Adherence to Treatment and Quality of Life during Hepatitis C Therapy
A Prospective, Real-life, Observational Study
Patrick Marcellin; Michel Chousterman; Thierry Fontanges; Denis Ouzan; Michel Rotily; Marina Varastet; Jean-Philippe Lang; Pascal Melin; Patrice Cacoub
Abstract and Introduction
Abstract
Background: Adherence is important for therapy of chronic diseases, but has still not been well studied in real life in chronic hepatitis C.
Aims: To assess adherence to hepatitis C combination therapy in routine clinical practice and to identify factors associated with imperfect adherence.
Methods: This cohort study included unselected chronic hepatitis C patients initiating peginterferon α-2b plus ribavirin. 100% adherence was defined by taking all the prescribed doses of both drugs for the full initially intended duration, as declared by the patient or believed by the physician. Quality of life was assessed using the short-form health survey (SF-36) questionnaire.
Results: 1860 patients were analysed, including 72% treatment-naive, 36% genotype 2/3, 23% psychiatric, 44% drug addicts and 3% human immunodeficiency virus (HIV)-positive patients. Early treatment discontinuation occurred in 30% of patients. Overall, 38% of patients reported 100% adherence. Patient- and physician-reported adherences were discordant, with a 20–30% overestimation by physicians. HIV co-infection [odds ratio (OR) 2.52, 95% confidence interval (CI) 1.36–4.67], no drug use during follow-up (2.37, 1.30–4.31), genotype 3 (1.55, 1.20–2.00) and treatment-naive (1.32, 1.03–1.69) were associated with 100% adherence. Quality of life worsened during treatment but returned to baseline after the end of treatment.
Conclusions: Imperfect adherence to combination therapy is common in routine patients. Adherence is markedly overestimated by physicians and is associated with some patient's baseline characteristics. Knowledge of these factors might help identify patients who are most in need of intervention and plan more frequent and accurate follow-up.
Introduction
The current standard of care for chronic hepatitis C is the antiviral combination therapy with peginterferon α and ribavirin.
[1,2] Peginterferon is administered subcutaneously once a week and ribavirin orally twice daily. Successful treatment, i.e. achievement of a sustained virological response, has been obtained in 79–93% of genotype 2 or 3 hepatitis C virus (HCV) patients after 24 weeks of treatment; as well as in 41–52% of genotype 1 patients and 77% of genotype 4 patients after 48 weeks of treatment in clinical trials.
[3–6] Good adherence to treatment is believed to enhance the rate of sustained virological response, as shown in patients from prior trials who received ≥80% of both their total peginterferon and ribavirin doses for ≥80% of the expected duration of therapy.
[7] However, adverse effects of interferon-based therapy – mainly flu-like symptoms, neuropsychiatric disorders, anaemia and neutropenia – are significant, justify frequent dose reduction of peginterferon and ribavirin and represent one major cause of premature treatment discontinuation.
[8] Shortened antiviral therapy may be considered in some patients depending on genotype, baseline viral load and viral kinetics,
[9–11] but only a minority of patients are concerned in routine practice, at least in France.
These data originate from randomised, controlled clinical trials. However, many patients in routine clinical practice are nonresponders or relapsers to previous interferon-based therapy and the majority present with a wide spectrum of comorbidities. In particular, the high prevalence of psychiatric or substance abuse disorders in patients with chronic HCV infection is well known.
[12] Conversely, people with chronic psychiatric disorders
[13,14] and intravenous drug users
[15–17] are far more likely to be infected with HCV compared with the general population. In addition, one-fourth to one-third of human immunodeficiency virus (HIV)-infected patients are co-infected with HCV.
[18,19] Besides strict contraindications, such diagnoses have constituted the most common reason patients have been denied treatment for HCV infection, not only in clinical trials
[3,4] but also, until recently, in the real life.
[20,21]
We performed a cohort study in chronic hepatitis C patients to evaluate the adherence to peginterferon α-2b and ribavirin combination therapy and to identify factors associated with good adherence in a real-life setting. We also assessed the coherence between patient-reported and physician-reported adherence, and the health-related quality of life.
