Hyperglycaemia as a predictor of outcome during Non Invasive Ventilation in decompensated COPD

Biswajit Chakrabarti ^1*, Robert M Angus ², Sanjeev Agarwal Dr², Steven Lane ³ and Peter Calverley ¹

¹ Clinical Sciences Centre, University Hospital Aintree, University of Liverpool, Liverpool, United Kingdom
² Aintree Chest Centre, University Hospital Aintree, Liverpool, United Kingdom
³ Centre for Medical Statistics and Health Evaluation, University of Liverpool, Liverpool, United Kingdom

^* To whom correspondence should be addressed. E-mail: biz{at}doctors.org.uk.

Accepted 29 April 2009

Abstract

Rationale: Hyperglycaemia predicts a poor outcome in ICU patients. Whether this is true for respiratory failure necessitating non-invasive ventilation (NIV) is not known.

Objectives: To determine whether hyperglycaemia within 24 hours of admission independently predicts outcome of NIV during acute decompensated ventilatory failure complicating COPD exacerbations.

Methods: COPD patients presenting with acute hypercapnic respiratory failure at University Hospital Aintree between June 2006 and September 2007 and receiving NIV within 24 hours of admission were prospectively studied. Random blood glucose levels were measured before NIV administration.

Measurements and main results: 88 patients (mean baseline pH 7.25, PaCO2 10.20 kPa, and PaO2 8.19 kPa) met inclusion criteria with NIV normalising arterial pH off therapy in 79 (90%). After multi-variate logistic regression, the following predicted outcome: baseline Respiratory Rate (OR 0.91; 95% CI 0.84-0.99), random glucose 7 mmol/l (OR 0.07; 95% CI 0.007-0.63) and admission APACHE II score (OR 0.75; 95% CI 0.62-0.90). The combination of baseline RR < 30 breaths per minute and random glucose < 7mmol/l increased prediction of NIV success to 97% whilst use of all 3 factors was 100% predictive.

Conclusions: In acute decompensated ventilatory failure complicating COPD, hyperglycaemia upon presentation was associated with a poor outcome. Baseline respiratory rate and hyperglycaemia are as good at predicting clinical outcomes as the APACHE 2 score. Combining these variables increases predictive accuracy providing a simple method of early risk stratification.