Adverse reactions to antituberculosis drugs in in-hospital patients: Severity and risk factors

Given Mycobacterium tuberculosis's characteristics, the treatment of tuberculosis (TB) infection is administered over a long period of time (for six months or more) with a combination of several drugs which could cause adverse reactions (AR). These can cause significant morbidity and compromise tuberculosis treatment regimens.

To determine the incidence and severity of and risk factors for major adverse reactions to antituberculosis drugs in in-hospital patients treated for active tuberculosis.

Retrospective analysis of clinical records of patients admitted to Pulido Valente Hospital (Pulmonology Unit III) with active TB treated with anti-tuberculosis agents April 1999 to July 2007. Adverse reactions resulting in modification or discontinuation of treatment or hospital admission were recorded. Patients' demographic characteristics and clinical data were used as independent variables. The relationship between independent variables and the frequency and severity of AR was studied using multivariate analysis using a logistic regression model. The data were analysed using the Student t test, one-way ANOVA and logistic regression. Statistical analysis was performed using the SPSS (Statistical Package for the Social Sciences) version 15.0.

We recorded 1400 in-hospital patients treated for active TB 1999 to 2007, of which 175 patients (12.5%), 118 male and 57 female, had at least one AR induced by antituberculosis agents, to a total of 192 events. Hepatotoxicity was the most prevalent AR (83/47.4%), followed by skin reactions (55/31.4%) and gastrointestinal intolerance (24/13.7%). In 76 patients (43.4%) AR caused prolonged hospital stay. Statistically significant differences (p < 0.001) were observed in the average hospital stay (58.4 days for patients with AR and 26 days for patients without AR). Isoniazid (62.2%) and rifampicin (51.9%) were the most frequently implicated drugs. It was possible to characterise the AR severity in 134 cases. In 106 cases (79%) AR resulted in discontinuation of the drug. The relationship between drug and AR was definitive in 23 cases (17%). Of the 13 patients (9.6%) who died, AR was directly implicated in the cause of death in six (4.4%). AR were associated with alcoholism (relative risk [RR] 3.0; 95% confidence interval [CI] 1.1-7.9) and CD4 levels < 350 cells/mm(3) (RR 2.6; CI 1.4-5). In the predictive model, hepatic reactions were associated with viral hepatitis B and/or C (RR 2.5, CI; 1.2-5.1) and that CD4 levels < 350 cells/mm(3) (RR5.5; CI 1.6-18.6).

Antituberculosis drugs are associated with a significant number of AR that can cause significant morbidity, prolonged hospital stay and even death. Our results show that alcoholism and levels of CD4 < 350 cells/mm(3) were significantly associated with a high risk of AR and hepatitis B and C and levels of CD4 < 350 cells/mm3 were also significantly associated with hepatotoxicity.