Validity of COPD diagnoses reported through nationwide health insurance systems in the People’s Republic of China

Figures & data

Table 1 Baseline characteristics of participants with incident COPD

Baseline characteristics	COPD cases retrieved for adjudicationa	COPD cases not adjudicated	All COPD cases
N	1,069	10,730	11,799
Age, years (mean ± SD)	62.9±9.2	62.4±13.9	62.4±12.8
Female (%)	46.3	50.2	50.7
Urban (%)	37.0	13.5	15.1
Follow-up time, years (mean ± SD)	3.7±1.7	4.7±6.3	4.6±4.9
Highest education completed (%)
None/primary school	61.6	58.8	57.6
Middle/high school	34.7	36.9	38.4
College/university	3.7	4.2	4.0
Annual household income (%)
<10,000 (Yuan)	11.6	12.2	11.8
10,000–34,000 (Yuan)	78.4	72.6	73.9
≥35,000+ (Yuan)	10.0	15.1	14.3
Smoking status (%)
Current regular	25.2	27.7	26.9
Ex-regular	10.9	7.8	7.9
Never regular	63.9	64.6	65.1
Height, cm (mean ± SD)	156.3±6.6	156.2±9.5	156.2±8.7
BMI, kg/m^2 (mean ± SD)	22.7±4.3	23.1±9.2	23.1±8.2
Self-reported health status (%)
Poor	32.1	21.4	23.0
Fair	39.7	46.2	45.5
Good	17.4	21.1	19.9
Excellent	10.7	11.3	11.5
Lung function (mean ± SD)
FEV1 (L)	1.46±0.65	1.81±1.47	1.80±1.47
FVC (L)	1.92±0.75	2.25±1.57	2.25±1.49
FEV1/FVC (%)	74.8±14.5	78.9±24.5	78.6±24.4
COPD severityb (%)
Grade 1–2	7.3	5.0	6.1
Grade 3	3.1	2.4	2.5
Grade 4	7.9	6.9	6.4
Grade 5	21.4	10.8	11.9
Self-reported comorbidity (%)
Hypertension	12.0	11.9	11.5
IHD	6.2	4.0	4.1
Diabetes	2.9	2.7	2.7
Stroke/TIA	1.4	1.7	1.7

Notes:

a Standardized for age, sex, and regions;

b data from baseline lung function (2004–2008): COPD Grade 1–2= FEV₁/FVC <LLN and (z-score of FEV₁ −2.5 to 1.0); Grade 3= FEV₁/FVC <LLN and (z-score of FEV₁ −3.0 to −2.5); Grade 4= FEV₁/FVC <LLN and (z-score of FEV₁ −3.5 to −3.0); Grade 5= FEV₁/FVC <LLN and (z-score of FEV₁ <−3.5); and P-values between two groups for all characteristics are <0.05. Eventually, medical records of 1,069 participants were retrieved for adjudication.

Abbreviations: SD, standard deviation; BMI, body mass index; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; IHD, ischemic heart diseases; TIA, transient ischemic attack; LLN, lower limit of normal.

Figure 1 Flow diagram summarizing the process of COPD adjudication.

Note: *Some misdiagnosed cases had multiple diseases.

Table 2 Distribution of refuted and confirmed COPD cases by regions, hospital ranking, reporting sources, sex, and age

	Number	Refuted cases n (%)	Confirmed COPD cases n (%)	aPPV % (95% CI)
Total (N)	1,069	71 (6.6)	911 (85.2)	85 (83–87)
Reporting source
HI only	961	71 (7.4)	806 (83.9)	84 (81–87)
Death registry and HI	108	0 (0.0)	105 (97.2)	97 (94–100)
Hospital tierb
Top rank (Tier 3)	365	25 (6.8)	303 (83.0)	83 (78–86)
Medium rank (Tier 2)	290	9 (3.1)	259 (89.3)	89 (85–93)
Low rank (Tier 1)	349	27 (7.7)	299 (85.7)	86 (83–91)
Admission year
2004–2007	104	6 (5.8)	92 (88.5)	88 (82–95)
2008	187	8 (4.3)	168 (89.8)	90 (85–94)
2009	228	15 (6.6)	191 (83.8)	84 (78–89)
2010	231	19 (8.2)	190 (82.3)	82 (77–88)
2011–2013	319	23 (7.2)	270 (84.6)	85 (80–89)
Sex
Male	611	23 (3.8)	547 (89.5)	89 (87–92)
Female	458	48 (10.5)	364 (79.5)	79 (75–84)
Age groups (in years)
<60	210	23 (11.0)	166 (79.0)	79 (73–85)
60–69	397	22 (5.5)	339 (85.4)	85 (82–89)
70+	462	26 (5.6)	406 (87.9)	88 (85–91)
Regions
Urban	531	35 (6.6)	434 (81.7)	82 (78–85)
Qingdao	120	3 (2.5)	105 (87.5)	87 (81–94)
Harbin	123	3 (2.4)	100 (81.3)	81 (74–89)
Haikou	40	3 (7.5)	37 (92.5)	92 (84–100)
Suzhou	125	8 (6.4)	106 (84.8)	85 (78–92)
Liuzhou	123	18 (14.6)	86 (69.9)	70 (60–80)
Rural	538	36 (6.7)	477 (88.7)	89 (86–91)
Sichuan	117	4 (3.4)	110 (94.0)	94 (90–98)
Gansu	79	13 (16.5)	58 (73.4)	73 (62–85)
Henan	113	15 (13.3)	93 (82.3)	82 (74–90)
Zhejiang	115	3 (2.5)	108 (93.9)	94 (89–98)
Hunan	114	1 (0.9)	108 (94.7)	95 (91–99)

Notes:

a Positive predictive value (subjects with confirmed COPD cases/total subjects selected for COPD adjudication) rounded up to no decimal place;

b hospital tier not available for 65 cases. Out of total 1,069 cases, 71 were refuted cases with no respiratory diagnoses and 87 cases were misdiagnosed as CODP although they were other respiratory diseases.

Abbreviations: PPV, positive predictive value; HI, health insurance; CI, confidence interval.

Figure S1 The location of ten survey sites in China Kadoorie Biobank (CKB).

Figure S2 Venn diagram showing the breakdown of sources for total COPD outcomes in China Kadoorie Biobank (CKB).

Figure S3 China Kadoorie Biobank (CKB) disease validation form for chronic respiratory diseases (CRD).

Abbreviations: OSAS, obstructive sleep apnea syndrome; IHD, ischemic heart diseases; FEV₁/FVC, forced expiratory volume in 1 second/forced vital capacity; DLCO, diffusing capacity of the lungs for carbon monoxide; CT, computed tomography; IgE, immunoglobulin E; PHD, Pulmonary Heart Disease; ID, identification.