Rosillon D, Willame C, Pladevall-Vila M, Zima J, van den Bosch JH, Bunge E, van Staa T, Boggon R, Baril L. Design and feasibility of a study using the Clinical Practice Research Datalink General Practice OnLine Database (CPRD GOLD) to assess the risk of new onset of auto-immune diseases (NOAD) following administration of the human papillomavirus (HPV)-16/18 AS. Poster presented at the 30th ICPE International Conference on Pharmacoepidemiology & Therapeutic Risk Management; October 2014. [abstract] Pharmacoepidemiol Drug Saf. 2014 Oct; 23(S1):277.

Background: The HPV-16/18 AS04-adjuvanted vaccine is indicated for protection against cervical cancer. When assessing the risk of rare events such as NOAD in vaccinated populations, large computerised clinical databases could be a valuable alternative to prospective field studies.

Objectives: To assess the feasibility of using CPRD GOLD to evaluate the risk of NOAD in adolescent and young adult women aged 9–25 years in the UK, after administration of the HPV-16/18 AS04-adjuvanted vaccine.

Methods: This observational study (NCT01953822) defines 4 cohorts (65,000 subjects each) in CPRD GOLD: 1 exposed female cohort (receiving =1 dose of HPV-16/18 vaccine Sep2008–Aug2010); and 3 unexposed cohorts: 1 concurrent male, 1 historical (Sep2005–Aug2007) male and 1 historical female. The male cohorts are internal controls for changes over time in reporting NOAD. Cases of pre-specified AD (using CPRD GOLD clinical terms) will be captured using predefined algorithms and confirmed by blinded independent experts by reviewing subject profiles, free text and hospital episode statistics available. Study feasibility was performed before study start.

Results: The feasibility assessment confirmed that CPRD GOLD contains a sufficient number of eligible subjects to define the 4 cohorts. Overall incidence rates of NOAD in the target population were close to estimates previously reported. Case ascertainment of a random sample of 55 potential cases of auto-immune disease (AD) classified 13 cases as non-confirmed AD (including 7 cases
of uveitis/iritis), 4 as non-AD, and 38 as confirmed AD. The positive predictive value of the algorithms was 69%.

Conclusions: CRPD GOLD is an adequate resource to assess the risk of rare events such as NOAD in a large vaccinated population. This feasibility assessment showed the need to have a robust case ascertainment method to increase clinical endpoint specificity. The positive output from this assessment resulted in study initiation (now ongoing).


Share on: