Clinical Impact and Healthcare Resource Utilization Associated with Early versus Late COPD Diagnosis in Patients from UK CPRD Database

Figures & data

Figure 1 Study design. *Index date was the time of the first recorded physician’s diagnosis of COPD during the identification time frame.

Abbreviation: COPD, chronic obstructive pulmonary disease.

Table 1 Baseline Characteristics of Early- versus Late-Diagnosed COPD Patients

Characteristics	Early Diagnosis of COPD (N = 3375)	Late Diagnosis of COPD (N = 6783)	P-value
Age
Mean ± SD	67.28 ± 10.21	69.32 ± 10.55	<0.0001
Age, ≥40 to <65 years	1297 (38.4%)	2160 (31.8%)
Age, ≥65 years	2078 (61.6%)	4623 (68.2%)
Gender, Male	2132 (63.2%)	3516 (51.8%)	<0.0001
BMI, mean ± SD	27.01 ± 6.16	27.5 ± 6.16	0.0009
Smoking status
Never smoker	137 (4.1%)	360 (5.3%)	0.0078
Ex-smoker	1637 (48.5%)	3752 (55.3%)
Current smoker	1593 (47.2%)	2649 (39.1%)
Missing	8 (0.2%)	22 (0.3%)
mMRC score
<2	1823 (54.0%)	3280 (48.4%)	<0.0001
≥2	897 (26.6%)	2514 (37.1%)
Missing	655 (19.4%)	989 (14.6%)
Lung function
FEV1 (L)^a, mean ± SD	1.88 ± 0.78	1.77 ± 0.72	<0.0001
Percentage predicted FEV1^b, mean ± SD	61.07 ± 18.76	62.78 ± 19.38	0.0011
FEV1/FVC (%)^c, mean ± SD	58.04 ± 16.94	59.40 ± 17.72	0.0014
FEV1/FVC ≥ 70% predicted	483 (19.5%)	1164 (24.4%)	<0.0001
FEV1/FVC < 70% predicted	1999 (80.5%)	3614 (75.6%)
Comorbidities in the pre-index period^d
Hypertension	1759 (52.1%)	3919 (57.8%)	<0.0001
Depression	878 (26.1%)	2197 (32.4%)	<0.0001
Anxiety	730 (21.6%)	1985 (29.3%)	<0.0001
Hyperlipidemia	751 (22.3%)	1757 (25.9%)	<0.0001
Diabetes	344 (10.2%)	931 (13.7%)	<0.0001
Pneumonia	188 (5.6%)	903 (13.3%)	<0.0001
Atrial fibrillation or other cardiac arrhythmias	280 (8.3%)	856 (12.6%)	<0.0001
Acute myocardial infarction	238 (7.1%)	642 (9.5%)	<0.0001
Stroke	208 (6.2%)	568 (8.4%)	<0.0001
Osteoporosis	142 (4.2%)	497 (7.3%)	<0.0001
Heart failure	109 (3.2%)	395 (5.8%)	<0.0001
Lung cancer	13 (0.4%)	95 (1.4%)	<0.0001
CCI score in the pre-index period, mean ± SD	2.19 ± 1.6	2.7 ± 1.91	<0.0001
Eosinophils (cells/µL)
<150	717 (33.57%)	1461 (32.68%)	0.0313
≥150 and <300	725 (33.94%)	1575 (35.23%)
>300	694 (32.49%)	1434 (32.08%)

Notes: Data presented as n (%) unless otherwise stated. ^aEarly (n = 2693) versus late (n = 5337) diagnosed patients. ^bEarly (n = 1966) versus late (n = 3909) diagnosed patients. ^cEarly (n = 2482) versus late (n = 4778) diagnosed patients. ^dPrior to the date of the first COPD diagnosis during the identification period.

Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index; COPD, chronic obstructive pulmonary disease; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; mMRC, modified Medical Research Council; SD, standard deviation.

Table 2 Hazard Ratio of First Exacerbation for Early versus Late COPD Diagnosis

Variables	Adjusted^a HR (95% CI)
Late versus early diagnosis of COPD	1.46 (1.38–1.55); P < 0.0001
Age, ≥ 65 years versus ≥ 40 to < 65 years	0.92 (0.87–0.98)
CCI score 0 versus ≥ 1 to ≤ 2	1.00 (0.68–1.46)
CCI score 0 versus > 2	1.09 (0.74–1.59)
Underweight versus normal weight^b	1.13 (0.99–1.28)
Overweight versus normal weight^b	0.98 (0.91–1.05)
Obese versus normal weight^b	0.90 (0.83–0.96)
Ex-smoker versus never smoker	1.02 (0.90–1.15)
Current smoker versus never smoker	1.13 (1.00–1.28)
GOLD B versus GOLD A	1.23 (1.11–1.37)
GOLD C versus GOLD A	1.54 (1.41–1.67)
GOLD D versus GOLD A	1.83 (1.68–1.99)
Adherence^c to COPD medication versus non-adherence	0.07 (0.07–0.08)
Patients with depression versus without depression	1.18 (1.09–1.27)
Patients with pneumonia versus without pneumonia	1.12 (1.06–1.18)
Patients with osteoporosis versus without osteoporosis	1.10 (1.00–1.22)

Notes: P-value from the Wald test obtained from an adjusted Cox model. ^aAdjusted for depression and the “a priori” variables, namely age, CCI, BMI, smoking status, GOLD severity, and adherence to COPD treatment. ^bWeight was measured based on BMI.

