Treatment Adherence and Support for People Who Inject Drugs Taking Direct-Acting Antiviral Therapy for Hepatitis C Infection

Phillip Read; Rosie Gilliver; John Kearley; Rebecca Lothian; Evan B. Cunningham; Karen J. Chronister; Gregory J. Dore


J Viral Hepat. 2019;26(11):1301-1310. 

In This Article



The Kirketon Road Centre (KRC) is a targeted primary healthcare facility in Sydney, Australia, focused on the prevention, treatment and care of HIV, viral hepatitis and sexually transmissible infections amongst a largely marginalized client population of "at risk" young people, sex workers and PWID. A multidisciplinary team provides a comprehensive range of medical, nursing, needle and syringe programme, social care, harm reduction and health education anonymously and free of charge. These services are provided both in-house and through an extensive outreach programme to local services utilized by PWID. KRC also has an integrated on-site "low threshold" OAT dosing programme. It is a highly integrated model of care designed to improve health outcomes of vulnerable populations.[20–22]

Adherence Support Evaluation and Intervention

The level of adherence support for each client was decided between the client and nurse based on social and drug-use stability, the client's ability to store medication and previous success taking an extended course of daily medication. Enhanced adherence support included at least weekly contact, either in-person or by telephone, flexible directly observed dispensing of medications either daily or weekly at KRC (with or without OAT), and liaison with other organizations or delivery of medications to prisons, psychiatric units and hospital wards. Standard of care adherence support included monthly contact and self-management of medications. The acceptability of this enhanced adherence support model to both client and staff has been reported elsewhere.[23]

Study Design and Participants

This was a prospective observational cohort study of all clients who commenced interferon-free DAA therapy for HCV at KRC prior to March 2018. This included all clients prescribed DAAs through the first two years of the Australian government-funded scheme (from March 2016) and a small number of clients who received the same model of care but their DAA therapy was prescribed through early access programmes. The enhanced adherence support sub-population of this cohort consisted of those clients who, following discussion with HCV clinic staff, chose to attend KRC either daily or weekly to receive DAAs. Detailed adherence evaluation was then undertaken on this sub-population. Not included in this sub-population were clients receiving enhanced support for DAA therapy in which KRC facilitated dosing through another service.

Study Variables and Data Sources

All clients attending KRC for clinical services complete a standardized intake survey, including demographic, social and drug-use information. This information along with clinical data was entered into an electronic clinical database. Demographic data included age, gender, Aboriginality and homelessness. A history of mental health concerns included a diagnosis of any affective, psychotic or anxiety-related disorder. Drug-use variables included ever injecting drugs and injecting within the last six months, age of first injection, drugs used in last 12 months, last drug injected, frequency and receptive sharing of injecting equipment. Clinical data included HCV genotype, HIV or hepatitis B co-infection, previous HCV treatment, degree of liver fibrosis (as determined by hepatic elastography, with Fibroscan), treatment regimen and duration, date of initiation and results of testing for sustained virological response (SVR). SVR testing was determined at least 12 weeks after the completion of treatment. SVR results obtained from clients who were considered lost to follow-up and later returned to KRC were included if the client did not have additional treatment elsewhere in the interim.

Additional variables were recorded and calculated for those clients in the enhanced adherence support sub-population. Data included medication dosing, gaps in dosing and daily adherence. Clients who missed dosing days were encouraged to continue attending until treatment was completed and all doses were taken. As such, the number of days that treatment was extended was also recorded. For clients choosing to attend daily for DAA dosing, each day was assigned 1 if the client attended and medication was dispensed or assigned 0 if the client did not attend. For clients choosing to attend weekly for DAA dosing, each day within the seven days of dispensed medication was assigned 1, assuming the client took each daily tablet. If a client did not attend at the end of the seven-day period to collect the next week's medication, each day was assigned 0 until the client attended to collect DAAs for another week. In some cases, clients switched between attending daily and weekly; dosing was recorded as indicated according to the expectation of attendance on the day. Doses provided to another service (eg hospital or prison) for a client were only recorded as taken if the service confirmed adherence; lacking confirmation, it was assumed the client did not receive DAAs on those days. Gaps in dosing were defined as each occasion a client did not attend to collect medication. The length (number of days) of each gap and number of gaps were recorded by the study team after reviewing dosing records. Clients who did not attend for medications for more than 15 days were considered lost to follow-up for the purpose of this analysis. If the client returned for DAAs, a clinical decision was made whether to continue therapy, based on planned treatment duration, the portion which had been completed and the ability of the client to continue or recommence treatment.

