Risk factors that may be driving the emergence of drug resistance in tuberculosis patients treated in Yangon, Myanmar

Survival analysis of patients with tuberculosis and risk factors for multidrug-resistant tuberculosis in Monrovia, Liberia

We reviewed the records of 337 confirmed cases of tuberculosis patients in Monrovia, the capital of Liberia, 2015. The risk factors affecting the survival and multidrug-resistance of tuberculosis patients were examined. Kaplan-Meier analysis and the log-rank test were used to assess the differences in survival among the patients, while Cox regression model was used for multivariate analysis. The qualitative data was tested with chi-square test in the single factor analysis of multidrug-resistant TB. Multivariate analysis was performed using binary logistic regression analysis. The significance level for all the tests were set at 0.05. The mean period of the follow-up of patients was 10 months. In the 337 patients, 33 (9.8%) died, the 21-month survival rate was 90.2%. The results of multivariate Cox regression analysis show that overcrowding (HR = 7.942, 95% CI 3.258-19.356), former smoking (HR = 3.773, 95% CI 1.601-8.889), current smoking (HR = 3.546, 95% CI 1.195-10.521), multidrug-resistance tuberculosis (HR = 4.632, 95% CI 1.913-11.217) were risk factors for death during anti-tuberculosis treatment in TB patients in Liberia. The results of binary logistic regression analysis show that extra-pulmonary (OR = 2.032, 95% CI 1.133-3.644), family history of TB (OR = 2.387, 95% CI 1.186-4.807) and current smoking (OR = 3.436, 95% CI 1.681-7.027) were risk factors for multidrug-resistant tuberculosis. These results can provide insights on local tuberculosis early intervention, increase public health awareness, and strengthen the control of factors that may affect the survival and multidrug-resistance of tuberculosis patients.

Determining Mycobacterium tuberculosis Drug Resistance and Risk Factors for Multidrug-Resistant Tuberculosis in Sputum Smear-Positive Tuberculosis Outpatients in Anhui Province, China, 2015-2016

BACKGROUND

Multidrug-resistant tuberculosis (MDR-TB) is currently a major problem in China. The prevention and treatment work for MDR-TB patients started late in Anhui province. To determine the prevalence of MDR-TB in sputum smear-positive TB patients (SSPTBPs) and analyze the risk factors for MDR-TB in Anhui province, we conducted an investigation of drug resistance among SSPTB outpatients from September 2015 to August 2016.

METHODS

A stratified cluster-randomized sampling method was used to obtain a representative sample. It was estimated that 2290 new cases and 440 previously treated cases of SSPTBPs needed to be recruited from 40 survey sites. Isolates were tested for resistance to six first-line and second-line anti-TB drugs. Information from patient questionnaire survey was used to identify factors linked to MDR-TB.

RESULTS

Finally, a total of 3047 SSPTBPs were recruited from 40 survey sites; of these, 2530 specimens were successfully cultured and had drug susceptibility testing done. The proportions of rifampin resistant (RR)-TB were 11.42% (289/2530, 95% CI: 10.18-12.66%), 7.64% (163/2133, 95% CI: 6.38-8.62%) and 31.74% (126/397, 95% CI: 27.38-36.60%) in all cases, new cases and previously treated cases, respectively, and the proportions of confirmed MDR-TB were 7.63% (193/2530, 95% CI: 6.59-8.66%), 4.97% (106/2133, 95% CI: 4.05-5.89%) and 21.91% (87/397, 95% CI: 17.83-26.00%), respectively. The ofloxacin resistance rate in previously treated SSPTBPs reached 21.66% (95% CI: 17.33-26.75%). Patients who had received two or more anti-TB treatment courses were significantly associated with MDR-TB compared to patients who have received one anti-TB course.

CONCLUSION

MDR-TB prevalence was high among SSPTBPs in Anhui province, and past anti-TB treatment course was associated with MDR-TB.

Invisible medicine sellers and their use of antibiotics: a qualitative study in Cambodia

BACKGROUND

Global attention to antimicrobial resistance has increased interest in tackling the widespread inappropriate dispensing of antibiotics by informal, for-profit healthcare providers (HCPs). We provide new evidence on an understudied group of informal HCPs: invisible medicine sellers (IMS) who operate without any marked facility. We investigated factors that influence community decisions on which HCPs to purchase medicines from, focusing on reasons for using IMS, and compared different HCPs' knowledge of antibiotic use.

METHODS

We conducted community focus group discussions (FGDs) in seven purposively selected villages representing high and low informal HCPs use in two peri-urban districts in Phnom Penh, Cambodia. Using information from the FGDs to identify HCPs that sell medicines, we interviewed 35 participants: 21 HCPs (including five IMS) and 14 key informants, including government HCPs and village leaders. We adopted an interpretative approach and conducted a thematic analysis.

RESULTS

Community members typically knew of several formal and informal HCPs selling medicines nearby, and IMS were common, as were doctors that sell medicines covertly. Two factors were most salient in influencing the choice of HCP for medicine purchasing. The first was trust in the effectiveness of medicines provided, judged by the speed of symptomatic relief. This pushed HCPs to provide several medicines, including antibiotics, at the first consultation. The second was the convenience offered by IMS and other informal HCPs: supplying medicines when other facilities are closed, accepting delayed payments, providing incomplete courses of medication and selling human antibiotics for animal use.

CONCLUSION

This first study focusing on IMS indicates that it is important, but challenging, for public health agencies to engage with them to reduce inappropriate use of antibiotics. Although public health facilities must fill some gaps that informal HCPs are currently addressing, such as access to medicines at night, reducing demand for unnecessary antibiotics is also critical.

Feature selection and prediction of treatment failure in tuberculosis

Background

Tuberculosis is a major cause of morbidity and mortality in the developing world. Drug resistance, which is predicted to rise in many countries worldwide, threatens tuberculosis treatment and control.

Objective

To identify features associated with treatment failure and to predict which patients are at highest risk of treatment failure.

Methods

On a multi-country dataset managed by the National Institute of Allergy and Infectious Diseases we applied various machine learning techniques to identify factors statistically associated with treatment failure and to predict treatment failure based on baseline demographic and clinical characteristics alone.

Results

The complete-case analysis database consisted of 587 patients (68% males) with a median (p25-p75) age of 40 (30–51) years. Treatment failure occurred in approximately one fourth of the patients. The features most associated with treatment failure were patterns of drug sensitivity, imaging findings, findings in the microscopy Ziehl-Nielsen stain, education status, and employment status. The most predictive model was forward stepwise selection (AUC: 0.74), although most models performed at or above AUC 0.7. A sensitivity analysis using the 643 original patients filling the missing values with multiple imputation showed similar predictive features and generally increased predictive performance.

Conclusion

Machine learning can help to identify patients at higher risk of treatment failure. Closer monitoring of these patients may decrease treatment failure rates and prevent emergence of antibiotic resistance. The use of inexpensive basic demographic and clinical features makes this approach attractive in low and middle-income countries.