Patients and Methods
This prospective, multicentre, observational study was carried out in University hospitals, non-University hospitals and private practice offices involved in the management of hepatitis C in France. All patients aged ≥18 years with chronic hepatitis C and initiating combination therapy with peginterferon α-2b and ribavirin were proposed to participate in the survey, whether naive for hepatitis C therapy or not. In accordance with the French law, Ethics Committee's approval was not required as the protocol was strictly observational and usual practice was unchanged. However, all patients gave informed consent to participate. Approval of the 'Commission Nationale de l'Informatique et des Libertés' was obtained, ensuring that patient data were kept confidential according to the French regulation.
As for all observational studies, there were no protocol-specific procedures or study visits. Patients saw their physician as per usual practice in the centre. The investigator and the patient completed an anonymous questionnaire each at inclusion, every 3 months during treatment and 6 months after the end of treatment. Patients filled in their questionnaires in the waiting room and either gave it back to the physician in a sealed envelope or returned it using a prepaid envelope.
Physicians recorded the socio-demographical characteristics, history of hepatitis C, selected comorbidities, process of care, concomitant treatments and adverse events. Comorbidities included past and current psychiatric history, HIV or HBsAg positivity, chronic diseases and use of psychoactive products. History of hepatitis C included source and duration of infection, viral load documented by quantitative PCR (Amplicor™, Roche Diagnostics, France), HCV genotype, liver fibrosis and necro-inflammation (Metavir, FibroTest-ActiTest®,
[22] Knodell score) and previous antiviral treatment. Process of care included the intended antiviral treatment regimen (drug doses and duration), any subsequent treatment modification (dose modification or treatment discontinuation), the reason for modification and patient's therapeutic education. Patients recorded the antiviral medication actually taken and self-assessed their quality of life using the validated French translation of the MOS 36-item short-form health survey (SF-36) questionnaire.
[23] Each SF-36 subscore ranges from 0 (worse) to 100 (best), except health transition, which ranges from 0 (best) to 100 (worse).
OutcomesAt each visit under treatment, patients reported the number of missed peginterferon injections in the past month (the last four injections) or missed ribavirin intakes in the past week (the last 14 intakes). To estimate patient-reported adherence at a given time point, we took into account the 'actual' doses as reported by the patient at each visit and compared it with the most recent physician prescription (thus integrating modifications of prescription that could have occurred over time). To estimate the patient-reported overall adherence, we took into account the adherence estimated at each visit, the actual treatment duration and the treatment duration initially intended by the physician. As a result, patients who declared having taken all the prescribed doses of both products for the full initially intended treatment duration were classified as 100% adherent. Those who declared having taken ≥80% of at least one product for ≥80% of the initially planned duration were classified as ≥80% adherent.
Physician-reported adherence at a given visit and overall was estimated similarly, using the response to the following questions: 'In your opinion, how many times did the patient missed a peginterferon injection in the past month (the last 4 administrations)?' or 'In your opinion, how many times did the patient missed a ribavirin intake in the past week (the last 14 intakes)?'.
Statistical Analysis
Statistical analysis was conducted using sas 8.2 (SAS Institute Inc., Cary, NC, USA). Tests were two-sided and type I error was set at 0.05. Descriptive statistics were performed using all available data. Bivariate group comparisons were carried out using the Kruskal–Wallis, chi-square or Fisher's exact test as appropriate. The relationships between adherence and a set of potential explanatory variables were analysed by forward stepwise logistic regressions. These variables included those for which the groups differed significantly (
P<0.05) at baseline in bivariate analyses (100 vs. <100% adherence and/or ≥80 vs. <80% adherence), and variables expected to be related to adherence. The quality of life data were analysed as recommended in the SF-36 manual and interpretation guide.
