Abstract

Résumé

Abstract

ASPECT OF INTEREST

Methods

We conducted a retrospective case series analysing patients commencing their first XDR-TB treatment episode in Peru during 2010–2012. In accordance with Peru's National TB Programme (NTP) recommendations, XDR-TB patients receive a baseline clinical assessment from a psychologist or psychiatrist to screen for mental health conditions using the International Classification of Diseases 10^th revision (ICD-10). Demographic and clinical data were abstracted from medical charts in the NTP's central archive onto paper forms during June–July 2015. The data were then single-entered and cleaned electronically using EpiData v3.1 (EpiData Association, Odense, Denmark), Excel 2011 (Microsoft Corp, Redmond, USA) and q (Ben-Attia Harel, Israel).

The prevalence of mental disorders—depression, anxiety, mixed anxiety-depressive disorder and psychosis—and alcohol and drug use was calculated. Using the Peruvian NTP definition for XDR-TB treatment outcomes (Table 1), we estimated the crude odds ratios (OR) for unfavourable (failure, loss to follow-up, death) outcomes for mental disorders and current drug/alcohol use using exact logistic regression using Stata 13.0 (StataCorp, College Station, TX, USA).

TABLE 1

Definitions of XDR-TB treatment outcomes, Peru National Tuberculosis Programme

Open in a separate window

The study was approved by the Human Ethics Research Committees at the Universidad Peruana Cayetano Heredia (Lima, Peru) and London School of Hygiene & Tropical Medicine (London, UK).

RESULTS

We abstracted treatment data for 53 XDR-TB patients who were mostly males aged 19–66 years living in metropolitan Lima/Callao. One patient was a secondary XDR-TB case, while the remainder had a median of 2.35 years of previous anti-tuberculosis treatment (interquartile range [IQR] 9.6 months–3.75 years). Of the 53 patients, 49 had documented psychiatric evaluations, of whom 11/49 were diagnosed with a mental disorder: five with depression, three with anxiety, two with mixed anxiety-depressive disorder and one with psychosis. Overall treatment outcomes were: 12/49 favourable (seven cured, five completed) and 37/49 unfavourable (11 lost to follow-up, 11 failures and 15 deaths). Among those diagnosed with mental disorders, 9/11 had an unfavourable treatment outcome. Three of the four patients without psychiatric assessments had unfavourable outcomes (Table 2).

TABLE 2

Prevalence of psychiatric conditions at baseline among 49 patients commencing XDR-TB treatment in Peru, 2010–2012

Open in a separate window

Regarding alcohol use, 8/49 patients were consuming alcohol at baseline, of whom two were diagnosed with alcoholism. A further 11/49 had a history of alcohol consumption. Six patients currently used drugs; five more had a history of drug use. Nearly half (22/49) had a current or prior history of alcohol/drug use; these patients tended to have unfavourable outcomes.

There was no statistical evidence that a diagnosis of mental disorder (OR 1.59, 95% confidence interval [CI] 0.26–7.6, P = 0.70) or current alcohol/drug use (OR 0.71, 95%CI 0.12–5.09, P = 0.69) were associated with increased odds of worse XDR-TB treatment outcomes.

DISCUSSION

To our knowledge, this retrospective series is the first to report on XDR-TB psychiatric comorbidity in Peru or elsewhere. The point prevalences of baseline depression and anxiety observed in this cohort of XDR-TB patients were greater than for the adult population of Lima/Callao (10.2% vs. 2.8% and 6.1% vs. 1.9%, respectively), while rates of psychosis and alcohol/drug use were comparable to the general population.⁷ Compared to the first 75 MDR-TB patients in Peru to receive individualised therapy (1996–1999), depression was less prevalent in the present study (10.2% vs. 52.2%), but the prevalence of anxiety was similar (6.1% vs. 8.7%).⁸ However, our results for all psychiatric conditions are comparable to a larger MDR-TB cohort in Peru (22.4% vs. 18.9%).² Nonetheless, we found no association of psychiatric conditions or drug/alcohol use with unfavourable XDR-TB outcomes.

Prospective studies and qualitative inquiry would shed further light on these unexpected findings. It may be that patients with pre-existing (prior to XDR-TB treatment) mental disorders were less likely to present for care, or had died before they could, and were therefore underrepresented in our sample. Alternatively, patients may have been too gravely ill to be evaluated with appropriate instruments, thereby underestimating the true prevalence of psychiatric disorders. Patients who were not using alcohol/drugs at baseline may have restarted once treatment improved their TB symptomatology. As it is unlikely that psychiatric disorders are protective factors during XDR-TB treatment, NTPs should assume that mental health interventions would increase patient wellbeing and reduce loss to follow-up. Initiatives such as the WHO End TB strategy⁹ and Zero TB,¹⁰ which include mental health support as core components, offer promise; however, more data are needed on specific comorbid mental disorders with XDR-TB to develop the most successful interventions.

Due to the small sample size and missing data, such as patients' socio-economic status and whether diagnoses were made by psychiatrists or psychologists, key confounders cannot be ruled out. As the mental health evaluations were performed at treatment initiation, we could not assess the incidence of mental disorders.

CONCLUSION

Baseline psychiatric disorders and drug/alcohol use were not associated with poor XDR-TB outcomes; however, overall treatment success was low. Appropriately powered cohort studies are needed to better understand mental health issues affecting XDR-TB patients throughout treatment and develop interventions that improve patients' survival and quality of life.

Acknowledgments

Footnotes

References