Reply: Essential need for research in hepatitis C

Dear Editor,

We have read the letter by J. S. McCombs1 and appreciate the comments. Our study compared during post therapy (on average 3–4 years) healthcare costs of patients who completed vs discontinued interferon therapy during the first 36–48 weeks of follow-up. To address his first comment regarding how savings attributed to completed therapy compared to costs, we have included a table with the overall healthcare costs during a revised observation period, which now includes the first 48 weeks post therapy initiation, stratified by hospitalizations, outpatient visits, ER visits and pharmacy costs (Table 1).

Table 1. Comparison of PPPY healthcare cost between beneficiaries in the complete vs discontinued HCV therapy cohorts including the first 48 weeks post therapy initiation.^a

PPPY healthcare costs (US $2012)	Complete therapy cohort	Discontinued therapy cohort	Complete vs discontinued therapy cohort
			Unadjusted cost difference	Adjusted cost difference^b [95% CI]	p Value
Number of patients, n	1066	1014
Observation period, days, mean (SD)	1665 (1021)	1514 (982)
All-cause healthcare costs, mean (SD)	$21,910 (28,734)	$22,187 (34,518)	−$277	$923 [−1730; 3517]	0.4945
Hospitalizations	$3262 (13,447)	$6237 (24,078)	−$2974	−$2639 [−4492; −1004]	<0.0001
ER visits	$230 (593)	$328 (1173)	−$98	−$89 [−159; −29]	0.0040
Outpatient visits	$5344 (15,815)	$6090 (10,498)	−$746	−$518 [−1528; 520]	0.3103
Pharmacy dispensing	$13,074 (11,535)	$9533 (11,805)	$3541	$4169 [3375; 4990]	<0.0001
CHC-related healthcare costs, mean (SD)	$11,128 (12,709)	$9215 (18,214)	$1913	$2426 [1030; 3703]	<0.0001
Hospitalizations	$1005 (8089)	$2556 (16,149)	−$1551	−$1504 [−2788; −325]	0.0100
ER visits	$3 (39)	$4 (34)	$0	$0 [−3; 3]	0.7888
Outpatient visits	$690 (1234)	$745 (1285)	−$55	−$28 [−145; 78]	0.6346
Pharmacy dispensing^c	$9430 (8849)	$5910 (7586)	$3520	$3958 [3484; 4431]	<0.0001
Interferon dispensing	$5841 (5348)	$3507 (3732)	$2334	$2598 [2307; 2894]	<0.0001

PPPY, per patient per year; SD, standard deviation; CHC, chronic hepatitis C; CI, confidence interval; ER, emergency room.

^aThe number of patients in the complete vs discontinued therapy cohorts reported are slightly higher than those reported in the manuscript since patients with 36–48 weeks of observation post therapy initiation, but no post therapy follow-up (due to end of insurance coverage or end of data availability [March 2012]) were also included, while in the manuscript only patients with at least 48 weeks post therapy initiation were included.

^bAdjusted cost differences were obtained using ordinary least-squares regressions, adjusting for baseline characteristics including age, gender, type of beneficiary, geographic region, payer type, year of index date, Quan-Charlson comorbidity index and healthcare costs. Confidence intervals (95% CI) were calculated using a non-parametric bootstrap with 999 replications.

^cCHC-related pharmacy dispensing costs were identified with NDC codes for any of these medicated treatments: peginterferon alfa-2a or alfa-2b, interferon alfa-2a or alfa-2b, or interferon alfacon-1, ribavirin, telaprevir and boceprevir.

In Table 1, we show that, after a mean follow-up of 1665 and 1514 days for patients who completed vs discontinued interferon therapy, respectively, the mean (SD) total healthcare costs per patient per year were $21,910 (28,734) vs $22,187 (34,518), with a non-significant adjusted cost difference of $923 in favor of incomplete therapy patients (p = 0.4945). These findings suggest that the higher costs of therapy for patients who completed vs those who discontinued interferon therapy during the first 48 weeks post therapy initiation were offset by their lower healthcare resource utilization after ∼3–4 years.

McCombs’ second comment is related to the high overall exposure to therapy of non-completers and sensitivities that could have been conducted on the adherence measure that was used to define patients who completed vs discontinued interferon therapy. In our study, interferon users were categorized according to their adherence to therapy, as defined by a persistence to therapy measure (i.e., no gap of 30 days or more between two refills) during the first 36 weeks of treatment. As mentioned by McCombs, the non-completers average was 391 days of exposure, suggesting that many non-completers likely started and stopped therapy multiple times. This is also what is suggested by the persistent therapy duration measure reported in Table 2 of our original manuscript, since non-completers had, on average, only 108 days of persistent therapy duration. We agree that the significant exposure to treatment by non-completers may have had an impact on their healthcare costs and likely reduced their costs and narrowed the difference in healthcare costs over the post-treatment period. The results presented in Table 1 of this letter provide some hindsight on the impact adherence to therapy has on healthcare costs, but sensitivities on adherence measures (e.g., using the proportion of days covered measure) to assess their impact was beyond the scope of the original study.

Regarding McCombs’ third comment that recent studies (including ours) did not study the predictors of treatment initiation or treatment response, we agree that it would be a very interesting topic that warrants future research, but it was again beyond the scope of the study on post therapy healthcare costs.

We are very grateful for the time McCombs took to review our manuscript and the opportunity to provide more information on our study findings.