.. The dimensionality of stigma: a comparison of its impact on the self of persons with HIV/AIDS and cancer. J Health Soc Behav 2000; 41:50–67. [PubMed] [Google Scholar]
- 32. Nielsen SF, Hjorthøj CR, Erlangsen A, Nordentoft M.. Psychiatric disorders and mortality among people in homeless shelters in Denmark: a nationwide register-based cohort study. The Lancet 2011; 377:2205–14. [DOI] [PubMed] [Google Scholar]
- 33. Facente SN, Grebe E, Burk K, et al. Estimated hepatitis C prevalence and key population sizes in San Francisco: a foundation for elimination. PLoS One 2018; 13:e0195575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Younossi ZM, Zheng L, Stepanova M, Venkatesan C, Mir HM.. Moderate, excessive or heavy alcohol consumption: each is significantly associated with increased mortality in patients with chronic hepatitis C. Aliment Pharmacol Ther 2013; 37:703–9. [DOI] [PubMed] [Google Scholar]
- 35. Elliott JC, Hasin DS, Des Jarlais DC.. Perceived health and alcohol use in individuals with HIV and hepatitis C who use drugs. Addict Behav 2017; 72:21–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Ferrer R, Klein WM.. Risk perceptions and health behavior. Curr Opin Psychol 2015; 5:85–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.