This study was conducted to assess the TBIC-related knowledge and practices of nurses in two secondary health facilities in Ibadan, Nigeria, and their association with the socio-demographic characteristics of the nurses. The results revealed mean knowledge and practice scores of 68.2% and 79.9% respectively. Furthermore, with the cut-off for good knowledge and practice scores set at 80% and 100% respectively, it was observed that the majority of them had poor knowledge and practices. This is consistent with reports from other studies in Nigeria where, generally, poor levels have been demonstrated among HCWs [18,19,23,24].
Various methods of assessing and scoring TB-related knowledge and practices have been observed from previous studies. While some of these simply stated the mean and used this as the cut-off to categorize the scores of the participants, others had arbitrary cut-off points [25,27]. Additionally, the categories may be varied: (i) good, moderate, and poor; (ii) good and poor; (iii) good, fair, and poor; and (iv) proper, and improper [7,15,27,28]. The majority of the nurses in this study (>80%) were able to correctly identify the constitutional symptoms of TB (cough of 2-3 weeks duration, bloody sputum, night sweats weight loss and fever). This is similar to the findings by Bhebhe et al. among HCWs in Lesotho, except that in their study, only 53.5% considered fever to be a symptom of TB . Most of the nurses gave correct responses to the questions related to the mode of transmission of TB. For instance, 95% recognized that TB can be transmitted by coughing (96.5%), and its transmission can be reduced by practising cough etiquette/hygiene (96%), and by opening the windows of a room that has a TB patient in it (90%). This agrees with the findings in studies conducted in Nigeria and Northwest Ethiopia [24,28]. It also aligns with a South African study, where Kanjee et al. reported that “most of the information (knowledge) items were answered correctly by over 70% of respondents with some exceptions” and that the “HCWs were generally well informed about TB transmission” . Other researchers have made similar observations among HCWs in Free State Province, South Africa .
The mean knowledge score of 68.2% reported in this study is higher than the findings of 61%, and 61.5% reported by previous investigators [9,33]. The poor knowledge level of TBIC conforms with the results of a study by Woith et al. among HCWs in Russia . In contrast to the poor knowledge noticed in the present study, some researchers have previously reported “good” or “adequate” TBIC knowledge among HCWs, although lower cut-off points were used in their studies. For instance, Bhebhe et al. reported that 89.2% of HCWs in their study in Lesotho had “appropriate” TBIC knowledge, but the cut-off used to define “good” was 70%, which is lower than 80% adopted for the present study . Furthermore, the mean score of 61.5% reported by them is lower than 68.2% observed in this study. Buregyeya et al. similarly reported that 69 % of HCWs in a study in Uganda had adequate TBIC knowledge, with the cut-off taken as 70% . A study in Ethiopia by Temesgen and Demissie revealed that 74.4% of health professionals had “good” knowledge, using a cut-off of 60% .
In terms of the practice of TBIC measures, only two out of the six items had more than 70% of the nurses reporting them as “always” practised: cough etiquette/hygiene and opening of windows. Ekuma et al. also reported similar poor “always” results for practice items in a study in Nigeria . This is, however, different from the reports of a South African study by Engelbrecht et al., where four items out of 12 had more than 80% of the respondents who “always” practised them: patient fast-tracking, screening, window-opening, and collection of sputum specimens from coughing patients . However, it is notable that the frequency of practice of the measures in that study, just like in the present study, was self-reported by the respondents; and the researchers noted a discrepancy between the self-reports and observed practices. We noticed in our study that the proportions of nurses reporting various TBIC measures as “always” practised were less than those that recorded correct responses to related questions on the knowledge scale. This incongruity suggests that, although good levels of nurses’ TBIC knowledge have been shown to be closely associated with good practices, it is not its only determinant . Other factors that influence proper TBIC practices include clear policy directives, appropriate triage system and separation of coughing patients, availability of personal protection equipment, reasonable workload, adequate and well-ventilated clinic space, among others . Findings from studies conducted in LMIC, where cost-effective TBIC measures are best suited, are in overwhelming support of the results of the present study. Inadequate practice of TBIC measures have been reported in Nigeria, South Africa, Lesotho and Ethiopia [8,9,15,28]. Tamir and his co-workers, using 80% as their cut-off, found that only 38% of HCWs in their Ethiopian study had overall proper TBIC practices . Even though Temesgen and Demissie reported an overall “good” TBIC practice in Ethiopia, with a cut-off of 50%, specific practices were still poor . Poor levels of implementation of TBIC measures were also reported by Bhebhe et al. and Kanjee et al. in Lesotho and South Africa respectively [7,13]. The divergence in the proportions with good knowledge and practices noticed between the present study and previous ones cited could be due to the different cut-off points and scoring systems used. It is noteworthy that higher cut-off points were used in the present study. However, the finding of large proportions of nurses with poor levels of knowledge and practices in this study is not completely unexpected because, although TBIC guidelines had been released in the country at the time the study was conducted, implementation was still in its infancy. Moreover, many facilities, including the study sites, were yet to benefit from the roll-out package. [14,19,20,23].
