Abstract
According to the World Health Organization, multidrug-resistant tuberculosis (MDR-TB) represents a major obstacle towards successful TB treatment and control. In Dakar, MDR-TB management began in 2010 with the strengthening of diagnostic resources. The objective of this study was to identify the factors associated with multidrug-resistant tuberculosis in Dakar between 2010 and 2016. We conducted a case-control study from January 10 to February 28, 2017 in tuberculosis centers in Dakar. of 169 cases and 507 controls. We used logistic regression with Epi-info version 7.2.1. to estimate the odds ratios of association. Factors significantly associated with MDR-TB were: residing in a periurban area (ORa=1.8; 95% CI (1.5-4.9); p=0.024), presence of MDR-TB in the entourage of patient (ORa=7.0; 95% CI (6.1-9.5); p=0.002), previous treatment failure (ORa=29.5; 95% CI (27.3-30.1); p=0.000), treatment not directly observed by a health care provider (ORa=4.3; 95% CI (4.1-7,2); p=0.000) and irregularity of treatment (ORa=1.7; 95% CI (0.5-5.4); p=0.037). Focusing interventions on population at-risk will prevent MDR-TB.
Key words: multidrug-resistant tuberculosis, factors, Dakar
Introduction
The emergence of drug-resistant strains is undermining efforts to control tuberculosis (TB) and is a public health problem. Worldwide, according to WHO, the number of reported cases of multidrug-resistant tuberculosis (MDR-TB) in the countries that are highly affected by this form of tuberculosis has reached a worrisome proportion. 1 Approximately 190,000 deaths have been recorded, representing a casefatality rate of 40%. In Africa, the proportion of reported MDR-TB cases is relatively low (2.2%). Given the lack of laboratories, it is likely that current figures underestimate the true burden of disease.2 Being male,3,4 age,5,6 previous treatment failure or an History of previous TB treatment,6–9 residing in a peri-urban area, history of multidrug- resistant tuberculosis in the entourage of patient are found to be factors associated with MDR-TB. The association between diabetes and MDR-TB is variously appreciated.10-12
In Senegal, the number of patients identified with multidrug-resistant tuberculosis has steadily increased from 36 cases in 2010 to 77 cases in 2015.In 2014, the proportion of MDR-TB was 0.5% in new cases of TB and 17.9% in previously treated cases (National Tuberculosis Control Program). In Dakar, the control of multidrug-resistant tuberculosis cases began in 2010 with the strengthening of diagnostic facilities. but the factors associated with MDR-TB are not yet known. The knowledge of these factors is crucial to guide the actors of TB care and TB prevention actions to overcome this scourge. The objective of this study was to identify the factors associated with MDRTB among tuberculosis patients in Dakar between 2010 and 2016.
Materials and Methods
Type of study
Case-control analytical study.
Study population
The study population consisted of confirmed tuberculosis patients. A case was defined as any tuberculosis patient with a test confirming infection with Mycobacterium tuberculosis, resistant to rifampicin alone or in combination with isoniazid, registered in Dakar between January 1, 2010 and December 31, 2016. A control was defined as any patient who developed confirmed tuberculosis, registered in Dakar and reported cured at first-line treatment between January 1, 2010 and December 31, 2016.
The study population consisted of confirmed tuberculosis patients. A case was defined as any tuberculosis patient with a test confirming infection with Mycobacterium tuberculosis, resistant to rifampicin alone or in combination with isoniazid, registered in Dakar between January 1, 2010 and December 31, 2016. A control was defined as any patient who developed confirmed tuberculosis, registered in Dakar and reported cured at first-line treatment between January 1, 2010 and December 31, 2016.
Sampling
We took 1 case for 3 unmatched controls and used the parameters of the Fleiss formula. Therefore, we got 676 for the size of our sample, including 169 cases and 507 controls. A random sampling was performed using the patient list.
Data collection
We conducted a literature review using information sources from registry and patient records and collected socio-demographic, clinical and therapeutic characteristics on a data collection sheet.
Statistical analyses
Proportions, central tendency and dispersion measurements, odds ratios were calculated.
In the univariate analysis, we calculated the Odds ratio (OR) measuring the link between each factor with MDR-TB; all variables with p≤0.20 were included in a multivariate logistic regression. Analyses were performed with Statistical Analysis Software Epi-info version 7.2.1. For the interpretation of association, the threshold of significance retained was 5%.
