Study design and setting
The original study, the INTRO-HCV trial, was designed as a multi-center, randomized controlled trial [19]. This study evaluated fatigue as a secondary outcome of the INTRO-HCV trial. We recruited PWIDs with chronic HCV infection who were eligible for HCV treatment with DAAs in accordance with Norwegian HCV treatment guidelines (Additional File 1). Participants were recruited from eight outpatient clinics providing opioid agonist therapy (OAT) in Bergen and Stavanger, Norway, as well as two community care centers (CCCs) in Bergen providing primary healthcare to PWIDs. Enrollment was conducted from May 2017 to June 2019. For a more comprehensive description, a published protocol is available [19].
Study sample
A total of 148 participants were randomized into the integrated HCV treatment group and 150 into the standard HCV treatment group (Figure 1). Ultimately, seven participants in the integrated treatment group and 15 in the standard treatment group were excluded due to death or lack of FSS-9 assessments. In total, 276 participants were included in the study – 141 in the integrated treatment group and 135 in the standard treatment group.
Inclusion and exclusion criteria
Inclusion criteria were defined as follows: 1) receiving OAT opioids in the OAT outpatient clinics or people injecting drugs receiving healthcare from the two CCCs; 2) having chronic HCV infection defined as detecting HCV with HCV polymerase chain reaction in two separate blood samples drawn with an interval of at least six months; 3) eligibility for treatment according to the Norwegian HCV treatment guidelines; and 4) willingness to sign a written informed consent to participate in the trial. We excluded people who 1) currently received treatment for HCV; 2) were co-infected with human immunodeficiency virus (HIV) or hepatitis B virus (positive surface antigen) at the time of inclusion; 3) had severe extrahepatic manifestations (e.g., cryoglobulinemia or membranoproliferative glomerulonephritis); 4) had chronic renal disease stages 4–5 (glomerular filtration rate < 30 ml/min/1.73 m2); and 5) had decompensated liver disease (Child-Pugh class B or C). Additionally, people who did not complete the FSS-9 questionnaire during the study period were excluded. For details on demographic and clinical variables, see Table 1.
Randomization and masking
Selected participants were randomized at a 1:1 ratio using blocks of 10 stratified by city and assigned into integrated (n = 148) or standard treatment (n = 150) for the trial. Complete blinding was considered impractical and would have reduced external validity [20], although some masking measures were taken [19]. In short, randomization was disclosed to clinical staff providing treatment and follow-up. Participants were informed of key elements in the delivery of the respective intervention and follow-up to which they were assigned, but no information was shared on treatment and follow-up alternatives or the hypotheses for the study.
Ethics approval and consent to participate
The present study was reviewed and approved by the Regional Ethical Committee for Health Research (REC) West, Norway (reference number: 2017/51/REK Vest, dated 29.03.2017/20.04.2017). All recruited participants were fully informed about the study, and their written informed consent was provided before their inclusion and randomization. All methods were carried out in accordance with relevant guidelines and regulations.
Data collection
Participants were evaluated prior to HCV treatment and 12 weeks after the end of treatment (EOT12) to record their health status, including fatigue level according to the FSS-9 score, sociodemographic data, current drug use, blood samples, transient elastography, and clinical examination. The health assessments were conducted by specialized research nurses in close collaboration with the clinics’ consultants in addiction medicine and infectious diseases. A medical team followed up with those who did not meet the criteria for inclusion in the study. Data from the health assessments prior to and after HCV treatment were defined as the study’s baseline and EOT12 (endpoint), respectively.
Measuring fatigue and drawing blood samples
We assessed fatigue using the FSS-9, including items considering mental and physical functioning, motivation, carrying out duties, and interfering with work, family, or social life. The FSS-9 is a well-known questionnaire to quantify fatigue during the week prior to the assessment [21-26], with high validity and reliability in people undergoing HCV treatment [27]. The FSS-9 items are answered on a Likert scale ranging from 1, no fatigue, to 7, worst fatigue, demonstrating the fatigue level. A high FSS-9 score indicates a high level of fatigue; a mean score greater than 4.0 reveals severe fatigue [26]. The FSS-9 employed had been translated and back-translated from US-English into Norwegian by qualified native Norwegian-speaking translators (Additional File 2) [28].
