Vaccine 21 (2003) 2003–2014 Systematic review of the effects of pertussis vaccines in children Tom Jefferson a,∗ , Melanie Rudin a , Carlo DiPietrantonj b a Health Reviews Ltd., Via Adige 28a, 00061 Anguillara Sabazia, Rome, Italy b Cochrane Vaccine Field, Alessandria, Italy Received 19 November 2002; accepted 27 November 2002 Abstract Objective: To assess the efficacy and safety of whole-cell and acellular pertussis vaccines administered to children singly or within diphtheria, tetanus and pertussis (DTP) vaccines. Data sources: We searched the Cochrane Library, MEDLINE, EMBASE, Biological Abstracts and Science Citation Index to December 2001. Specialised websites and bibliographies of retrieved articles and reviews were assessed. Vaccine manufacturers and investigators were contacted for additional data. Review methods: We included randomised and cohort studies comparing efficacy and/or safety of pertussis vaccines with placebo, DT, no intervention or each other. Results: We included 52 studies (49 randomised controlled trials (RCTs), 3 cohort studies). All tested whole-cell and acellular vaccines were significantly more effective than placebo against pertussis. Absolute efficacy of whole-cell DTP varied from 37 to 92%. One- and two-component acellular vaccines had lower absolute efficacy (67–70%), than vaccines with ≥3 components (80–84%). Whole-cell vaccines were associated with significantly higher incidences of swelling and induration (odds ratio (OR) 11.67, 95% confidence interval (CI) 8.83–15.44), fever (OR for fever >39 ◦ C 3.36, 95% CI 2.06–5.49) and crying for >2h (OR 4.72, 95% CI 2.94–7.59) than placebo or DT. Differences in incidence of hypotonic hyporesponsive episodes (HHE) and convulsions were not statistically significant. Acellular pertussis vaccines did not cause a higher incidence of local signs, fever, convulsions, HHE or prolonged crying than placebo or DT. Conclusion: All tested pertussis vaccines were efficacious. Whole-cell vaccines show variable efficacy, making interpretation of direct comparisons unreliable. Acellular vaccines with ≥3 antigenic components showed higher efficacy than one- and two-component vaccines. The adverse event profile of acellular vaccines was similar to that of placebo and considerably better than that of whole-cell vaccines. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Pertussis vaccines; Systematic review; Meta-analysis; Effectiveness; Safety 1. Introduction Vaccination against pertussis (whooping cough) is widely practised and coverage with DTP (pertussis combined with diptheria and tetanus toxoids) is almost universal in Eu- rope [1]. Whole-cell vaccines against pertussis (wP) were developed in the 1940s but concerns about possible associ- ations with infantile spasms, convulsions and epilepsy led to the development of acellular vaccines (aP) in the 1970s [2,3]. These contain up to five of the Bordetella pertus- sis antigens: pertussis toxin, filamentous heamagglutinin, pertactin and three serotypes of fimbrial antigens or ag- glutinogens. Identification of the best pertussis vaccines is important to ensure effective prevention and to ensure effi- cient use of healthcare resources, since whole-cell vaccines are considerably cheaper than acellular vaccines. However, the array of pertussis vaccines within DTP, differences in ∗ Corresponding author. Tel.: +39-06-999-00-989. E-mail address: toj1@aol.com (T. Jefferson). case definitions, ascertainment procedures and vaccination schedules have made it difficult to review the evidence about their relative efficacy [4–11]. We report a systematic review of studies assessing the efficacy and safety of DTP vaccines. 2. Methods 2.1. Searching We searched the Cochrane Library, MEDLINE, EM- BASE, Biological Abstracts, Science Citation Index and OLDMEDLINE up to December 2001. We also searched the bibliographies of retrieved articles and reviews and the Vaccine Adverse Event Reporting System website [12]. Vaccine manufacturers and authors of relevant studies were contacted to identify further published or unpublished stud- ies and to answer queries about the conduct or outcome of studies. 0264-410X/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0264-410X(02)00770-3