[23]
Results
PatientsA total of 184 investigators enrolled and followed up 2001 HCV patients in the survey between November 2002 and January 2005. Of these patients, 141 were excluded from analysis because they did not receive combination therapy with peginterferon α-2b and ribavirin combination therapy (n=37), duration of combination therapy was not available (n=79) or virological data were not available (n=25). Compared with the analysed population, excluded patients were more often genotype 1 carriers (66 vs. 55%, P=0.013) and less often genotype 3 carriers (185 vs. 25%, P=0.072). They had more severe liver disease (Metavir A2–A3: 66 vs. 54%, P=0.022; F3–F4: 47 vs. 34%, P=0.007), and had received more often a previous HCV treatment (43 vs. 28%, P<0.001). All other baseline data were not significantly different.
The analysed population included 1860 patients. Their baseline characteristics are provided in
Table 1. Most of them (72%) were naive for hepatitis C therapy and 36% had genotype 2 or 3 HCV infection. Liver disease was assessed by biopsy in the majority of patients (
n=1606, 86%) and estimated using noninvasive markers in the remaining patients. Two-thirds of patients had significant fibrosis (F2, F3, F4: 65%), including 15% of cirrhotics.
Comorbidities were frequent. The cohort included 22% of psychiatric patients, 44% of drug users and 3% of HIV-co-infected patients. Current psychiatric disorders were diagnosed by a psychiatrist in 62% of the cases and included mainly depressive (
n=203, 11%), anxiety (
n=129, 7%), psychotic (
n=19, 1%) and bipolar (
n=8, 0.4%) disorders. Other comorbid chronic diseases were hypertension (
n=176, 36%), diabetes mellitus (
n=107, 22%) and asthma (
n=42, 9%). Health-related quality of life was altered, each SF-36 subscore being moderately lower than that in the general population in France.
[23]
Antiviral TreatmentThe mean dose regimen initially prescribed was 1.37 μg/kg/week (median 1.5, n=1837) for peginterferon α-2b and 922 mg/day (median 1000, n=1820) for ribavirin. A total of 1089 (58.5%) patients received various forms of therapeutic education at the discretion of the physician during the first 3 months of treatment. The average dose of both drugs progressively decreased over time, reaching 0.89 μg/kg/week (median 0.8, n=1729) for peginterferon and 595 mg/day (median 545, n=1728) for ribavirin at month 12. According to the physician, 915/1860 (49%) patients did not complete therapy as intended initially. The main reason for not doing so was virological criteria (n=302, 16%), lost to follow-up (n=244, 13%), safety (n=182, 10%), patient's request (n=78, 4%), another reason (n=51) or not specified (n=58). However, as calculated subsequently, only 563 (30%) patients had 'insufficient' treatment duration, i.e. <80% of the recommended duration (24 weeks with genotype 2 or 3, 48 weeks with the other genotypes). Adverse events were reported in 1598 (86%) patients overall.
Adherence to Treatment
Patient-reported adherence to combination therapy was 100% for the full initially intended treatment duration (overall adherence) in 580/1510 (38%) patients; it was ≥80% in 747 (50%) patients. The proportion of patients with good adherence at a given time point was stable over time, at 53–58% for 100% adherence and 64–66% for ≥80% adherence (Fig. 1).
Proportion of patients with 100% and ≥80% adherence to combination therapy (patient report).
Patient- and physician-reported adherences to treatment were discordant. As shown quarterly in Figure 2, the proportion of patients reporting a 100% adherence to peginterferon at a given time point was 76–79%, whereas physicians believed it was above 97%. Moreover, the proportion of patients reporting a 100% adherence to ribavirin was 62–66%, whereas physicians believed it was above 90%.
Figure 2. Proportion of patients with 100% adherence to (A) peginterferon α-2b and (B) ribavirin over time, as reported by the patient and as perceived by the physician. |
Adherence to ribavirin was always worse than adherence to peginterferon (Fig. 2) and adherence to both products (Fig. 1) was always worse than adherence to ribavirin. Therefore, most patients with imperfect adherence to one product were different from those with imperfect adherence to the other product.
