A total of 61% (n = 868) of all human diseases and 75% of emerging human pathogens are zoonotic 1 . Besides the fact that many emerging human diseases are zoonotic 2 3 4 5 , its only now that they have been demonstrated by quantitative analysis as risk factors for disease emergence 1 . Both domestic and wild animals have been shown to be important reservoirs of zoonoses 6 7 .

In Africa, bovine tuberculosis, brucellosis, anthrax, sleeping sickness, and rabies are still widespread 8 9 and in Tanzania, African trypanosomiasis, plague, rabies, brucellosis, anthrax and echinococcosis have been documented as being among the most common zoonoses 6 . In a study conducted in northern Tanzania in 2002, nineteen diseases were recorded as zoonoses by household members with rabies, tuberculosis, anthrax and brucellosis the top four zoonoses in pastoral, agro-pastoral and smallholder dairy farming systems 10 . Although the majority of households practiced at least one risk activity for transmission of zoonoses there was general lack of knowledge about the diseases 10 11 .

Although human brucellosis is a notifiable disease in many countries, official figures do not fully reflect the number of people infected each year and the true incidence has been estimated to be between 10 and 25 times higher than what reported figures indicate 12 . Cases very often remain unrecognized and are thus treated as other diseases or as "fever of unknown origin" 12 . In Uganda, it was noted that despite the reported increase in the number of individuals infected with Trypanosoma gambiense species in the 1990s, WHO estimated that the figures represent only 10–15% of the actual number of infected individuals 13 . Poor referral systems, limited surveillance coverage, difficulty and delay in diagnosis by the health facilities have been contributing to the under-reporting of zoonoses. Patients on the other hand have been seeking alternative services such as those offered by traditional healers and hence delay to present to health facilities or failing to present at all making data on their diseases not available for epidemiological records 13 14 15 16 .

Individuals as well as societies have been slow to act on zoonoses 17 . This could be due to insufficient systematic continuing education and opportunities to acquire new knowledge on zoonoses for those working in health institutions 18 . A physician attending to an ill veterinarian or a zookeeper will immediately suspect a wide array of diseases other than zoonoses; likewise a pediatrician attending to a sick child who recently received a puppy will not suspect an animal transmitted disease. All these underscores the fact that medical professionals have not been giving due consideration of animals as carriers of diseases that can be transmitted to humans 19 . This has resulted in poor quality of epidemiological data on zoonoses and their control measures on animal and human populations in particularly sub-Saharan Africa 20 21 .

Translation of knowledge into proper care of patients is among the critical areas in health care delivery 22 . This is only possible if health service providers have the right knowledge of health problems they are dealing with. In some countries active continuing education programmes have been intensified to consolidate the knowledge of health workers 23 . Innovative educational approaches have also been addressed in target specific groups of health workers to facilitate the implementation of guideline-based recommendations in the management of patients 24 . Because targeted education is an integral part in improving the diagnosis of diseases 25 , assessing the knowledge of practitioners could be an important step in identifying target receptors for public health education in Tanzania.

In total, four medical officers, six assistant medical officers and 27 clinical officers participated in the questionnaire (Table 1). Based on the location of the hospitals, seventeen medical practitioners were classified as working in rural hospitals and 20 in urban hospitals.

When level of training was considered in the analysis, there was no significant difference between the two groups (medical officers combined with assistant medical officers compared to clinical officers) with respect to knowledge of zoonoses, showing that level of training was not a factor in determining the type of response. Analysis was therefore conducted basing on comparison of responses from medical practitioners in urban and rural hospitals with "Partial" knowledge responses omitted.

