Additional value of spirometry in risk assessment of acute respiratory disease in winter among patients with mild to moderate asthma or COPD

Abstract

Background:

Spirometry is frequently performed by Dutch general practitioners to diagnose and monitor asthma or COPD. However, its additional value to anamnestic data in predicting the occurrence of acute respiratory disease (ARD) is unknown. We developed a clinical prediction model, assessing the additional value of spirometry, for the risk of asthma/COPD exacerbation or pneumonia during an influenza epidemic.

Methods:

Data were used from a case-control study among patients with asthma or COPD aged 18-64 years in general practice. During the influenza epidemics of 1998-1999 and 1999-2000, data on demographics, prior health care use, medical consumption, type of lung disease and comorbidity were collected from 87 patients with an exacerbation or pneumonia and 363 controls matched on age en sex using computerized medical records. Additionally, FEV₁ and PEF were measured before and after inhalation of a β-agonist. Conditional multivariate logistic regression analysis and receiver-operator curve (ROC) analysis was used to assess the discriminative value of anamnestic data and spirometry.

Results:

Independent predictors of the occurrence of ARD were ‘prior exacerbation’ (OR 6.3 95% CI 3.3-12.3), ‘use of bronchodilators’ (OR 2.0, 95% CI 1.3-3.8), ‘treatment by pulmonologist’ (OR 2.0, 95% CI 1.2-3.4) and ‘COPD’ (OR 1.7, 95%CI 0.9-3.0). The fit of the model was good (Hosmer-Lemeshow goodness-of-fit-test p=0.36) and discrimination satisfactory (AUC =0.75). Both FEV₁ %predicted and PEF%predicted were significant predictors, but did not improve discriminative power of the model. Subgroup analyses showed no effect modification by age, type of lung disease, or risk classification on the additional discriminative value of spirometry.

Conclusion:

Using routinely available information, a reliable estimation of the risk of ARD in patients with mild to moderate lung disease aged 18-64 years during an influenza epidemic can be made. There is no additional value of lung function to history taking in this patient group.

Full Text

The Full Text of this article is available as a PDF (18.2 KB).