Efficacy of various vaccines against pneumonic pasteurellosis in cattle: A meta-analysis

The objective of this study was to compare the protective efficacy induced by various experimental and commercial vaccines used in the past 12 years (1982-1994) against bovine pneumonic pasteurellosis. All studies using an experimental-challenge model to evaluate protective efficacy induced by vaccine preparations were identified using a combination of VETCD, CAB, and MEDLINE searches. A total of 72 trials were identified in 27 studies. Age groups of animals used, sample sizes used, vaccine types, route of vaccination, dose of vaccine, adjuvant types, booster times, time of challenge since first vaccination, challenge methods, challenge doses, challenge route, observation times postchallenge, and raw mean lung-lesion scores (defined as percent lung involved) were extracted from each publication and used in the analysis. Vaccine types were grouped into 11 categories as controls, live, potassium thiocyanate (KSCN) extract, phosphate buffered saline extract (PBSE), 30 kDa recombinant protein, bacterin, streptomycin dependent live (Sm-D live), sodium salicylate extract (SSE), sarkosyl outer membrane protein extract (sarkosyl OMP), or culture supernatant (CS) categories based on the subunit content in each preparation. In a second analysis, the chemically altered live vaccine was grouped together with live Sm-D group as modified live vaccines. The least-squares means of the postchallenge lung-lesion scores of each vaccine type were then compared after controlling for all the above-mentioned variables in an analysis of variance using Fischer's protected least significant difference. The statistical analysis was done using the general linear models procedure of Statistical Analysis Systems Institute Inc.. Protective indices for each vaccine type were calculated as a function of independent probabilities of protection among vaccinates and controls. The protective indices of each vaccine type were then compared in a similar model. Cattle vaccinated with live vaccines had the smallest lung-lesion scores (9 +/- 8) which significantly differed from that of controls (44 +/- 8), bacterins (23 +/- 6), and live Sm-D (38 +/- 12) or modified live (31 +/- 10). KSCN extract (9 +/- 10), bacterin, and CS (20 +/- 9) also had significantly lower (P < 0.05) lung-lesion scores compared to controls. The CS and KSCN vaccines showed a significantly (P < 0.05) better protective index (PI) (0.70 and 0.68 respectively) than bacterins (0.50) or bacterin + CS (0.44) vaccines. Saline extract and live vaccines had better (P < 0.05) PI (0.60 and 0.70 respectively) compared to bacterins.