Results
Socio-demographic characteristics of patients of the study
The average age was 31.2 ± 10.4 for cases and 33.6 ± 12.9 for controls. Ages ranged from 8.0 to 72.0 years for cases and 7.0 to 81.0 years for controls. Weight loss was noted in cases and weight gain in controls. (Table 1).
The proportion of patients with previous treatment (number of previous treatment episode greater than or equal to 1) was 47.9% in cases and 6.5% in controls. The average time to first treatment was 5.4 days for cases and 2.4 days for controls.
Univariate analysis
The socio-demographic characteristics associated with MDR-TB were 15-59 years of age and suburban residence (Table 2). The clinical characteristics associated with MDR-TB were: co-morbidity (diabetes), alcohol consumption and having MDR-TB in entourage of patient (Table 3). Therapeutic characteristics associated with MDR-TB were: previous treatment failure, treatment not directly observed (by a health worker) and irregularity of treatment (Table 4).
Multivariate analysis
After adjustment for all independent variables in the study (Table 5), the factors that remained significantly associated with MDR-TB were suburban residence (ORa=1.8; 95% CI (1.5-4.9); p=0.024), having MDR-TB in entourage of patient (ORa=7.0; 95% CI (6.1-9.5); p=0.002), previous failure (ORa=29.5; 95% CI (27.3-30.1); p=0.000), treatment not directly observed (by a health provider) (ORa=4.3; 95% CI (4.1-7,2); p=0.000), irregularity of treatment (ORa=1.7; 95% CI (0.5-5.4); p=0.037).
Discussion
In our study, the significant factors associated with MDR-TB were residing in a peri-urban area, history of multidrug-resistant tuberculosis in the entourage of patient, previous treatment failure, treatment not directly observed (by a provider) and irregularity of treatment.
Socio-demographic characteristics
Our study was not able to demonstrate an association between age and MDR-TB. A survey conducted in the Congo reported that this age group was less exposed to multidrug- resistant tuberculosis.10 However, a 2013 study conducted in France showed that these patients were more exposed to MDR-TB than in our study.13 This difference in results could be explained by the fact that in France, multidrug-resistant tuberculosis prevails in foreigners, who are most often young immigrants in quest of employment and unfavorable living conditions (precariousness, promiscuity, imprisonment ...) and therefore exposed to the disease. We found in our survey that residing in a peri-urban area was associated with multidrug resistance. The peri-urban area is comprised mainly of vulnerable population living in difficult socio-economic conditions. This peripheral zone of Dakar is mainly characterized by poorly controlled urbanization, difficult access areas, lack of basic social infrastructure, low household income, and high human concentration in households. André MK10 showed that “people living in peri-urban households were exposed to multidrug resistance”.
Table 1.
Description of MDR-TB cases and controls, Dakar, 2010-2016.
| Characteristics | Case n (%) | Control n (%) | |
|---|---|---|---|
| Age (year) | 0-4 | 0 (0,0) | 0 (0,0) |
| 5-14 | 18 (10,7) | 50 (9,9) | |
| 15-59 | 149 (88,2) | 408 (80,5) | |
| ≥ 60 | 2 (1,2) | 49 (9,6) | |
| Sex | Female | 55 (32,5) | 128 (25,2) |
| Male | 114 (67,5) | 379 (74,8) | |
| Profession | Student | 34 (20,1) | 79 (15,6) |
| self-employed | 94 (55,6) | 308 (60,8) | |
| Employee (State) | 5 (3,0) | 21 (4,1) | |
| Retirement | 1 (0,6) | 19 (3,7) | |
| Other | 5 (3,0) | 4 (0,8) | |
| No occupation | 30 (17,7) | 76 (15,0) | |
| Average weight in Kg | At diagnosis (W0) | 54,4±11,4 | 51,9±10,6 |
| (+/- standard deviation) | At the 1st Control (W1) | 53,7±11,0 | 54,7±11,1 |
| At the 2nd Control (W2) | 52,8±11,1 | 57,5±11,6 | |
| Weight variation (Kg) | W1-W0 | -0,7 | 2,8 |
| W2-W1 | -0,9 | 2,8 | |
| W2-W0 | -1,6 | 5,4 | |
| Number of previous | 0 | 88 (52,1) | 474 (93,5) |
| treatment episode | 1 | 66 (39,0) | 33 (6,5) |
| 2 | 12 (7,1) | 0 (0,0) | |
| 3 | 3 (1,8) | 0 (0,0) | |
| Average time to first treatment (+/- standard deviation) | 5,4±10,3 | 2,4±4,9 |
W0: Weight at diagnostic, W1: 1st weight control at 2-3 months, W2: 2nd weight control at 6-8 months.