We drew blood samples, including hepatitis B virus surface antigen, HIV antigen/antibodies, thrombocytes, and aspartate aminotransferase, as well as HCV antibodies and HCV polymerase chain reactions. Liver stiffness was measured by calculating the aspartate aminotransferase to platelet ratio index and performing transient elastography at baseline (Additional File 3). Transient elastography calculates liver stiffness using the median value of ten repeated measurements on an empty stomach [29, 30].
Intervention – standard HCV treatment
Participants in the standard HCV treatment group were referred to the centralized outpatient infectious disease clinic at the collaborating referral hospital for HCV treatment. Their clinical assessment could involve additional blood samples and imaging before initiating HCV treatment. In the first year of the study, HCV consultation with a consultant in infectious diseases was mandatory, but with increasing clinical experience and growing evidence, the primary assessment became voluntary. Participants were offered follow-up assessments, including blood samples, during treatment in the infectious disease outpatient clinic every four weeks. They were responsible for retrieving and adhering to their prescriptions, and attending assessment appointments. At EOT12, blood samples, including HCV polymerase chain reaction, were drawn at infectious disease outpatient clinics, OAT clinics, and CCCs. In addition, participants met at OAT clinics or CCCs to assess their FSS-9 levels.
Intervention – integrated HCV treatment
All assessments and medications for participants in the integrated treatment groups were provided onsite at the OAT clinics or CCCs, including DAAs, blood samples, and FSS-9 assessments. The OAT clinics differed from the CCCs by offering OAT medications in addition to psychosocial approaches. In both settings, nurses and social workers, in cooperation with peer counselors, provided most of the participants’ daily follow-ups. For those eligible for HCV treatment, DAAs were administered by a nurse at OAT clinics/CCCs after a prescription from a consultant in infectious diseases. All HCV treatment and scheduled follow-up during treatment were given in parallel with the observed intake of OAT medications and other care, in line with the study protocol. The number of deliveries of OAT and DAA medications per week was adapted to the level of functioning of each participant. The multidisciplinary team planned assessments with participants, or drop-in approaches were applied.
Statistical analyses
We used Stata SE version 17 (StataCorp, TX, USA) for descriptive analyses and linear mixed model analyses, and IBM SPSS version 26.0 for expectation-maximization calculation. The threshold for statistical significance was set to p < 0.05 for all analyses unless otherwise stated. All statistical analyses were conducted following CONSORT and SPIRIT guidelines [31, 32]. The sample size was calculated for the primary outcome of SVR, defined as undetectable HCV RNA 12 weeks after HCV treatment completion, in the INTRO-HCV trial [19].
We dealt with any missing values in FSS-9 scores at baseline and EOT12 as “missing at random” when running expectation-maximization algorithm [33, 34]. We identified missing values in 1.4% of FSS-9 scores at baseline and 29% at EOT12, and all were replaced with estimated values. The expectation–maximization algorithm for computing data iteratively performs maximum likelihood estimation in the presence of latent variables [35], recommended for optimizing the mixed models. Sensitivity analyses without estimated values were conducted in all regression models.
The FSS-9 sum scores at baseline and EOT12 were calculated by summarizing the points generated by the nine items. We created Pen’s parades in which the FSS-9 sum score at baseline were listed chronologically per participant and spikes were performed to express changes in FSS-9 scores from baseline to EOT12 in the integrated and standard HCV treatment groups. Additionally, linear mixed models were applied to investigate changes in FSS-9 sum scores (outcome variable) from baseline to EOT12, with the interaction between treatment groups (dichotomized as standard (0) versus integrated (1)) and time (dichotomized as baseline (0) and EOT12 (1)) as a predictor variable. The linear mixed models were random intercept fixed slope regression models. The restricted maximum likelihood was set as the estimator [36, 37]. The full information of maximum likelihood ensured that all available FSS-9 sum score were used. The LMM analysis was run as intention-to-treat and per-protocol analyses and as a sensitivity analysis without computed data.