Performance of COPD population screener questionnaire in COPD screening: a validation study and meta-analysis

Abstract

This study aimed to validate the chronic obstructive pulmonary disease (COPD) Population Screener (COPD-PS) questionnaire as a screening tool in a cohort of Chinese subjects who underwent a health examination, and to summarise its overall performance through a meta-analysis. We enrolled 997 subjects aged ≥40 years who underwent a health examination, both lung function and COPD-PS data were collected. The screening performance of COPD-PS was evaluated with a receiver operating characteristic (ROC) curve analysis, using the area under the curve (AUC) to assess the screening accuracy. A standard diagnostic meta-analysis was used to summarise the screening performance of COPD-PS for COPD. Of the 997 subjects, 157 were identified as having COPD. The COPD-PS score was significantly higher in COPD patients than controls (5.03 ± 5.11 vs. 2.72 ± 1.80, p < .001). At a cut-off of 4, the sensitivity and specificity of COPD-PS for identifying COPD were 74.52 and 70.24%, respectively, with an AUC of 0.79. Eight studies (including this study) were included in this meta-analysis. The pooled estimates for COPD-PS were as follows: sensitivity of 0.66 (95% CI: 0.47-0.63), specificity of 0.86 (95% CI: 0.84-0.89), positive likelihood ratio of 3.00 (95% CI: 1.65-5.47), negative likelihood ratio of 0.43 (95% CI: 0.35-0.52) and diagnostic odds ratio of 7.24 (95% CI: 3.91-13.40). The AUC of the summary ROC curve was 0.78. COPD-PS appears to be a useful tool for screening individuals with a high risk of COPD and guiding the selection of individuals for subsequent spirometry examination.KEY MESSAGESCOPD-PS is a simple and useful method to screen COPD.The combination of COPD-PS with other tools may improve the screen performance.

Keywords: chronic obstructive pulmonary disease population screener; Chronic obstructive pulmonary disease; meta-analysis; screening.

Figure 1.

Receiver operating characteristic (ROC) curve of the use of COPD-PS to discriminate between COPD patients and controls. COPD-PS: chronic obstructive pulmonary disease population screener.

Figure 2.

Quality assessment of studies on COPD-PS. COPD-PS: chronic obstructive pulmonary disease population screener.

Figure 3.

Forest plot of sensitivity COPD-PS with random-effects model. The point estimates of sensitivity from each study are shown as solid circles. Error bars indicate 95% CIs. COPD-PS: chronic obstructive pulmonary disease population screener.

Figure 4.

Forest plot of specificity of COPD-PS with random-effects model. The point estimates of specificity from each study are shown as solid circles. Error bars indicate 95% CIs. COPD-PS: chronic obstructive pulmonary disease population screener.

Figure 5.

Forest plot of diagnostic odds ratio of COPD-PS with random effects model. The point estimates of specificity from each study are shown as solid circles. Error bars indicate 95% CIs. COPD-PS: chronic obstructive pulmonary disease population screener.

Figure 6.

Summary receiver operating characteristic (SROC) curve of COPD-PS as a screening tool for COPD. COPD-PS: chronic obstructive pulmonary disease population screener. Q: The maximum joint value of sensitivity and specificity of COPD-PS.

Figure 7.

Deeks’ funnel plot assessing the likelihood of publication bias. The statistically nonsignificant p-value of 0.37 for the slope coefficient suggests symmetry in the data and a low likelihood of publication bias. ESS: effective sample size, corresponding to 4 × Ncontrol × N_COPD/(Ncontrol + N_COPD).