Mortality and prognostic factors in patients with obesity-hypoventilation syndrome undergoing noninvasive ventilation

Abstract

Objectives: The incidence of obesity-hypoventilation syndrome (OHS) has greatly increased over time, but data on long-term outcome are limited. We investigated survival and prognostic factors in these patients undergoing noninvasive positive pressure ventilation (NPPV).

Design: Retrospective descriptive analysis of patients with OHS and NPPV up to 10 years.

Methods: Long-term mortality and predictors of survival were assessed. Additionally, we evaluated changes in lung function, blood gas and laboratory parameters 5.7 +/- 2.5 months after initiation of NPPV.

Results: 126 patients (BMI 44.6 +/- 7.8 kg m(-2); PaCO(2) 55.5 +/- 7.7 mmHg) were followed for 41.3 +/- 27.6 months. Upon follow-up, blood gases (daytime and nighttime), as well as pulmonary function improved, whilst haemoglobin and BMI decreased (P < 0.001 each). Adherence to NPPV was high (94.5% continuing NPPV 6.5 +/- 2.3 h day(-1)). All-cause mortality was 12.7%, with 1-, 2- and 5-year survival of 97.1%, 92.0% and 70.2%, respectively. In univariate analysis, patients with PaO(2) <50 mmHg, C-reactive protein > or = 5.1 mg L(-1), leucocytes > or = 7.8 x 10(3) microl(-1), or pH > or = 7.44 at baseline had poor prognosis (P < 0.05 each). In Cox multivariate analysis, PaO(2), pH and leucocytes were independent predictors of mortality. Reduction in nocturnal PaCO(2) by > or =23.0% and haemoglobin at follow-up was associated with improved survival (P < 0.05 each) whilst a decrease in pH was a predictor of increased mortality. In contrast, neither baseline BMI nor its change was linked to survival.

Conclusion: Gas exchange and lung function in OHS were improved after initiation of NPPV. Hypoxemia, high pH and elevated inflammation markers predicted poor survival. Overall, NPPV was well tolerated and survival was excellent when compared with data from historical matched controls.