Table 2.
Quantitative Studies of Association Between ADR and Treatment Adherence or Outcome in MDR-TB Patients
| Study Design | Setting | Target | Data Collection | Summary of Result | Reference |
|---|---|---|---|---|---|
| Cross sectional | Guizhou, China, January-June 2018 | MDR-TB Patients (n= 202) | Pill count, interview | ADR (OR: 2.82, 95% CI:1.41–5.61) were the top predictor of treatment interruption. | Wang et al, 201949 |
| Cohort retrospective | Yerevan, Armenia, November 2011 | DR-TB patients who initiated treatment from 2005 to 2011 (n= 381) | Ratio of treatment doses received under DOT divided by the number of doses prescribed. | Number of ADR (aRR 1.18, 95% CI 1.09–1.27) was associated with loss to follow up from treatment. | Sanchez-Padilla et al, 201446 |
| Retrospective cohort | Oromia region, Ethiopia, November 2012-December 2017 | DR-TB patients (n= 406) | Patient records | Drug adverse effect (aHR = 6.1; 95% CI: 2.5, 14.34) were independent predictors of loss to follow up. | Woldeyohannes et al, 202152 |
| Case control | Philippines, April-July 2014 | DR-TB patients, ≥ 18 years old, started treatment from July-December 2012 (lost to follow-up cases, n=91; control, n= 182) | Interview | Vomit was the only ADR that associated with loss to follow-up in multivariate analysis (OR 1.10 (1.01–1.21), p value 0.03) and reported as the primary reason for stopping treatment. | Tupasi et al, 201650 |
| Cross sectional | Kano and Kaduna States, Nigeria, January 2015-June 2016 | DS-TB and DR-TB patients, adult (n=390) | Tuberculosis Medication Adherence Questionnaire (TBMAQ) | 56.7% reported that side effects of anti TB drugs affected their treatment adherence. Antituberculosis drugs side effects were significant variable in bivariate analysis (p < 0.001), but it was insignificant predictor in multivariate analysis (p = 0.191). | Iweama et al, 202153 |
| Cross sectional | Rajkot, India | MDR-TB cases during 2010–2013 (n =125) | Registered data | Defaulter was associated with ADR (χ2 = 20.214, df = 5, p = 0.001). Treatment adherence was associated with no occurrence of ADR (χ2 = 18.614, df = 1, p = 0.000). | Dela et al, 201754 |
| Retrospective cohort | Metro Manila, Philippines | MDR-TB patients treated from 1999 to 2006 (n=583) | Patient records | Uncontrolled ADR was not significantly associated with defaulter (p = 0.35). | Gler et al, 201555 |
| Retrospective cohort | Uganda | MDR-TB patients treated from January 2012-May 2016 (n=227) | Registered data | ADR was not significantly associated with non-adherence (p value = 0.357) | Batte et al, 202156 |
| Retrospective cohort | Kerala, Delhi and West Bengal, India | MDR-TB in the three States from January 2009 to December 2011. | Treatment card and registered data | ADR has no significant association with unfavorable outcome (p value = 0.09) | Nair et al, 201757 |
Abbreviations: HRQoL, health-related quality of life; ADR, adverse drug reaction; MCS, mental component summary; PCS, physical component summary; aHR, adjusted hazard ratio; OR, odds ratio; CI, confidence interval; DR-TB, drug resistant tuberculosis; MDR-TB, multi-drug resistant tuberculosis.