The nurses’ TBIC knowledge was not significantly related to all the socio-demographic factors considered. This result is in conformity with the observations from previous studies. Temesgen and Demissie reported that TBIC knowledge among HCWs was not significantly associated with work experience and age category . Buregyeya et al. also noted that it was not associated with age and sex. On the contrary, Gizaw et al. showed an association between knowledge and work experience in a study in Ethiopia, with HCWs who had more than six years’ work experience being more knowledgeable than those with less than three years’ experience [25,27]. The lack of association they observed between knowledge and age, however, agrees with our findings. Besides the revelation from the present study that the more experienced nurses had statistically significant lower odds of having good practices, the observations concerning the other socio-demographic factors are in alignment with some previous studies. According to Mugomeri et al., the nurses’ age and TB ward work experience did not significantly influence their practice of TBIC measures in a study in Lesotho . Similarly, Temesgen and Demissie noted that the practices were not related to work experience and age category . Also, work experience, age, gender and marital status were not statistically related to practice in the study by Tamir et al. . The more experienced nurses in the present study were less likely to obtain good practices maybe because most of them, who are usually in the senior cadre, serve as unit heads or managers and their work duties include mainly managerial functions such as managing work schedule and roster, ensuring compliance with work policies and protocols, commodity management, among others. They may not necessarily carry out actual TBIC work practices, in line with official responsibilities assigned to the different professional levels. Some investigators in Nigeria have reported that nurse managers tend to focus more on their managerial duties at the expense of their clinical roles [36,37]. The distribution of the respondents in terms of sex and marital status was heavily skewed in favour of females (97.0%) and married ones (91.5%). This pattern reflects the profile of the nursing workforce in Nigeria, as reported by other investigators . Furthermore, the wide confidence intervals recorded for the regression analysis using the two factors as separate independent variables, and knowledge and practice as dependent variables indicate low precision. These elements should be considered when interpreting the results related to sex and marital status.
The association between TBIC knowledge and practices was not statistically significant, although the nurses with good knowledge scores were less likely to obtain good practice scores. This should, however, be interpreted with caution as the 2x2 contingency table showed that there were no nurses with good knowledge scores that also had good practice scores, hence the use of exact logistic regression approach to produce the point estimate and confidence interval, in line with the recommendations by Hosmer and Lemeshow for analyzing cells with zero or sparse counts . This absence of association between knowledge and practices aligns with the findings by Gizaw et al. . It, however, contracts with reports from South Africa and Northwest Ethiopia, where a significant association was reported between them [15,26]. Furthermore, TBIC knowledge was reported to be significantly associated with training on TBIC in some of the previous studies referenced here [25-27]. Knowledge might not have significantly influenced practices in the present study because at the time the study was conducted, nurses at the study sites were yet to be trained on the newly-introduced TBIC package [14,23]. The findings from this study suggest that diffusion of actions and professional socialization might have influenced the routine work practices of the nurses, including performance of TB preventive measures, irrespective of their knowledge and socio-demographic characteristics as these were generally not significantly associated with their TBIC practices [40,41]. However, the positive influence of TBIC training on specific aspects or overall practice of TB preventive measures revealed in some studies, and the well-known interplay of training and knowledge on practice underscore the importance of conducting trainings on TBIC for the nurses in order to equip them with essential skills to improve their practices [25,26,42].
Limitations of the study
It was difficult to carry out direct observation of TBIC practices by the nurses as this is time-consuming and requires the engagement of more research personnel as observers. Self-reports were relied on and there is the likelihood of a discrepancy between this and direct observation of actual practices, as previously reported by Engelbrecht et al. . There could have been over-reporting of the practices by the participants because of the social acceptability of being perceived as doing the proper thing (social desirability bias). Also, for the purpose of obtaining an all-inclusive sample, nurses in all the units of the hospitals (out-patient units, in-patient wards, emergency room, operating theatre, among others) were encouraged to participate in the study. Although the nurses rotate through various work units, some of the duties and assigned tasks may not be directly related to TB care and it might have been challenging for participants who had not recently been involved in providing TB-related care to respond correctly to the practice items (recall bias). Furthermore, the study questionnaire tended towards administrative and environmental control measures (specifically, natural ventilation system) as managerial control measures are more in the purview of the facility management while personal protection equipment (respirators) are mostly available for use in specialized MDR-TB centres. It would be beneficial to further review and expand the items in the study instrument to make for the implementation of a more comprehensive package of TBIC measures.