This is the first time that TB knowledge has been assessed in this region. Our community study measured knowledge in ordinary villagers about this important infection in a typical region of sub-Saharan Africa. We also measured the HIV knowledge in the same survey and we were thus able to compare the community knowledge of both infections. We consider the results from this knowledge comparison to be important for policy makers who have to look at priorities for infectious disease programs in the context of limited district health budgets. The strength of our study was that we used a mixed design with quantitative and qualitative methods which enhanced the understanding on our topic and improved the validity of the findings. To the best of our knowledge, we could not find any published information from sub-Saharan Africa on a comparison between TB and HIV knowledge: therefore we consider our study as a contribution to the literature. Additionally, misconceptions about disease aetiology, treatment and prevention of TB and HIV are important to inform the delivery of effective TB and HIV treatment and prevention programs in the district. Service utilization and adherence to treatment and prevention protocols should ideally be based on these findings.
The TB knowledge in the population was unexpectedly low with a mean summary score of 33%. This result is classified as seriously deficient according to a district rating system suggested by Janowsky et al. and which is frequently used . This low TB knowledge is corroborated by the additional finding that only 31% of study participants knew that TB is an airborne infection. Some of the “misconceptions” about TB transmission which we found were also reported in the other studies from eastern Uganda and elsewhere [7, 11]. Sharing cups emerged as a very frequent response by participants trying to explain the transmission of TB in our study, but was not reported in the other studies [4, 6, 11, 15]. Although this belief may have some truth in it (as saliva contamination could possibly transmit Mycobacterium tuberculosis from one person to another person when cups are shared without cleaning them, and two people sharing food and drinks are likely close enough for airborne transmission to occur), it distracts somewhat from the most important fact that the community must know that the major mode of TB transmission is airborne. This popular belief is not necessarily harmful as proper hygiene and not sharing cups and utensils lead generally to preventative health practice.
Surprising to us was the understanding by some that TB is an untreatable hereditary disease. This was also mentioned in each focus group discussion. It has also been described elsewhere in Uganda and in Tanzania [6, 11]. This is a serious misconception and should be vigorously targeted in education programs, since conceptualising TB as a hereditary disease which cannot be treated has huge potential to negatively affect decisions to seek TB care. Another significant finding was that 20% of respondents did not know that TB can be cured which was also confirmed in the focus group discussions by participants. This has also serious implications for control: if people do not believe that their illness can be treated they will not seek care. This last finding is different from the other studies in Uganda, Tanzania, and Rwanda that found that people universally understood that TB was curable [6, 9, 11].
Older age and rural residence predicted higher TB knowledge. This is a surprising result, as we expected the opposite. The survey from Ethiopia on TB found that literacy predicted better knowledge of TB causation, transmission, and prevention . As literacy is usually higher in younger people , our study finding of better TB knowledge in older persons is contrary to the findings from the Ethiopian study. Our hypothesis for an explanation of this finding is that TB education in the study area was a major focus within a Primary Health Care program in the mid 1980’s and 1990’s which was supported by the German Government and focused on rural areas in Kabarole District. As such, individuals who were old enough during that period to understand TB information presented then were positively impacted and increased their knowledge. Later on, this program was abandoned and priorities were given to HIV prevention and care. Therefore younger people may have remained uninformed about TB issues and consequently had a lower knowledge about TB. These previous program funding developments would also explain why TB knowledge was better in rural areas compared to urban areas, as one would expect the opposite.
HIV knowledge was much higher than TB knowledge with highly statistically significant differences (58% vs 34% knowledge, p < 0.001). The HIV knowledge was comparable to a survey conducted in 2005 in the same study area using a similar methodology and similar knowledge questions. In the 2005 survey the HIV knowledge in the general population was 59% . Predictors of higher HIV knowledge were younger age and urban residence, similar to other studies which found that younger persons living in urban areas generally have better health knowledge than older persons residing in rural areas of sub-Saharan Africa. Knowledge about mother to child transmission of HIV was very low in our sample, with only 14% of respondents mentioning it as a means of HIV transmission, much lower than 75% of respondents reported by the study in Ethiopia . This in line with other data from Kabarole District, where knowledge of mother-to-child transmission of HIV was found to be seriously deficient and the services of PMTCT (Prevention of Mother-to-Child Transmission of HIV) grossly underutilized .
The lower TB knowledge compared to the HIV knowledge was not unexpected. We suspect that it is a reflection of recent shifts in program development where there has been a major focus on antiretroviral treatment of HIV infection and a corresponding allocation of resources to HIV/AIDS programs. This shift of resources has been confirmed in a previous study in Kabarole, where it was found that human and financial resources were allocated to HIV/AIDS programs to the disadvantage of other communicable disease programs . Health care workers in the study area could only informally confirm the poor management and outcomes of TB control in the district, as no hard reliable data is available. However, while we expected a lower knowledge of TB compared to HIV, we were surprised about the magnitude of the knowledge difference. TB’s status as a neglected infection [20, 21] seems to be reinforced if we looked on community knowledge as an indicator.
The interpretation of responses for both the questionnaire and the FGDs pose some ambiguity, given that they can be evaluated either on a strictly scientific basis level or on a more inclusive approach, whereby peoples’ perceptions and beliefs are considered. An even more inclusive approach would be to consider the cultural factors of a particular population and develop a cultural model of health communication based on sound scientific information, community perceptions and the local culture. According to our experience from sub-Saharan Africa, this approach has rarely been selected. To achieve this, an interesting follow-up project could be undertaken here with participation of medical, public health and anthropological experts and a broad community involvement.
We used a non-standardized questionnaire; therefore, interpretation of the level of knowledge and its comparability with other studies is limited. However, the comparison of the TB and HIV knowledge score within our study can be considered valid as we used similar types of questions and the same methodology. The questionnaire results were reliable, as shown in the test-re-test agreement and valid when compared to the results from the qualitative study component which confirmed most of the quantitative results.
We cannot exclude interview bias and/or social desirability bias, as our study dealt with a sensitive topic. We minimized it by using highly trained interviewers with local knowledge of the culture and the language and supervised them closely during the data collection process.
The ambiguity of responses and how they are interpreted is acknowledged. In this study, we were leaning toward a more scientific evaluation of the responses, which may not be always the best approach to better health communication in communities.
Because participants were recruited during weekdays, there may be an under-representation of individuals employed in the formal sector since they would not have been home during the day. It was not possible within the time limits of the field work to include those individuals by making additional visits to the homes.
Predictors of the knowledge scores are based on cross-sectional data and therefore cannot be considered as causal.