Table 2.
Sociodemographic characteristics associated with MDR-TB cases Dakar, 2010-2016 in univariate analysis.
| Socio-demographic characteristics | Cases n | Controls n | OR 95% CI | p | |
|---|---|---|---|---|---|
| (%) | (%) | ||||
| Age (year) | Less than 15 or more than 59 | 20 (11.8) | 99 (19.5) | 1 | |
| 15-59 | 149 (88.2) | 408 (80.5) | 1.8 (1.1-3.0) | 0,023 | |
| Sex | Male | 114 (67.5) | 379 (74.8) | 1 | 0,064 |
| Female | 55 (32.5) | 128 (25.2) | 1.4 (0.9-2.1) | 0.064 | |
| Profession | Employees | 6 (3,5) | 21 (4.1) | 1 | 0,733 |
| Unemployed | 163 (96.5) | 486 (95.9) | 1.2 (0.5-2.9) | 0.733 | |
| Area of residence | Other | 126 (74.6) | 419 (82.6) | 1 | 0,021 |
| Peri-urban | 43 (25.4) | 88 (17.4) | 1.6 (1.1-2.5) | 0.021 |
Table 3.
Clinical characteristics associated with MDR-TB cases Dakar, 2010-2016 in univariate analysis.
| Clinical characteristics | Cases n | Controls n | OR 95% CI | p | |
|---|---|---|---|---|---|
| (%) | (%) | ||||
| HIV Infection | No | 165 (97.6) | 487 (96.1) | 1 | |
| Yes | 4 (2.4) | 20 (3.9) | 0.6 (0.2-1.7) | 0.337 | |
| Diabetes | No | 163 (96.4) | 502 (99,0) | 1 | |
| Yes | 6 (3.6) | 5 (1.0) | 3.7 (1.1-12.3) | 0.026 | |
| Alcohol | No | 165 (97.6) | 505 (99.6) | 1 | |
| Yes | 4 (2.4) | 2 (0.4) | 6.1 (1.1-33.7) | 0.018 | |
| Tobacco | No | 160 (94.7) | 488 (96.2) | 1 | |
| Yes | 9 (5.3) | 19 (3.8) | 1.4 (0.6-3.2) | 0.373 | |
| History of MDR-TBis in the entourage of patient | No | 145 (85.8) | 505 (99.6) | 1 | |
| Yes | 24 (14.2) | 2 (0.4) | 41.8 (9.8-178.9) | 0.000 |
Table 4.
Therapeutic Characteristics associated with MDR-TB Cases Dakar, 2010-2016 in univariate analysis.
| Therapeutic characteristics | Cases n | Controls n | OR 95% CI | p | |
|---|---|---|---|---|---|
| (%) | (%) | ||||
| Number of previous episodes of tuberculosis treatment | 0 | 88 (52.1) | 474 (93.5) | 1 | |
| ≥1 | 81 (47.9) | 33 (6.5) | 13.2 (8.3-21.0) | 0.000 | |
| Previous treatment failure | No | 126 (74.6) | 503 (99.2) | 1 | |
| Yes | 43 (25.4) | 4 (0.8) | 42.9 (15.1-121.8) | 0.000 | |
| Directly Observed Treatment | Sanitary | 70 (41.4) | 420 (82.8) | 1 | |
| Non-sanitary | 99 (58.6) | 87 (17.2) | 6.8 (4.7-10.0) | 0.000 | |
| Irregularity of treatment | No | 145 (85.8) | 489 (96.4) | 1 | |
| Yes | 24 (14.2) | 18 (3.6) | 4.5 (2.4-8.5) | 0.000 |
Table 5.
