Case finding of tuberculosis among mining communities in Ghana

Background Data on active TB case finding activities among artisanal gold mining communities (AMC) is limited. The study assessed the yield of TB cases from the TB screening activities among AMC in Ghana, the factors associated with TB in these communities and the correlation between the screening methods and a diagnosis of TB. Methods We conducted secondary data analyses of NTP program data collected from TB case finding activities using symptom screening and mobile X-ray implemented in hard to reach AMC. Yield of TB cases, number needed to screen (NNS) and the number needed to test (NNT) to detect a TB case were assessed and logistic regression were conducted to assess factors associated with TB. The performance of screening methods chest X-ray and symptoms in the detection of TB cases was also evaluated. Results In total 10,441 people from 78 communities in 24 districts were screened, 55% were female and 60% (6,296) were in the aged 25 to 54 years. Ninety-five TB cases were identified, 910 TB cases per 100,000 population screened; 5.6% of the TB cases were rifampicin resistant. Being male (aOR 5.96, 95% CI 3.25–10.92, P < 0.001), a miner (aOR 2.70, 95% CI 1.47–4.96, P = 0.001) and age group 35 to 54 years (aOR 2.27, 95% CI 1.35–3.84, P = 0.002) were risk factors for TB. NNS and NNT were 110 and 24 respectively.; Cough of any duration had the strongest association with X-ray suggestive of TB with a correlation coefficient of 0.48. Cough was most sensitive for a diagnosis of TB; sensitivity of 86.3% (95% CI 79.4–93.2) followed by X-ray, sensitivity 81.1% (95% CI 71.7–88.4). The specificities of the symptoms and X-rays ranged from 80.2% (cough) to 97.3% (sputum). Conclusion The high risk of TB in the artisanal mining communities and in miners in this study reinforces the need to target these populations with outreach programs particularly in hard to reach areas. The diagnostic value of cough highlights the usefulness of symptom screening in this population that may be harnessed even in the absence of X-ray to identify those suspected to have TB for further evaluation.


Background
Data on active TB case finding activities among mining communities is limited. The study assessed the yield of TB cases from the TB screening activities among mining communities, in Ghana, the factors associated with TB in these communities and the correlation between the screening methods and a diagnosis of TB.

Methods
We conducted secondary data analyses of NTP program data collected from TB case finding activities using symptom screening and mobile X-ray implemented in hard to reach mining communities. Yield of TB cases, number needed to screen (NNS) and the number needed to test (NNT) to detect a TB case were assessed and logistic regression were conducted to assess factors associated with TB. The performance of screening methods chest X-ray and symptoms in the detection of TB cases was also evaluated.

Conclusion
The high risk of TB in the mining communities and in miners in this study reinforces the need to target these populations with outreach programs particularly in hard to reach areas. The diagnostic value of cough highlights the usefulness of symptom screening in this population that may be harnessed even in the absence of X-ray to identify those suspected to have TB for further evaluation.   in Ghana, the factors associated with TB in these communities and the correlation between 40 the screening methods and a diagnosis of TB. 41

42
We conducted secondary data analyses of NTP program data collected from TB case finding 43 activities using symptom screening and mobile X-ray implemented in hard to reach mining 44 communities. Yield of TB cases, number needed to screen (NNS) and the number needed to 45 test (NNT) to detect a TB case were assessed and logistic regression were conducted to assess 46 factors associated with TB. The performance of screening methods chest X-ray and 47 symptoms in the detection of TB cases was also evaluated. 48

61
The high risk of TB in the mining communities and in miners in this study reinforces the 62 need to target these populations with outreach programs particularly in hard to reach areas. 63 The diagnostic value of cough highlights the usefulness of symptom screening in this 64 population that may be harnessed even in the absence of X-ray to identify those suspected to 65 have TB for further evaluation. High risk populations such as miners especially those engaged in illegal mining activities 106 may however be in hard to reach areas where they may encounter challenges accessing health 107 services including those for diagnosing TB [9,10]. Strategies such as deployment of mobile 108 teams proved to be useful in identifying people with TB in such hard to reach populations and 109 mining communities [10][11][12]. The implementation of active case finding (ACF) activities 110 using mobile teams in hard to reach mining communities in Myanmar identified significantly 111 higher TB prevalence in these townships than in the rest of the country [11]. In a study 112 conducted among South African miners, death from TB was more than 2 times higher among 113 those with TB diagnosed through passive case finding than those diagnosed through active 114 screening exercises [13]. 115 The Ghana NTP with funding from the Global Fund supported interventions to improve TB 116 case finding among mining communities in the country [14]. Given the limited data on active 117 TB case finding activities among mining communities in Ghana, the aim of the study was to 118 assess the yield of TB cases from the TB screening activities among mining communities, the 119 factors associated with TB in these communities and the correlation between the screening 120 methods and a diagnosis of TB. 121 in consultation with the respective Regional Health Directorates (RHD) selected 3, 6 and 12 130 districts respectively from each region with intense mining activities. 131

