Non-tuberculous mycobacterial pulmonary diseases in France: an 8 years nationwide study

Abstract

Background: The objective of the study was to describe the epidemiology, management and cost of non-tuberculous mycobacteria pulmonary disease (NTM-PD) in France. Methods: A retrospective analysis was performed using the SNDS (“Système national des données de santé”) database over 2010–2017. Patients with NTM-PD were identified based on the ICD10 codes during hospitalizations and/or specific antibiotics treatment regimens. The study population was matched (age, sex and region) to a control group (1:3) without NTM-PD. Results: 5628 patients with NTM-PD (men: 52.9%, mean age = 60.9 years) were identified over the study period and 1433 (25.5%) were treated with antibiotics. The proportion of patients still receiving treatment at 6 and 12 months was 40% and 22%, respectively. The prevalence of NTM-PD was estimated at 5.92 per 100,000 inhabitants and the incidence rate of NTM-PD remained stable over time between 1.025/100,000 in 2010 and 1.096/100,000 in 2017. Patients with NTM-PD had more co-morbidities compared to controls: corticoids (57.3% vs. 33.8%), chronic lower respiratory disease (34.4% vs. 2.7%), other infectious pneumonia (24.4% vs. 1.4%), malnutrition (based on hospitalization with the ICD-10 code reported during a hospital stay as a main or secondary diagnosis) (22.0% vs. 2.0%), history of tuberculosis (14.1% vs. 0.1%), HIV (8.7% vs. 0.2%), lung cancer and lung graft (5.7% vs. 0.4%), cystic fibrosis (3.2% vs. 0.0%), gastro-esophageal reflux disease (2.9% vs. 0.9%) and bone marrow transplant (1.3% vs. 0.0%) (p < 0.0001). The mean Charlson comorbidity index score was 1.6 (vs. 0.2 for controls; p < 0.0001). NTM-PD was independently associated with an increased mortality rate with a hazard ratio of 2.8 (95% CI: 2.53; 3.11). Mortality was lower for patients treated with antibiotics compared to untreated patients (HR = 0.772 (95% CI [0.628; 0.949]). Annual total expenses the year following the infection in a societal perspective were € 24,083 (SD: 29,358) in NTM-PD subjects vs. € 3402 (SD: 8575) in controls (p < 0.0001). Main driver of the total expense for NTM-PD patients was hospital expense (> 50% of the total expense). Conclusion: Patients with NTM-PD in France were shown to have many comorbidities, their mortality risk is high and mainly driven by NTM-PD, and their management costly. Only a minority of patients got treated with antibiotics and of those patients treated, many stopped their therapy prematurely. These results underline the high burden associated with NTM-PD and the need for improvement of NTM-PD management in France.