Bivariate comparison of patients who took 100% of both products for the full initially intended treatment duration with those who did not is provided in
Table 2. They differed significantly for the following items. 100% adherent patients had longer transport time to the physician office (
P=0.001). They were more often genotype 2 or 3 HCV carriers (
P=0.009) and naive for antiviral HCV therapy (
P=0.037). In addition, they had less frequent diabetes (
P=0.026), they consumed lower alcohol amounts at baseline (
P=0.023) and they were less often regular drug users (
P=0.017), but were more often HIV co-infected (
P=0.003). They also used illicit drugs less frequently during follow-up (
P=0.007) and, although not significantly, drunk less frequently >20g/day alcohol (
P=0.053). They did not differ markedly for the HCV treatment actually received and occurrence of adverse events.
The potential explanatory variables proposed to the multivariate model included genotype, remoteness of the centre (transport time), Metavir activity and fibrosis scores, sex, age, BMI, previous HCV treatment, serum HCV RNA, HIV co-infection, psychiatric disorders (ever and at baseline), diabetes, alcohol consumption >20g/day (at baseline and during follow-up), drug use (at baseline and during follow-up) and therapeutic education. As a result (
Table 3), the factors significantly associated with 100% adherence to combination therapy were HIV co-infection, no illicit drug use during follow-up, HCV genotype 3, HCV treatment-naive and, to a lower extent (odds ratio close to 1), remoteness of the centre.
Quality of LifeThe change from baseline of SF-36 subscores is displayed in Figure 3. During the treatment period (up to month 12), all physical and mental domains progressively worsened, in particular those reflecting problems at work or in daily activities that result from the physical (role physical) and mental (role emotional) status. The mean change from baseline of the mental and physical composite scores, which integrate all SF-36 domains except health transition, was, respectively, −6.4 and −6.5 points at month 6, and −5.7 and −5.3 points at month 12. The health transition score, which estimates the change of perceived health condition compared with 1 year before, also worsened during treatment as shown by a mean change from baseline at +6.6 points at month 6 and +5.5 points at month 12. After the end of treatment, quality of life was returned above the pretreatment level; the mean change from baseline was +2.6, +2.4 and −6.2 points for the mental composite score, physical composite scores and health transition score respectively.
Figure 3.
Mean change from baseline of SF-36 subscores (point). MCS, mental composite score; PCS, physical composite score. Physical domains: BP, bodily pain; GH, general health; PF, physical functioning; RP, role physical. Mental domains: MH, mental health; RE, role emotional; SF, social functioning; VT, vitality. HT, health transition
Almost all SF-36 domains were significantly worse at baseline in patients who reported an adherence <100% compared with patients with 100% adherence (P<0.05 for PF, BP, GH, VT, SF, MH, PCS and HT; not significant for RP, RE and MCS). During and after treatment, changes of quality of life were parallel in both groups; there were no significant differences between groups for the change from baseline of each SF-36 subscore, whatever the time point.
Discussion
To our knowledge, this is the largest study assessing adherence to HCV therapy using data on dose taking. Imperfect adherence was common. Overall, only 38% of our routine patients reported strict adherence to peginterferon α-2b and ribavirin, i.e. full-dose, persistent therapy as initially intended by the physician. In addition, 76–79% of patients on treatment reported having taken all peginterferon doses in the last 4 weeks; 62–66% reported having taken all ribavirin doses in the last 7 days; and 53–58% reported having taken all doses of both drugs at months 3, 6, 9 and 12. We also provide information on adherence in terms of early treatment discontinuation (30%) and dose decreases, as usually referred to by clinicians in HCV infection.
