+49 30 838 62618
Veterinarians provide a wide range of services and play an important role in protecting animal welfare and maintaining animal and public health. Publications are important to implement knowledge into the practice despite the difficulties of a continuous increase of literature and variations in quality. Thus, the systematic evaluation of the quality of veterinary literature may help to imply best available care for patients of the highest available evidence. The overall objective of this study was to evaluate the quality of actual veterinary literature and the manageability of its assessing. First an overview of evidence-based veterinary medicine, its strengths and weaknesses regarding the establishment in a workaday life was shown. In the first study the quality of a randomly selected sample of published literature on reproduction of cattle, horses and dogs was evaluated and compared. A literature search was conducted in the databases Medline and Veterinary Science. Approximately five times more articles on clinical bovine reproduction (n = 25 910) were found compared to canine (n = 5 015) and equine (n = 5 090) reproduction. For the evaluation of the literature a checklist was used. A subset of 600 articles published between 1999 and 2008 was randomly selected. After applying exclusion criteria a total of 268 trials (86 for cattle, 99 for horses and 83 for dogs) was evaluated and used for further analysis. Data of this study demonstrated for the field of canine and equine reproduction fewer clinical trials with a control group compared to bovine reproduction (cattle 66%, horses 41% and dogs 41%). For all three species investigated, few publications were identified (4%) with the highest level of evidence, i.e. controlled, randomized and blinded trials or meta-analyses. In cattle 33% of the publications were graded adequately to draw sound conclusions. Only 7% and 11% were graded adequately in dogs and horses, respectively. Altogether, the results of this study demonstrate deficits and differences of published literature on reproduction in cattle, dogs and horses. Hence, improvement of the quality of well-designed, conducted and reported clinical trails is required. The objective of the second study was to determine the inter-observer agreement utilizing an existing checklist for the evaluation of scientific literature in the field of animal reproduction. Three publications on bovine reproduction (one case report, one randomized controlled study without blinding and a blinded, randomized controlled study) were selected. Fourteen international recognized scientists in the field of animal reproduction were utilized. Each reviewer was provided with the three articles, three checklists, and supplementary explanations. Altogether, 13 of the respondents filled out over 90% of the answers of the three evaluation forms. The repeatability for the respondents using a Fleiss’ kappa was 0.35. By combining ‘strongly agree’ and ‘agree’ as well as ‘strongly disagree’ and ‘disagree’ the kappa value increased to 0.49. There was a very strong agreement among the respondents concerning the classification of the level of evidence (98% of all evaluation forms). Only in one case a randomized, controlled trial was considered as a controlled trial. Evaluation of information provided regarding housing (35% identically answers) and preconditions or pretreatments (42%) of the animals varied widely in all publications. These data illustrate that by combination of the extreme positions the kappa value raised from a fair to a moderate agreement. Even if the repeatability of classification was moderate, repeatability of important review categories was high. Our data provide evidence that such a checklist does provide a reasonable and practical tool to assess the quality of publications. Overall, the two studies demonstrated the manageability of using a checklist to assess the quality of publications. Additionally, publications in the field of animal reproduction vary widely in their quality and the deficits are species specific. The lack of high quality trials hampers the implementation of high evidence knowledge. Therefore, the veterinarian should always try to assess the quality of information before implementing results into practice.