We have demonstrated previously that only a small number of veterans with a diagnosis of HCV were prescribed treatment for HCV (6
). In this study, we have shown that only a small proportion of HCV-infected veterans who initiate treatment complete a 48-week course. Determination of treatment completion rate is important independent of the rates of treatment initiation or eligibility, because it may help us identify modifiable factors that could improve treatment completion rates, and clinical outcomes.
Treatment completion rates in clinical trials for HCV range between 86 and 88% (1
). However, clinical trials recruit participants who have been carefully screened to exclude significant comorbidity and who are generally well motivated to initiate and complete treatment. Treatment completion rates in clinical practice settings may be quite different. Our study demonstrates that less than one-quarter of the persons initiated on treatment for HCV in a large national cohort of veterans completed a 48-week course of treatment. Those who completed treatment were less likely to have pre-treatment anaemia and depression. Presence of major medical, psychiatric or substance abuse comorbidities are common reasons for not initiating treatment for HCV in both HCV monoinfected as well as HCV–HIV-coinfected persons (6
), but to our knowledge, they have not been studied as factors influencing treatment completion in a national sample of HCV-infected persons.
In our study, pre-treatment anaemia emerged as an important independent predictor of completion, with the likelihood of completion decreasing with the severity of anaemia. Subjects with most severe anaemia (haemoglobin < 10 g/dl) were only 0.22 times as likely to complete a course of treatment compared with persons with no anaemia (haemoglobin > 14 g/dl). The standard of care for HCV therapy includes ribavirin, which causes a dose-dependent haemolytic anaemia in a significant number of treated subjects. In one study, almost 50% of the subjects treated with 1000–1200 mg/day of ribavirin experienced a haemoglobin drop of > 3.0 g/L (34
). Presumably, those with baseline anaemia had a worsening of anaemia on therapy, necessitating treatment interruption or withdrawal. This underscores the need for aggressive evaluation and management of anaemia in patients with HCV when treatment is being contemplated.
The best management strategy for pre-treatment anaemia remains to be established. Recently, erythropoietic growth factors have been shown to maintain ribavirin dosing, improve haemoglobin levels and improve health-related quality of life in HCV-infected patients (35
). Compared with ribavirin dose reduction to manage anaemia, the cost of epoetin- α per additional quality-adjusted life year was US$60 600 for genotype 1 infected persons and US$64 311 for genotype 2 or 3 infected persons (37
). However, these potential benefits need to be carefully reassessed in view of recent studies demonstrating no survival benefit, and possibly an increased mortality risk with correction of anaemia with erythropoietin in certain populations (38
Severe psychiatric illness is associated with a lower rate of treatment prescription for HCV (6
). Our current study demonstrates that pre-existing depression is also significantly associated with failure to complete treatment. This finding underscores the need to evaluate and treat HCV-infected patients for depression before treatment initiation. More studies are warranted to determine whether aggressive treatment of depression would improve treatment completion rates in HCV-infected persons.
Substance abuse and dependence have traditionally been considered relative contraindications to treatment for HCV. In the most recent VA guidelines, evaluation of patients with ongoing drug and alcohol use is encouraged, and antiviral therapy recommended for patients enrolled in methadone maintenance programmes. For patients with ongoing alcohol abuse, counseling is recommended while offering them therapy (31
). We did not find a statistically significant independent association between alcohol or drug abuse and dependence and treatment completion for HCV, providing further evidence that these patients be evaluated for treatment while appropriately addressing the substance abuse issues.
We found that individuals treated with pegylated interferon were substantially more likely to complete a 48-week course of treatment compared with those on standard interferon. This is likely because of the lesser frequency of administration (once weekly for pegylated interferon vs. thrice weekly for standard interferon). We did not assess whether a difference in the rate of adverse events could account for this difference, but previous clinical trials have not shown major differences in the rates of systemic or laboratory adverse events that would necessitate discontinuation of therapy (1
). Because a significant number of interferon-related adverse events are injection related, reduction in injection frequency would be expected to be associated with lesser adverse events, and consequently improved treatment completion rates.
