Oxman 2005

Methods	Randomised, placebo‐controlled, double‐blind study at 22 sites in the US Time of follow‐up: at least 7 years of surveillance for HZ
Participants	N = 38,546 participants 60 years of age or older, with history of varicella or had resided in the continental United States for at least 30 years Median age in both groups was 69 years ˜59% male 95.4% white race
Interventions	1. Zoster vaccine (frozen) (18,700 to 60,000 plaque‐forming units per dose (pfu/dose) and more than 90% of vaccinated participants received 32,300 pfu or less) SC; N = 19,270 2. Placebo SC; N = 19,276
Outcomes	Confirmed cases of HZ, cases of HZ within 30 days of vaccination, confirmed HZ cases and all adverse events occurring within 42 days after vaccination and during the whole study Participants with follow‐up, participants with 1 or more AEs (systemic or injection site), participants with serious AEs, vaccine‐related AEs (systemic or injection site), death, varicella‐like rash at injection site and not at injection site, herpes zoster‐like rash, rash unrelated to HZ, participants hospitalised, hospitalisation related to HZ
Purpose of the Study	"To determine whether vaccination with a live attenuated varicella‐zoster virus vaccine would decrease the incidence, severity, or both of HZ and postherpetic neuralgia in adults 60 years of age or older"
Funding sources	"Supported by the Cooperative Studies Program, Department of Veterans Affairs, Office of Research and Development; by a grant from Merck (to the Cooperative Studies Program); and by a grant from the James R. and Jesse V. Scott Fund for Shingles Research (to Dr. Oxman). The vaccine and placebo used for the study were supplied by Merck; famciclovir was supplied by SmithKline Beecham and Novartis Pharmaceuticals"
Notes	"Zoster vaccine and placebo were lyophilised, held frozen at ‐15°C until reconstituted with sterile water, and administered within 30 minutes" 132 participants withdrew from the study and 113 were lost to follow‐up 1588 participants died during the study, but it was not described whether these were related to the protocol or not Only a subgroup of patients had a safety assessment (zoster vaccine N = 3345; placebo N = 3271), being the adverse event sub‐study This study performed 2 ITT analyses, with all individuals developing HZ and only with those who developed after 30 days from the vaccine injection (modified ITT). For the meta‐analysis we considered the modified ITT There was a break in surveillance for cases of HZ of approximately 15 months between the completion of the Shingles Prevention Study surveillance in September 2003 and resumption of follow‐up in the Short‐Term Persistence Substudy in December 2004. Beginning in October 2005, open‐label zoster vaccine was offered without charge to Shingles Prevention Study placebo recipients. Placebo recipients enrolled in the Short‐Term Persistence Substudy completed the study upon receiving zoster vaccine, since they could then no longer serve as unvaccinated controls. The Short‐Term Persistence Substudy participants who were zoster vaccine recipients in the Shingles Prevention Study continued to be followed until the initiation of the Long‐Term Persistence Substudy in March 2006 The 2012 publication evaluated the effectiveness of the vaccine for up to 7 years after the participants had been vaccinated. However, the data available in this publication report different dates for the collection of outcomes in the intervention and in the placebo groups. The data from the zoster vaccine group are from December 2004 to March 2006 (16 months). In the placebo group, data are reported only from December 2004 to September 2005 (10 months), because in October 2005 the zoster vaccine was also offered to participants in the placebo group, as stated by the authors reported above We contacted the authors of this study asking for the data corresponding to the period from December 2004 to September 2005 (10 months) for both groups (vaccine and placebo). They replied to our email but did not provide this information and suggested instead that we should "assume a uniform rate of events and calculate the estimated number of cases from that"
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not described
Allocation concealment (selection bias)	Low risk	"Each study site received randomly ordered vials of zoster vaccine and placebo in separate boxes for each age stratum"
Blinding (performance bias and detection bias) All outcomes	Low risk	"All other study personnel were blinded to study treatment assignments"
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Since the reconstituted zoster vaccine had a different appearance from the placebo, reconstitution and administration were performed by technicians who did not otherwise interact with participants, evaluate outcomes or adverse events, answer the telephone or enter study data."
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not described
Incomplete outcome data (attrition bias) All outcomes	Low risk	Clear patient flow
Selective reporting (reporting bias)	Low risk	All data on effectiveness and adverse events that the authors proposed in their methodology were described in the results for both groups
Other bias	Unclear risk	Not described