Treatment period-based adherence was determined for each of the first 12 weeks of treatment, the first 2 weeks of treatment (days 1 through 14), the second 2 weeks (days 15 through 28) and three 4-week periods (days 1–28, 29–56 and 57–84). Adherence for each period was calculated as the number of doses taken during the period divided by the number of days in the period. Overall adherence was calculated in two ways. First, adherence was calculated for the intended duration of treatment (ie 56, 84 or 168 days) by dividing the number of days doses were taken in that time period by the number of days prescribed. Using this definition of treatment adherence, a patient was determined to be adherent to therapy if they were adherent on at least 90% of the expected treatment days. In cases where treatment was continued beyond the planned treatment duration (due to sub-optimal adherence), a second calculation was performed for adherence by dividing the number of days doses were taken in the total treatment period (planned plus extended) by the intended number of days of dosing. For DAA regimens with divided daily dosing, partial daily adherence was not considered, such that any dose taken equalled adherent on that day. Finally, on-treatment adherence was calculated as the proportion of days in treatment that dosing occurred until permanent discontinuation or loss to follow-up (eg a client who dosed each day of the first four weeks of treatment but then discontinued for the remaining 8 weeks would have overall adherence of 33%, but 100% adherence while actually on-treatment). In all cases, adherence calculations were adjusted for the small number of clients who died (unrelated to treatment) or whose care was transferred to another service to include only those days/weeks when clients were expected to attend KRC for treatment.

Statistical Methods

Data were analysed using descriptive statistics to summarize the entire treatment cohort and further stratified by inclusion in the enhanced adherence support sub-population. Differences between the enhanced adherence support sub-population and the remaining cohort were analysed using Pearson's chi-square test or Fisher's exact tests and binomial tests of significance as appropriate. Data from the enhanced adherence support sub-population were further examined with respect to adherence. The level of support received (eg daily or weekly), treatment outcome (eg completed, transferred elsewhere, lost to follow-up), the number of gaps per client and the client maximum, average and median gap length were summarized with descriptive statistics. The mean difference between overall treatment adherence with and without extension beyond planned treatment duration was compared using a paired t test.

Logistic regression analyses were used to assess factors associated with being adherent to treatment (≥90%). Demographic and drug-use characteristics (ie frequency of injecting, receptive syringe sharing), HIV or hepatitis B coinfection and the number of gaps in treatment during the first (two/four) weeks were also examined. Unadjusted logistic regression was performed to estimate the odds ratio (OR) and 95% confidence intervals (95% CI) to identify which factors were associated with an overall adherence of at least 90% without treatment extension. Variables with P < .20 in the unadjusted analyses were considered for inclusion in a multivariate logistic regression model using a backward stepwise approach yielding adjusted odds ratios (aOR) and 95% CIs. Generalised estimating equation (GEE) methodology was used to investigate the association between time since the initiation of treatment and daily adherence. GEE methods were employed to account for the correlated nature of repeated measurements amongst individual participants. GEE models were specified using a Gaussian family function. ORs with corresponding 95% CIs and P-values were calculated per week of treatment. Statistically significant differences were assessed using two-sided P-values considered significant if P < .05. All analyses were performed using Stata v14.2.


This study was approved by the South Eastern Sydney Local Health District Human Research Ethics Committee (17/POWH/646) and the ethics committee of the Aboriginal Health and Medical Research Council of New South Wales (1339/17).