Our results add to the literature as among the few studies that assessed adherence in terms of dose taking, none did so in the whole population of routine HCV patients and over the whole treatment period. In a clinical trial including 401 mono-infected HCV patients,
[24] at least 95% of patients reported having taken all peginterferon doses in the past 4 weeks at months 1, 3, 6, 9 and 12; the rate of patients who reported having taken all ribavirin doses in the past 4 days decreased from 91% at month 1 to 43% at month 12. In a cross-sectional study involving 180 routine HCV patients,
[25] 7% of patients reported having missed at least one peginterferon dose in the last 4 weeks and 21% having missed at least one ribavirin dose in the last 7 days. Patients were under treatment since 19.3 ± 13.4 weeks in average. In a retrospective study where adherence was estimated using pharmacy refill data in 188 HCV US Veterans,
[26] 73% of patients were found with at least 100% adherence to peginterferon and 68% with at least 100% adherence to ribavirin during the initial 3 months of treatment. Lastly, in a cohort of 63 HCV/HIV-co-infected patients,
[27] 23% of patients discontinued treatment early and 98% of those on treatment reported having taken all peginterferon and ribavirin doses in the past 2 weeks at months 3, 6 and 12. Such a high adherence is not surprising as dose-taking adherence is nowadays routinely stressed in HIV patient care.
However, the ability of physicians to recognise nonadherence was poor. Our study physicians markedly overestimated adherence to combination therapy, by 20–30% compared with patient self-report. This phenomenon has already been shown in other chronic diseases such as HIV infection or diabetes but not yet in chronic HCV infection.
[28,29] Several reasons exist including for example poor or judgemental provider–patient communication.
[30,31] Moreover, in this indication, electronic monitoring has provided much lower rates of adherence than self-reported adherence,
[24] suggesting that patients also overestimate adherence to combination therapy. This medication-taking behaviour is also well known in other chronic disorders.
[30] Therefore, healthcare providers should be more vigilant about adherence to HCV combination therapy in their daily management, especially in terms of missed ribavirin doses. Indeed, standard ribavirin dosing is complex and ribavirin dose reductions, at least in the first weeks of treatment, may alter virological outcome.
[26, 32–34] Patient-related reasons for nonadherence may include forgetfulness, the decision to omit doses, lack of information and emotional factors.
[30] Clinician-related reasons may include, in addition to poor communication with the patient, failing to explain the treatment benefits and side effects and not giving consideration to a patient's lifestyle. More flexibility in indication for treatment could have a positive impact on the individual prognosis of patients and the overall control of the disease burden. Although there is no perfect method to assess adherence to medication, patient self-report is probably the simplest and most effective method of measurement.
This study provides other clinically relevant information. Patients who did not present with the following baseline characteristics of HCV infection, HIV co-infection, HCV genotype 3 and HCV treatment-naive, and patients who used illicit drugs during HCV treatment were at higher risk of imperfect adherence. Interestingly, adherence was not associated with history of addiction or psychiatric disorders, suggesting that these conditions should no longer limit access to HCV antiviral therapy. In addition, treating intravenous drug users should have a positive impact on prevention of transmission, with a chance of reducing the incidence of new cases. Knowledge of factors predictive of poor adherence is a useful resource for physicians to help identify patients who are most in need of intervention and plan more frequent and accurate follow-up.
Careful assessment of health-related quality of life was regularly performed in our patients using a validated questionnaire. As expected, quality of life was impaired before treatment initiation compared with the general population
[23,35] and both the mental and physical domains worsened during treatment. However, 6 months after the end of treatment, it was returned to baseline, even slightly better than before treatment. Quality of life changes were parallel in patients with perfect and imperfect adherence. These results confirm previous reports of temporary worsening.
[36,37] Therefore, physicians should reassure patients and encourage them to persist with therapy despite frequent side effects and worsened quality of life.
In conclusion, this cohort study brings potential clinically relevant information by emphasising the following points. Imperfect adherence to combination therapy with peginterferon plus ribavirin is common in routine chronic hepatitis C patients. Adherence is markedly overestimated by physicians and is associated with some patient's baseline characteristics. These findings suggest that assessment of adherence to HCV combination therapy by physicians should be improved. This could be easily carried out by the wide use of standardised adherence measurement tools such as a self-questionnaire, keeping in mind that patients may overestimate the true figures. The need to enhance communication would be triggered by discordance between the physician and patient assessments. Knowledge of baseline characteristics associated with adherence might help adjust the monitoring in a subset of patients at higher risk of nonadherence.
Hepatitis C virus infected females have a higher risk of advanced fibrosis and graft loss after liver transplantation than males