^cCalculated based on MPR. The number of patients with good adherence versus poor adherence was very low, especially in patients not having an exacerbation. This may be attributed to the difficulties around recording of medication in the CPRD database.

Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; CPRD, Clinical Practice Research Datalink; GOLD, Global Initiative for Chronic Obstructive Lung Disease; HR, hazard ratio; MPR, medicine possession ratio.

Figure 2 Time-to-first exacerbation in the early and late-diagnosed COPD patients.

Abbreviation: COPD, chronic obstructive pulmonary disease.

Table 3 Annual Rate of Exacerbations in the 1-, 2-, and 3-Year Periods Following the Index Date

Variable	Number of Exacerbations	Person-Years	Exacerbation Rate (Per 100 Person-Years)	Unadjusted*	Adjusted^†
				Rate Ratio (95% CI); P-Value	Rate Ratio (95% CI); P-Value
1-year follow-up after the index date
Late COPD diagnosis	6052	6430.72	94.11	1.99 (1.84–2.15); P < 0.0001	1.67 (1.56–1.80); P < 0.0001
Early COPD diagnosis	1513	3163.26	47.83
2-year follow-up after the index date
Late COPD diagnosis	11,521	11,406.45	101.00	1.96 (1.84–2.09); P < 0.0001	1.68 (1.58–1.79); P < 0.0001
Early COPD diagnosis	2924	5560.81	52.58
3-year follow-up after the index date
Late COPD diagnosis	15,715	14,425.96	108.94	1.95 (1.83–2.07); P < 0.0001	1.68 (1.59–1.79); P < 0.0001
Early COPD diagnosis	4013	7011.69	57.23

Notes: P-value from the Wald test obtained from both the adjusted and unadjusted negative binomial regression model. *Used only late/early as independent variable. ^†Adjusted for pneumonia, depression, baseline eosinophil levels, osteoporosis, atrial fibrillation, and the “a priori” variables, which include age, CCI, BMI, smoking status, GOLD severity, and adherence to COPD treatment.

Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; GOLD, Global Initiative for Chronic Obstructive Lung Disease.

Table 4 Annual Rate of COPD-Related Clinic Visits, A&E Visits, and Hospitalizations After 1, 2, and 3 Years from the Index Date

Variable	Number of Visits	Person-Years	Rate (Per 1000 Person-Years)	Adjusted^a
				Rate Ratio (95% CI)	P-value
Clinic visits^b
Within 1 year after the index date
Late COPD diagnosis	10,253	6431	159.44	0.91 (0.87–0.94)	<0.0001
Early COPD diagnosis	5544	3163	175.26
Within 2 years after the index date
Late COPD diagnosis	19,229	11,406	168.58	0.99 (0.96–1.02)	0.5024
Early COPD diagnosis	9614	5561	172.89
Within 3 years after the index date
Late COPD diagnosis	25,712	14,426	178.23	1.03 (1.00–1.06)	0.0687
Early COPD diagnosis	12,453	7012	177.60
A&E visits
Within 1 year after the index date
Late COPD diagnosis	325	6431	5.05	1.03 (0.79–1.33)	0.8488
Early COPD diagnosis	127	3163	4.01
Within 2 years after the index date
Late COPD diagnosis	706	11,406	6.19	1.11 (0.91–1.35)	0.3140
Early COPD diagnosis	254	5561	4.57
Within 3 years after the index date
Late COPD diagnosis	999	14,426	6.93	1.19 (1–1.42)	0.0482
Early COPD diagnosis	345	7012	4.92
Hospitalizations
Within 1 year after the index date
Late COPD diagnosis	3241	6431	50.40	1.05 (0.93–1.19)	0.3916
Early COPD diagnosis	1196	3163	37.81
Within 2 years after the index date
Late COPD diagnosis	7192	11,406	63.05	1.12 (1.02–1.23)	0.0165
Early COPD diagnosis	2476	5561	44.53
Within 3 years after the index date
Late COPD diagnosis	10,606	14,426	73.52	1.18 (1.08–1.28)	0.0001
Early COPD diagnosis	3538	7012	50.46

Notes: P-value from the Wald test obtained from an adjusted negative binomial regression model. ^aAdjusted for the “a priori” variables plus for each parameter presented in the table. Clinic visits: depression, lung cancer, and diabetes. A&E visits: pneumonia, acute myocardial infarction, depression, and atrial fibrillation. Hospitalizations: baseline eosinophil groups, pneumonia, atrial fibrillation, hypertension, heart failure, and hyperlipidemia. Overall model: pneumonia, baseline eosinophil group, atrial fibrillation, depression, heart failure, and diabetes. ^bEach occurrence of a COPD-related clinic visit at different dates from the clinical or referral CPRD dataset with the COPD code.

Abbreviations: A&E, accident and emergency; CI, confidence interval; COPD, chronic obstructive pulmonary disease; CPRD, clinical practice research database.