Observational studies were carried out in four industrial swine farms. The observations were performed on a 24 hours basis and mummified fetuses, stillborn and born alive piglets were recorded by observers (OB) and by employees (EM). In all farms, it was also obtained a retrospective data previous to the observational study period, comprising the same number of farrowings. Differences among mummified fetuses and total born piglets recorded by OB and EM in all farms (P<0.05) were observed. The recorded number of stillborn piglets differed between OB and EM in three farms (P<0.05) and the number of born alive piglets in one unit (P<0.05). In all farms during the analysis period, the maximal percentage of recorded mummified fetuses, stillborn, born alive and total born piglets, were 67.8%, 34.5%, 1.8% and 5.1%, respectively. The comparison of retrospective and EM data showed a maximal percentage of not recorded mummified fetuses, stillborn, born alive and total born piglets by EM of 70.4%, 82.1%, 10.5% e 16.5%, respectively. These errors can induce to a wrong interpretation of the farm data and the efforts to maximize the productivity could be directed to other sectors than the farrowing house.
record collection errors; mummified; fetuses; stillbirth; litter
