This supplement article by Oxman and Levin describes VA Cooperative Study #403, a randomized, double-blind, placebo-controlled trial at 22 US sites enrolling 38,546 immunocompetent adults ≥60 years (median 69; 41% female) between November 1998 and September 2001, with mean surveillance of 3.13 years and >95% retention via an automated telephone follow-up system. Diagnosis of suspected zoster cases was adjudicated by a blinded clinical evaluation committee and validated by real-time PCR distinguishing wild-type from Oka vaccine strain; 88.4% of 1,308 suspected cases were PCR-diagnosed, and vaccine-strain virus was never detected in confirmed cases .
Reported efficacy: HZ burden of illness reduced 61.1% (95% CI 51.1–69.1); PHN incidence reduced 66.5% (95% CI 47.5–79.2; 27 vs 80 cases; 0.46 vs 1.38/1000 person-years); HZ incidence reduced 51.3% (95% CI 44.2–57.6; 5.4 vs 11.1/1000 p-y). Efficacy against HZ incidence was lower in subjects ≥70, while PHN efficacy was preserved across age strata. Safety signals were limited to injection-site reactions (erythema 36%, pain/tenderness 35%); post-vaccination SAE rates in the adverse-event substudy were 1.9% (vaccine) vs 1.3% (placebo, P=.034) with no clinically meaningful differences on subject-by-subject review .
Strengths are genuine: massive randomized sample, active monthly surveillance, laboratory confirmation distinguishing wild-type from vaccine strain, blinded adjudication, and prespecified success criteria. The Zoster Brief Pain Inventory–derived severity-by-duration endpoint (HZ BOI) is innovative but composite—it pools incidence and severity, so component effects (51.3% incidence reduction plus modest severity reduction, 141.2 vs 180.5 mean score) must be disaggregated, as the authors themselves do .
Conflicts of interest are substantial and disclosed: the study chairman's trial was partly funded by Merck, coauthor Levin held a partial patent interest in the vaccine and sat on Merck's speakers' bureau, and the supplement itself was sponsored by Osaka University Biken, GlaxoSmithKline, the Sabin Vaccine Institute, CDC, and the March of Dimes . Only 0.6% of subjects were lost to follow-up, but immunocompromised persons—a group at highest HZ risk—were excluded, limiting generalizability, and the ~3.1-year window leaves durability unknown. Later observational work showed live-vaccine effectiveness waning from 49.1% in year 1 to 31.8% by year 8, and one review reports 4.2% effectiveness by year 8 for Zostavax, while the recombinant adjuvanted vaccine (ZOE-50/70) achieved 97.2% (age ≥50) and 89.8% (age ≥70) with ~89% long-term efficacy through ~9.6 years .
An independent 2005 Antelope Valley analysis (lower-quality journal, methodology caveats) also raised the hypothesis that universal childhood varicella vaccination reduces exogenous immune boosting and could elevate adult HZ incidence—a confounder the SPS did not measure, and which the paper acknowledges as a hypothesis from other investigators .
Bottom line: For its population (immunocompetent US adults ≥60) and its ~3-year window, the SPS provides strong, internally consistent randomized evidence that the live attenuated vaccine reduced HZ-related pain morbidity and PHN. Confidence is high for the reported period; it is explicitly not evidence for long-term durability, immunocompromised populations, or superiority over the later recombinant vaccine. What would change the conclusion: equivalent or higher morbidity in vaccine recipients in independent replication, loss of durability, or safety signals outweighing benefits—the waning-effectiveness data partially confirm the durability caveat.
Know what changed, what holds up, and what remains uncertain. Every Friday. No ads.