This study was conducted to confirm the presence of bovine tuberculosis (BTB) and determine its prevalence, based on skin test reactivity, in cattle reared under extensive farming conditions in the Govuro district, Inhambane province, Mozambique. The study was comprised of a primary screening test using the single intradermal test (SIT) in randomly selected animals from Colonato and Sede dip tanks in Govuro. Positive reactors to the SIT were tested again with bovine and avian tuberculin using the single intradermal comparative test (SICTT) 7 weeks after the SIT. The sample size was calculated using Win Episcope 2.0 based on 95% confidence to detect a 2% expected prevalence using the SIT, with a 1% accepted error and accounting for a total population size of 7208. The calculated sample size was 682 animals. To compensate for the probability of 20% default in reading, the sample size was increased to 853. During the testing process (SIT), it was evident from the first 3 reading days that the apparent prevalence (61, 94%) was higher than expected (2%), hence we decided to stop when the total number of cattle was 530. During the testing process (SIT), it was evident from the first 3 reading days that the apparent prevalence (61.94%) was far higher than expected (2%), hence we decided to stop when the total number of cattle was 530. This was due to the fact that, at such a high prevalence, it would not be necessary to achieve as high a precision as 1% accepted error. A sample size of 530 would be sufficient to achieve a precision of 4% accepted error, which was regarded as more than adequate. The 530 cattle, 3 or more years of age, were selected using systematic random sampling from the two dip tanks (Colonato 371 and Sede 159 animals). All animals were identified by numbers painted, dorsally on the sacral region. Out of 530 tested cattle by SIT, 268 were read, and 166/268 (61.94% with 95% confidence interval [CI]: 55.8 – 67.8%) were found positive, with visible swallow at the injection site. Apparent prevalence (AP) was found to be 61.94% while the true prevalence (TP) was 75.92%. The predictive value of a positive result (PV+) was found to be 87.9%. No significant difference in apparent prevalence between the two areas was detected by Fisher’s exact test (P = 0.11). By SICTT, out of 28 animals positive reactors to SIT, 21 were possible to read, and 13/21 (61.9%; 95% CI: 55.1 – 89.3%) were found positive. A three year old bull, positive reactor to the SIT, was slaughtered, and a detailed post mortem was carried out and organs with visible lesions were collected for further laboratory testing (histopathology, culture and isolation of M. Bovis and PCR). Later on, 30 more positive reactors to the SIT test were slaughtered: 25/30 (83.3%) showed visible lesions compatible with BTB, and total condemnation of carcass was made in 3/25 (12%) due to generalized lesions. The high prevalence rate of skin test positive animals as well as gross lesions and histopathology were confirmed to be BTB by the isolation and identification of M. Bovis by culture and PCR. Our results suggest that bovine tuberculosis is highly prevalent in Govuro district and may thus represent a potential health problem of zoonotic tuberculosis in humans. Our results suggest that BTB has reached the plateau phase of endemicity in cattle in Govuro district. In this context, the positive predictive value of the SIT is very high and thus the use of the SICTT as a confirmatory test has a limited value and should not be advocated. Our results further indicate that no other prevalence study of BTB should be conducted in the next few years in Govuro district, unless comprehensive control measures are implemented. The focus of further studies should be on the isolation and the molecular characterization of M. Bovis from cattle and humans in order to assess transmission routes and the role played by BTB in human TB cases in Govuro district. Copyright
Dissertation (MSc (Veterinary Tropical Diseases))--University of Pretoria, 2010.