Validity of the Handheld Expiratory Flowmeter for COPD Screening in the Primary Care Setting of China

Figures & data

Table 1 Characteristics of All Participants

	Non-COPD (N=1258)	COPD (N=229)
Area
Urban	824	121
Rural	434	108
Sex
Men	689	188
Women	569	41
Age (years)
40–49	215	7
50–59	568	73
60–69	372	92
≥70	103	57
BMI (kg/m^2)
<18.5	84	32
18.5–23.9	653	129
24.0–27.9	403	57
≥28.0	118	11
Smoking status
Never-smoker	775	70
Smoker	482	159
Respiratory symptoms
With respiratory symptom	251	111
Without respiratory symptom	1006	118
Lung function
FEV1/FVC >0.7	1258	–
GOLD stage I	–	105
GOLD stage II	–	91
GOLD stage III	–	28
GOLD stage IV	–	5

Figure 1 Correlation of FEV₁ measured by the conventional spirometry with FEV₁ measured by the handheld expiratory flowmeter. (A) Relationship between FEV₁ measured by spirometry and FEV₁ measured by the handheld expiratory flowmeter in total group (r₁=0.889, P<0.001), non-COPD group (r₂=0.869, P<0.001) and COPD group (r₃=0.907, P<0.001). (B) Relationship between FEV₁ measured by spirometry and the handheld expiratory flowmeter in the groups of GOLD stage I (r_I=0.810, P<0.001), stage II (r_II=0.802, P<0.001), stage III (r_III=0.637, P<0.001) and stage IV (r_IV=0.844,P<0.001). (C) Bland–Altman graph of FEV₁ measured by spirometry and the handheld expiratory flowmeter. 4.5% (67/1487) plots were out of the 95%LoA(−0.445to 0.816L).

Figure 2 Correlation of FVC measured by the conventional spirometry with FEV₆ measured by the handheld expiratory flowmeter. (A) Relationship between FVC measured by spirometry and FEV₆ measured by the handheld expiratory flowmeter in total group (r₁=0.835, P<0.001), non-COPD group (r₂=0.865, P<0.001) and COPD group (r₃=0.807, P<0.001). (B) Relationship between FVC measured by spirometry and FEV₆ measured by the handheld expiratory flowmeter in groups of GOLD stage I (r_I=0.737, P<0.001), stage II (r_II=0.724, P<0.001), stage III (r_III=0.574, P=0.0014) and stage IV (r_IV=0.615, P=0.269). (C) Bland–Altman graph of FVC by spirometry and FEV₆ by the handheld expiratory flowmeter. 5.2% (77/1487) plots were out of the 95%LoA(0.514–1.297L).

Figure 3 Correlation of FEV1/FVC measured by the conventional spirometry with FEV1/FEV₆ measured by the handheld expiratory flowmeter. (A) Relationship between FEV1/FVC measured by spirometry and FEV1/FEV₆ measured by the handheld expiratory flowmeter in total group (r₁=0.647, P<0.001), non-COPD group (r₂=0.343, P<0.001) and COPD group (r₃=0.686, P<0.001). (B) Relationship between FEV1/FVC measured by spirometry and FEV₆/FVC measured by the handheld expiratory flowmeter in groups of GOLD stage I (r_I=0.197, P<0.044), stage II (r_II=0.641, P<0.001), stage III (r_III=0.715, P<0.001) and stage IV (r_IV=0.784, P=0.117). (C) Bland–Altman graph of FVC by spirometry and FEV₆ by the handheld expiratory flowmeter.

Table 2 The Measures on Sensitivity and Specificity at Different Cut-Off Points of FEV₁/FEV₆

Cut-Off Point	SE%	SP%	PPV%	NPV%	LR+
≤0.65	25.8	98.8	80.8	88.0	23.2
≤0.70	41.9	97.9	78.7	90.3	20.3
≤0.75	60.3	93.0	60.5	92.8	8.4
≤0.77	71.2	89.8	55.8	94.5	6.9
≤0.80	77.7	80.0	41.3	95.2	3.9

Abbreviations: SE, sensitivity; SP, specificity; PPV, positive predictive value; NPV, negative predictive value; LR+, positive likelihood ratio.

Figure 4 Area under ROC curves for FEV₁/FEV₆ measured by the handheld expiratory flowmeter. (A) FEV₁/FEV₆ by the handheld expiratory flowmeter and DIF-Age(“lung age” by the handheld expiratory flowmeter–actual age) to identify airflow obstruction.(B) FEV₁/FEV₆ by the handheld expiratory flowmeter to identify airflow obstruction in the population with respiratory symptoms and the population without respiratory symptom. (C) FEV₁/FEV₆ by the handheld expiratory flowmeter to identify airflow obstruction in the smokers (including current smokers and ex-smokers) and never-smokers. Using post-bronchodilators FEV₁/FVC <70% as a “gold standard” for determination of airflow obstruction.