Factors associated with multidrug-resistant tuberculosis, Dakar, 2010-2016 in multivariate analysis.
| Factors | Cases n | Controls n | Adjusted OR | p | |
|---|---|---|---|---|---|
| (%) | (%) | (95% CI) | |||
| Age | Less than 15 or more than 59 | 20 (11.8) | 99 (19.5) | 1 | |
| [15-59] | 149 (88.2) | 408 (80.5) | 0.7 (0.3-1.7) | 0.468 | |
| Sex | Male | 114 (67.5) | 379 (74.8) | 1 | |
| Female | 55 (32.5) | 128 (25.2) | 0.9 (0.4-1.2) | 0.105 | |
| Area of residence | Other | 126 (74.6) | 419 (82.6) | 1 | |
| Peri-urban | 43 (25.4) | 88 (17.4) | 1.8 (1.5-4.9) | 0.024* | |
| Diabetes | Non | 163 (96.4) | 502 (99,0) | ||
| Yes | 6 (3.6) | 5 (1.0) | 1.3 (0.1-1.6) | 0.114 | |
| Alcohol | No | 165 (97.6) | 505 (99.6) | 1 | |
| Yes | 4 (2.4) | 2 (0.4) | 1,5 (0.2-2.0) | 0.235 | |
| History of multidrug-resistant | No | 145 (85.8) | 505 (99.6) | 1 | |
| Tuberculosis in the entourage of patient | Yes | 24 (14.2) | 2 (0.4) | 7.0 (6.1-9.5) | 0.002* |
| Number of previous episodes of tuberculosis treatment | 0 | 88 (52.1) | 474 (93.5) | 1 | |
| ≥1 | 81 (47.9) | 33 (6.5) | 1.4 (1.1-2.0) | 0.321 | |
| Previous treatment | No | 126 (74.6) | 503 (99.2) | 1 | |
| failure | Yes | 43 (25.4) | 4 (0.8) | 29.5 (27.3-30.1) | 0.000 |
| Directly Observed | Sanitary | 70 (41.4) | 420 (82.8) | 1 | |
| Treatment (DOT) | Non-sanitary | 99 (58.6) | 87 (17.2) | 4.3 (4.1-7.2) | 0.000 |
| Irregularity of treatment | No | 145 (85.8) | 489 (96.4) | 1 | |
| Yes | 24 (14.2) | 18 (3.6) | 1.7 (0.5-5.4) | 0.037 |
*Statistical significant.
Clinical characteristics
Our study showed that diabetes was not significantly associated with MDR-TB. In this regard, the association between diabetes and MDR-TB is variously appreciated by the authors.10-12 Patients having multidrug-resistant TB in their entourage were more exposed to MDR-TB. Other authors found a history of multidrug-resistant tuberculosis in the family of cases.10 The presence of MDRTB cases in the family is a significant determinant in the onset of the disease. Patients in close contact with MDR-TB who have tuberculosis may be suspected of multidrug resistance. Exposure to multidrug resistance is so high in these patients that it would be acceptable, according to some authors,7,14-16 to implement a treatment similar to that of the index case while waiting for the results of the anti-tuberculosis drug susceptibility test, even if changes in treatment are required later following the test result.
Therapeutic characteristics
According to WHO, the classification of patients according to therapeutic backgrounds, makes it possible to note in our study that there was an association between multiresistance and some groups of patients (previous treatment failure). According to the meta-analysis of Faustini et al., which represents a systematic review of 29 publications, failure of prior anti-tuberculous therapy increased exposure to a multidrugresistant strain by up to 10 times.7 Just like this aforementioned study, Ahmad et al reported that treatment failure was the main factor associated with MDR-TB.17 Treatment failure is mainly associated with poor adherence to treatment. In our study, patients who had irregular treatment were exposed to MDR-TB. This can be explained by the lack of systematic application of DOT (Directly Observed Treatment). DOT was not applied in 58.6% of the cases and in 17.2% of the controls. Taking anti-tuberculosis drugs in front of a health provider every day is an act of safety for the patient and those around him; indeed, there is an association between unsupervised drug intake and multidrug resistance (p=0.000).
Conclusions
Significant factors associated with the occurrence of the MDR-TB were: residing in a peri-urban area, presence of multidrugresistant tuberculosis in the entourage of patient, previous treatment failure, treatment not directly observed (by a provider) and irregularity of treatment. Active search of MDR-TB cases in the patient’s environment and adherence to the standard of care of sensitive tuberculosis would help prevent multiresistance.