We found a strong and independent association between number of subjects initiated on treatment at each site and rate of treatment completion for HCV. At sites where 200 or more subjects were initiated on treatment during the study duration, rate of treatment completion was 87% higher than among sites with < 200 subjects initiated on treatment. There was no significant association between number of HCV-infected subjects diagnosed at each site and treatment completion, suggesting that experience in treating HCV-infected persons at local level is an important determinant of treatment completion. We hypothesize that providers at high-volume treatment sites may have devised management strategies that ensure better treatment completion rates. Whether there are better support structures at these sites (e.g. dedicated pharmacists, social workers, alcohol and drug abuse counselors) is not known.
Of interest is our finding that HIV coinfection does not appear to be associated with a lower rate of completion of treatment for HCV at 48 weeks. Several studies have demonstrated that HCV–HIV-coinfected subjects have higher prevalence of medical, psychiatric and substance abuse comorbidities, are less likely to be initiated on treatment for HCV, and have poorer response to therapy (9
). Our results suggest that, among those selected for treatment, treatment completion is not substantially different by HIV status.
There are many strengths to our study. We used a national sample of HCV-infected persons in a setting where financial incentives or disincentives are minimal. The VA is the largest integrated healthcare system in the USA, and because of the computerized integration of records, and national coverage of eligible veterans, this system is able to track, follow and treat patients even when they move from one geographical area to another. Fully computerized and comprehensive pharmacy records allow investigators and administrators to track all the prescriptions written, quantity prescribed and the duration of prescriptions. We performed several sensitivity analyses to ensure the stability of our findings across a number of possible assumptions. The general associations and the magnitude were similar in all analyses.
Despite the strengths of our analyses, some limitations of large database analyses, and the VA data need to be understood. Some veterans may have received additional treatment outside the VA, though this number is expected to be minimal because of costs. In addition, some veterans may have received treatment in clinical trials settings. However, the VA generally requires drugs to be prescribed through local pharmacies and that is likely to be captured in the PBM database. We defined a full course of treatment as 48 weeks of treatment. However, patients infected with HCV genotype 2 or 3 require only a 24-week course of treatment according to the current guidelines. As pointed out earlier, 79–87% of HCV-infected veterans are infected with HCV genotype 1, which requires a 48-week course of therapy. Because we did not have complete genotype information, this may have led to a misclassification in subjects with genotype 2/3 infection. We did not assess the adherence to therapy in our study. The comorbidity diagnoses (except anaemia) were based on ICD-9 codes, and while our group and others have validated them and found to be quite accurate, we did not perform an independent chart review validation. We studied treatment completion between 1998 and 2003, and it may be argued that treatment for HCV was less well accepted in that time frame. However, since we are studying treatment completion rates among those who already initiated treatment, we believe that our study is still valid. The management of treatment adverse events and provider comfort may have increased with experience, and we did not assess these factors. Another limitation of analysing such large database is finding statistical significance when no true clinical relevance exists.
It has been argued that the veterans in care are a non-representative sample for the US population in general. This is true when comparing all veterans in care with the US population since veterans are predominantly male, and older. However, with the exception of women, the HCV epidemic among veterans in care is largely representative of the national epidemic. According the estimates from the Centers for Disease Control and Prevention, most HCV-infected persons in the US are between 30 and 49 years old, are more likely to be black and have a higher rate of drug use. Further, the prevalence of HCV is about twice as high in men as women. Thus, the HCV epidemic in the veterans is not much different from the general population of HCV-infected men (43
In conclusion, less than one-quarter of HCV-infected veterans who are initiated on treatment complete a 48-week course of treatment. Non-completion is associated with pre-treatment anaemia, depression and a smaller number of treated patients at the site. Strategies to address these comorbidities should be instituted before universal advocacy of HCV treatment for every infected person, if therapeutic success at the population level is to be achieved. Effect of consolidating services to higher volume centers experienced in such treatment, or enhanced training of the healthcare providers at lower volume center need further study.