Diagnosis and hence reporting of diseases depend largely on the level of understanding of the diseases. Knowledge of reservoirs of zoonoses and the way they are transmitted to humans has enabled not only their diagnosis and reporting but also their control 29 30 . For instance, knowledge of animal reservoir and transmission modes has enabled the identification and control of zoonoses outbreaks in the world such as Rift Valley fever in Kenya and Somalia 31 , Nipah virus in Malaysia and Singapore 32 , West Nile virus in USA 33 and Hendra virus in Australia 34 . It has however been noted in some countries that zoonoses are causing prolonged and unnotified human suffering 19 . Where malaria is endemic, diseases such as brucellosis and anthrax have been under-diagnosed because of their similarities in clinical presentations 12 . It is therefore important to optimize the diagnosis of other diseases such as zoonoses that have significant socio-economic impact on human life.

According to the population and housing census conducted in Tanzania in 2002, over 75% of the Tanzanian population is in rural areas and keep more livestock compared to their urban counterparts 26 35 . One would therefore expect more awareness of zoonoses in rural than in urban areas. In the current study however, it was found that practitioners in the rural areas had poor knowledge of how sleeping sickness is transmitted to humans and clinical features of anthrax and rabies. In both areas the practitioners had poor knowledge of how echinococcosis is transmited, how it presents, and how to go about investigating bovine tuberculosis in humans. Inadequate knowledge of any aspect of a disease is a potential contributing factor to misdiagnosis. For instance, if a practitioner is not well informed of how the diseases manifests or does not know how to investigate for its presence, there is a higher chance of misdiagnosis as one would not know which disease to investigate and how.

Adequate knowledge of animal reservoirs and transmission routes enable practitioners to focus on key areas related to the disease and hence reach the definitive diagnosis easier and earlier enough for prompt management of the disease 36 . The results of the study indicate that it is possible some zoonoses are missed by those entrusted with the duty of identifying them. Many reasons could explain these findings. These include practitioners concentrating on endemic diseases or diseases that have been common in their areas and ignore zoonoses that are increasingly becoming of public health importance and the possibility that teaching curricula in medical training institutions do no put due emphasis on zoonoses. The location of practitioners in the distant underprivileged rural areas could also explain the differences in the level of knowledge. Refresher programmes, seminars and workshops are more convenient and cheaper to conduct amongst staff in urban than in rural areas.

In Tanzania, like in other sub-Sahara African countries, there is a poor diagnostic capacity of many diseases including zoonoses. Laboratories particularly in the rural areas where there is the majority of the population are poorly equipped and can not diagnose most of the emerging and re-emerging diseases. Besides this fact, increased awareness of diseases among health workers and the community is still the most important area in disease control. Before more weight is put on increasing the diagnostic capacity of diseases, efforts should be made to equip the practitioners and the general community with adequate knowledge of zoonoses.

The study has shed some light on an area that has not been thought before as a possible contributing factor to under-reporting of diseases such as zoonoses. A larger study covering a wider area using random sampling could provide more conclusive results. In the current study, the prevailing weather condition necessitated a convenient sampling which could not be the best method to conduct such a study. The distribution of practitioners as presented in the study represents the distribution of health staff in Tanzania. The majority of senior health staff are found in urban areas of big cities where patients from rural facilities are referred for expertise management. However, it was established in the study that there was no significant difference of level of knowledge of zoonoses amongst different levels of practitioners.

Knowledge of medical practitioners of zoonoses could be among the contributing factors to under-diagnosis and under-reporting of zoonoses in Tanzania. There is a need for refresher programmes amongst the practitioners but also their teaching curricula should put more emphasis on zoonoses. This will provide practitioners with adequate knowledge of zoonoses prevention and control in the community. The study suggests that any intervention to be undertaken should give priority to practitioners in the rural areas.

Collaboration between veterinary and medical personnel should also be strengthened in Tanzania. This should include sharing knowledge on zoonoses and working together to identify and control zoonoses which are increasingly becoming diseases of public health importance in Tanzania.

The authors declare that they have no competing interests.

KJ, RK and GM were involved in the design of the study, KJ supervised field work, KJ was involved in data analysis and KJ, RK and GM were involved in manuscript write-up. All authors have read and approved the final manuscript.