Acknowledgements
The authors express our sincere gratitude to the WAFETP program and the institutions that supported us in carrying out this work: Ministry of Health of Senegal, Center for Disease Control and Prevention, West African Health Organization.
Funding Statement
Funding: none.
References
- 1.World Health Organization. Global tuberculosis report 2018. 2018. Available from: https://www.who.int/tb/publications/global_report/en/. [Google Scholar]
- 2.World Health Organization, editor. Antimicrobial resistance: global report on surveillance. Geneva, Switzerland, 2014. Availablefrom: https://apps.who.int/iris/bitstream/handle/10665/112642/9789241564748_eng.pdf [Google Scholar]
- 3.Yimer SA, Agonafir M, Derese Y, et al. Primary drug resistance to anti-tuberculosis drugs in major towns of Amhara region, Ethiopia. APMIS 2012;120: 503–9. [DOI] [PubMed] [Google Scholar]
- 4.Abate D, Taye B, Abseno M, Biadgilign S. Epidemiology of anti-tuberculosis drug resistance patterns and trends in tuberculosis referral hospital in Addis Ababa, Ethiopia. BMC Res Notes 2012;5:462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sharma P, Lalwani J, Pandey P, Thakur A. Factors associated with the development of secondary multidrug-resistant tuberculosis. Int J Prev Med 2019;10: 67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Alene KA, Viney K, McBryde ES, et al. Risk factors for multidrug-resistant tuberculosis in northwest Ethiopia: A case-control study. Transbound Emerg Dis 2019;66:1611-8 [DOI] [PubMed] [Google Scholar]
- 7.Faustini A, Hall AJ, Perucci CA. Risk factors for multidrug resistant tuberculosis in Europe: a systematic review. Thorax 2006;61:158–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Mulu W, Mekonnen D, Yimer M, et al. Risk factors for multidrug resistant tuberculosis patients in Amhara National Regional State. Afr Health Sci 2015;15:368–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mulisa G, Workneh T, Hordofa N, et al. Multidrug-resistant Mycobacterium tuberculosis and associated risk factors in Oromia Region of Ethiopia. Int J Infect Dis 2015;39:57-61. [DOI] [PubMed] [Google Scholar]
- 10.Misombo-Kalabela A, Nguefack-Tsague G, Kalla GCM, et al. [Risk factors for multidrug-resistant tuberculosis in the city of Kinshasa in the Democratic Republic of Congo]. Pan Afr Med J. 2016;23:157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mdivani N, Zangaladze E, Volkova N, et al. High prevalence of multidrugresistant tuberculosis in Georgia. Int J Infect Dis 2008;12:635-44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mejri I, Saad SB, Daghfous H, et al. Facteurs de risque de tuberculose multirésistante. Revue des Maladies Respiratoires 2015;32:A233–4. [Google Scholar]
- 13.Robert J. Résistance de Mycobacterium tuberculosis aux antituberculeux en France. 2013;42 Available from: http://www.infectiologie.com/UserFiles/File/medias/JNI/JNI13/2013-JNI-BKR-franceRobert.pdf [Google Scholar]
- 14.Caminero JA. Multidrug-resistant tuberculosis: epidemiology, risk factors and case finding. Int J Tuberc Lung Dis 2010;14:382-90. [PubMed] [Google Scholar]
- 15.Agence de la santé publique du Canada. Normes canadiennes pour la lutte antituberculeuse 7e édition [Internet]. aem. 2014. Available from: https://www.canada.ca/fr/sante-publique/services/maladies-infectieuses/normescanadiennes-lutte-antituberculeuse-7eedition.html [Google Scholar]
- 16.Turkova A, Tebruegge M, Brinkmann F, et al. Management of child MDR-TB contacts across countries in the WHO European Region: a survey of current practice. Int J Tuberc Lung Dis 2017;21: 774-7. [DOI] [PubMed] [Google Scholar]
- 17.Ahmad AM, Akhtar S, Hasan R, et al. Risk factors for multidrug-resistant tuberculosis in urban Pakistan: A multicenter case-control study. Int J Mycobacteriol. 2012;1:137–42. [DOI] [PubMed] [Google Scholar]