Study Design and Setting
These 21 districts consisted of an estimated total population of 2,120,660 in 2017 with TB 132 notification rates ranging from 37 to 394/100,000 population. In consultation with the RHD 133 and District Health Directorates (DHD), 78 communities were selected from these districts 134 for TB screening on the basis of TB case burden and being hard to reach thus with difficulty 135 accessing health services. 136 The NTP deployed mobile teams consisting of a team leader to superintend the screening 137 team, field coordinator to coordinate the field activities, radiographer to operate the digital X-138 ray machine, a physician-assistant to read the digital X-ray information technologist to 139 maintain network connectivity for the transmission and storage of X-ray images and data, 140 engineer to maintain the X -ray equipment and electrical generators, support team members 141 to conduct screening and collect sputum samples and drivers for the digital X-ray van and 142 vehicle to transport the team and other logistics. A week before the arrival of the team to the 143 community, the district TB focal person in collaboration with the community volunteers 144 undertook community mobilization using various methods including house to house 145 sensitization and community radio announcements informing the community member of the 146 upcoming free TB screening exercise, the date and the place where the team will be located. 147 148

Study population and data collection 149
All community members 15 years and above in the designated communities were eligible to 150 be included in the screening exercise though the dataset showed that a handful of children 151 less than 15 years were also screened. On the day of the exercise, community members 152 voluntarily presenting themselves at the location of the mobile team were registered and their 153 socio-demographic information recorded. A TB symptom screening was administered using a 154 questionnaire inquiring about the presence or absence of cough, fever, sputum production, 155 chest pain, weight loss, night sweats and fever. With the exception of the following; pregnant 156 or possibly pregnant women, those who for one reason or the other could not take or declined 157 taking the chest X-ray, everyone's X-ray was taken using the digital X-ray machine. A 158 physician assistant trained to read digital X-ray images categorized the X-ray finding into 3 159 groups: normal, abnormal suggestive of TB and abnormal but unlikely to be TB. Those 160 reported to have an abnormal X-ray suggestive of TB, those admitting having a cough of 2 161 weeks or any duration with at least one other symptom were presumed to have TB and were 162 requested to produce 1 spot sample of sputum. Also included in the presumed TB group and 163 asked to produce sputum for testing were those unable to have an X-ray taken as indicated 164 above and HIV positive individuals. The samples were transported to designated laboratories 165 with a GeneXpert machine for testing. The GeneXpert results were reported as 166 Mycobacterium tuberculosis (MTB) not detected, MTB detected rifampicin sensitive and 167 MTB detected rifampicin resistance detected. The field coordinator followed up on the results 168 and arranged for those found to have TB to be registered at the district level to start TB 169 treatment. The analyses of the treatment outcomes of these TB patients were beyond the 170 scope of this paper. 171

Statistical analyses 172
The dataset for the analyses consisted of the population that was screened. Descriptive 173 analyses were conducted for the following variables district, age, sex, occupation, X-ray 174 reading, presence of TB symptoms and GeneXpert results according to region. Among the 175 population screened, those with TB were compared with those without TB using Fisher's 176 exact test. The prevalence of TB was calculated and so were the number needed to screen 177 (NNS) and the number needed to test (NNT) to detect a TB case. Univariate logistic 178 regression was initially conducted to assess factors associated with TB. Given the few 179 characteristics being examined, all variables were then entered into a multivariable logistic 180 regression model. Odds ratio and 95% confidence (95%CI) interval were assessed. The 181 association between symptoms and having X-ray suggestive of TB was also determined by 182 assessing Cohen Kappa's coefficient as well as multivariable logistics regression. Finally, the 183 performance of screening methods chest X-ray and symptoms in the detection of TB cases 184 was also evaluated. In the sensitivity analysis all those not tested were regarded not to have  Table 1   Among those who had a chest X-ray done (10,329), almost 14% (1,448)

Factors associated with TB 251
In the bivariate analyses shown in Table 3, there was significant association between sex and 252 occupation and TB but not age. 253 254  Those with TB were more likely to be male, compared to females. In the Ashanti Region, all 261 but 1 of the 25 TB case identified were male. Even though miners formed 8 % of those 262 screened, 25% of those with TB were miners with a range of 18% in Brong Ahafo to 27% in 263 Western Region. 264 In univariate analyses, shown in Table 4, being a male, miner and unemployed were risk 265 factors for having TB as was coming from Western and Ashanti regions. In multivariate 266 analyses also in Table 4

