Risk factors affecting treatment outcomes for pulmonary tuberculosis in Finland 2007-2014: a national cohort study.

Background: Major transition in tuberculosis (TB) epidemiology is taking place in many European countries including Finland. Monitoring treatment outcome of TB cases is important for identifying gaps in the national TB control program, in order to strengthen the system. The aim of the study was to identify potential risk factors for non-successful TB treatment outcomes, with a particular focus on the impact of comorbidities. We also evaluated the treatment outcome monitoring system.

Methods: All notified microbiologically confirmed pulmonary TB cases in Finland in 2007-2014 were included, except multi-drug resistant (MDR) cases. Nationwide register data were retrieved from: Infectious Diseases Register, Population Register, Cause of Death Register and Hospital Discharge Register. Non-successful outcomes were divided into three groups: death, unsatisfactory outcomes and non-defined outcomes. Logistic regression analyses were used to identify risk factors for non-successful outcomes.

Results: Treatment outcomes were notified for 98.6% of study cases (n = 1396/1416). Treatment success rate was 75%. The main reason for non-successful outcome was death (16%), whereas outcomes failed and lost to follow-up were rare (1% together). In a multivariable model, risk factors for death as outcome were increasing age, male gender and Charlson comorbidity index ≥1, for unsatisfactory outcomes non-MDR drug resistance and TB registered in the first study period, and for non-defined outcomes non-MDR drug resistance. Among 50 cases with unsatisfactory outcomes, we observed false outcome allocations in eight (16%), and > 2% of the cases transferred to another country or disappeared before or during treatment.

Conclusions: With a high proportion of older population among tuberculosis cases, death is a common treatment outcome in Finland. Comorbidity is an important factor to be incorporated when interpreting and comparing outcome rates. There was a considerable inconsistency in outcome allocation in the monitoring system, which implies that there is need to review the guidelines and provide further training for outcome assessment.