Association between TB symptoms and X-ray suggestive of TB and TB diagnosis 277
In the analyses to assess the correlation between the symptoms and X-ray, shown in Table 5,  278 each of the symptoms, cough of any duration, sputum, chest pain, weight loss, night sweats 279 and fever was found to be significantly associated with X-ray suggestive of TB with cough 280 ranking the highest with Kappa Coefficient of 0.46. Findings were similar in the multivariate 281 analyses controlling for sex and age. 282 283 Miners and the communities in which they live are identified as being at higher risk for TB 322 than the general population [15]. Forging ahead with the End TB Strategy, responsive 323 programs for such key populations require data that will inform and guide stakeholders in the 324 appropriate planning and delivering of services [4,16]. In this study, which is the first to the 325 best of our knowledge to highlight on active TB case finding among mining communities in 326 Ghana using mobile X-ray and symptom screening, the overall rate of TB among the people 327 screened was found to be 910 per 100,000 people screened with 5.3% of the TB cases having 328 rifampicin resistant TB. The NNS and NNT were 110 and 24 respectively. Risk factors for 329 TB were being male, a miner and location in Ashanti and Western Regions. Cough regardless 330 of duration correlated with X-ray findings suggestive of TB and a diagnosis of TB. 331 The prevalence of TB among the communities screened in this study was more than two and 332 half times what was reported for the general population in Ghana in the TB prevalence survey 333 (356 bacteriologically confirmed TB per 100,000) re-iterating the data on the higher risk of 334 TB among mining communities [9,15,17]. The overall community TB prevalence was 335 however lower than has been reported in other mining communities such as the copper belt in 336 Zambia (1.2%) and mining communities in Myanmar (2.7%) [11,18]. The TB prevalence 337 among the miners in our study (2.65%), while also lower than that cited in other studies in 338 South Africa (5.4%) and Zambia (9.5%), was higher than that in miners in Myanmar (1.7%) 339 [11,18,19]. These differences may be related to several factors including varying methods for 340 diagnosing TB in the miners such as sputum microscopy as opposed to GeneXpert, the type 341 of mining, the exposure to silica dust and the duration of exposure to mining conditions [18]. 342 In the Zambia study, study participants were underground miners in copper mines while the 343 miners in our study were likely involved in gold and other mineral exploration using different 344 mining methods ranging from surface, to underground mining [18,[20][21][22]. 345 The rate of rifampicin resistance found among our study population (5.3%) was slightly 346 higher than the 4.3% drug resistance recorded among South African gold miners and their 347 dependents but much higher than the 1.3% reported from the 2018 national surveillance data 348 from Ghana [1,23]. The setting in the South Africa study was a formal gold mine 349 employment with a well-functioning TB control program that was keeping track of drug 350 resistant TB. In contrast, but for the outreach program to these mining communities, it is not 351 possible to tell how long it would have taken for these drug resistant cases to be detected. 352 This highlights the importance of mainstreaming TB screening services for such high risk 353 populations [24]. 354 In general males are more at risk for TB in Ghana and other countries and our study findings 355 are in line with this [17,24]. It is reported that in illegal artisanal small-scale mining settings 356 in Ghana, women may constitute up to 50% of the labour force and play various roles 357 [22,25,26] in reading of X-rays. [33,35] In our study the sensitivity (86.3%) of cough of any duration as 370 a screening tool for the detection of TB was relatively high and even ranked higher than that 371 of chest X-ray even though it was comparable. The combination of cough and X-ray raised 372 the sensitivity to 95%. This finding is similar to what van't Hoog and colleagues reported in a 373 prevalence study in Kenya even though the cough in their study was 2 weeks or more in 374 duration [33]. 375 In this TB case detection outreach to the mining communities, those who were eligible to 376 produce sputum for testing consisted of those who had an X-ray suggestive of TB, symptoms 377 suggestive of TB and those who for some reason could not take X-ray totalling about 23% of 378 the screened population. This is relatively higher compared to the 13% identified during the 379 TB prevalence survey [17]. The difference may be related to the expertise in reading the X-380 rays (radiologists in the prevalence survey as opposed to physician assistants in our survey) 381 and the cough criteria (2 weeks cough in the prevalence survey). The yield of TB cases was 382 however much higher and may justify the use of this approach of combining symptom and x-383 ray screening used in the mining communities as opposed to a sequential algorithm of 384 symptom screening following by X-ray. Further studies assessing a cost benefit analyses and 385 factoring in yield of TB cases may shed more light on the more beneficial and cost-effective 386 approach to use in high risk populations in hard to reach communities. 387 This study has a number of limitations which include the point that the diagnosis of TB was 388 based on results from GeneXpert to the exclusion of those who may have had clinically 389 diagnosed TB. Secondly, having used routine data already collected, there was no control 390 over standardizing the measurement and reporting of variables to enable further analyses of 391 interest nor was it not possible to apply quality control measures or checks. For example, in 392 the data set the recording cough of any duration without always specifying the duration for 393 each participant limited the ability to assess the sensitivities of the different durations of 394 cough and the diagnosis of TB. It is therefore being recommended to the NTP to have 395 standardized data collection tools for all teams engaged in outreach programs to enable 396 optimum analyses of data to inform programming. Finally, with this study being conducted in 397 mining communities the findings may not be generalizable to other communities in the 398 country. 399

400
The high risk of TB in the mining communities and in miners as shown in this study 401 reinforces the need to target these populations with outreach programs particularly given that 402 they may be in hard to reach areas. Even though the combination of cough and chest X-ray 403 had the highest sensitivity, the diagnostic value of cough highlights the usefulness of 404 symptom screening in this population that may be harnessed even in the absence of X-ray to 405 identify those suspected to have TB for TB diagnostic evaluation. 406