Prescribing Pathways to Triple Therapy: A Retrospective Observational Study of Adults with Chronic Obstructive Pulmonary Disease in the UK

Figures & data

Table 1 Baseline Characteristics for the Study Population

Characteristics	Study Population (N = 82,300)
Age at COPD diagnosis, years
Mean (SD)	64.7 (10.6)
Median (IQR)	64.8 (57.4–72.4)
Male, n (%)	43,889 (53.3)
Comorbidities,^a n (%)
Cardiovascular disease	55,076 (66.9)
Depression/anxiety	30,810 (37.4)
Gastroesophageal reflux disease	18,846 (22.9)
Diabetes	11,978 (14.6)
Heart failure	7216 (8.8)
Atrial fibrillation	6711 (8.2)
Osteoporosis	6519 (7.9)
BMI (closest value to COPD diagnosis)
n (%) with BMI available	79,564 (96.7)
Mean (SD)	27.1 (6.3)
Median (IQR)	26.3 (22.7–30.5)
Phenotype (asthma), n (%)
Asthmatic	18,800 (22.8)
Non-asthmatic	63,500 (77.2)
Age at first triple therapy, years
Mean (SD)	69.0 (10.6)
Median (IQR)	69.5 (61.8–76.8)
Follow-up period in months
Mean (SD)	46.6 (33.2)
Median (IQR)	40.0 (19.3–68.4)
Initiation of triple therapy, n (%)
Before COPD diagnosis	6787 (8.2)
At COPD diagnosis	4641 (5.6)
After COPD diagnosis	70,872 (86.1)
Pre-triple therapy exacerbation rate, per patient per year
n (%) with data available	70,872 (86.1)
Mean (SD)	1.8 (5.0)
95% CI	1.8–1.8
Median (IQR)	0.9 (0.2–2.2)
Min–max	0.0–365.3
Triple therapy exacerbation rate, per patient per year
n (%) with data available	82,300 (100.0)
Mean (SD)	2.3 (3.5)
95% CI	2.3–2.3
Median (IQR)	0.8 (0.0–3.2)
Min–max	0.0–34.8
GOLD 2011 category, n/N (%) [95% CI]
A or B at COPD diagnosis^b	22,126/57,223 (38.7) [38.0–39.3]
C or D at COPD diagnosis^b	35,097/57,223 (61.3) [60.8–61.8]
A or B at triple therapy initiation^c	17,933/70,872 (25.3) [24.7–25.9]
C or D at triple therapy initiation^c	52,939/70,872 (74.7) [74.3–75.1]

Notes: ^aProportions reflect those with non-missing values; ^bOnly for patients with at least 6 months of data prior to the date of COPD diagnosis; ^cOnly for patients with index date after COPD diagnosis.

Abbreviations: BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; GOLD, Global Initiative for Chronic Obstructive Lung Disease; IQR, interquartile range; SD, standard deviation.

Table 2 Time to, and Adherence to, Triple Therapy

Variables	Study Population (N = 82,300)
Time to triple therapy (months)^a
n (%) with data available	70,872 (86.1)
Mean (SD)	60.9 (63.1)
95% CI	60.4–61.3
Median (IQR)	42.5 (13.9–87.4)
Adherence to triple therapy, measured as a proportion of days covered (%)
Mean (SD)	81.8 (15.5)
95% CI	81.7–81.9
Median (IQR)	84.4 (70.5–95.7)

Note: ^aOnly includes patients with index date after COPD diagnosis.

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; IQR, interquartile range; SD, standard deviation.

Table 3 Most Frequently Recorded Treatment Pathways Prior to Triple Therapy

Pathways, n (%)	Study Population^a (n = 75,513)
ICS+LABA	21,042 (27.9)
LAMA	9882 (13.1)
No therapy prior to triple therapy	9139 (12.1)
ICS → ICS+LABA	7734 (10.2)
ICS	1929 (2.6)
ICS+LABA → LAMA	1691 (2.2)
LABA → ICS+LABA	1639 (2.2)
ICS → ICS+LAMA	1586 (2.1)
LAMA → ICS+LABA	1055 (1.4)
ICS → ICS+LABA → ICS → ICS+LABA	883 (1.2)
LAMA → LABA+LAMA	792 (1.0)
ICS+LAMA	760 (1.0)
Other^b	17,381 (23.0)

Notes: ^aOnly patients whose first instance of triple therapy was on or after COPD diagnosis were included, and only the period on and after COPD diagnosis was considered. Short-acting bronchodilators were omitted from the treatment pathways; ^bPathways with <1% of patients.

Abbreviations: COPD, chronic obstructive pulmonary disease; ICS, inhaled corticosteroid; LABA, long-acting β₂-agonist; LAMA, long-acting muscarinic antagonist.

Table 4 Factors Influencing the Pathway to Triple Therapy (Multivariate Analysis) [N = 70,872]^a,b

Pathways	OR (95% CI)	p value
Phenotype: Asthmatic vs Non-Asthmatic
ICS+LABA	5.37 (4.85–5.95)	<0.001
LAMA	1.16 (1.03–1.30)	0.014
ICS → ICS+LABA	4.85 (4.35–5.42)	<0.001
ICS	3.58 (3.11–4.12)	<0.001
ICS+LABA → LAMA	4.09 (3.54–4.73)	<0.001
LABA → ICS+LABA	2.70 (2.31–3.16)	<0.001
ICS → ICS+LAMA	4.16 (3.59–4.83)	<0.001
LAMA → ICS+LABA	2.08 (1.72–2.52)	<0.001
ICS → ICS+LABA → ICS → ICS+LABA	5.09 (4.28–6.06)	<0.001
LAMA → LABA+LAMA	1.03 (0.79–1.34)	0.831
ICS+LAMA	4.34 (3.62–5.21)	<0.001
Other	3.94 (3.55–4.38)	<0.001
Gender: male vs female
ICS+LABA	0.89 (0.83–0.95)	<0.001
LAMA	1.01 (0.94–1.09)	0.786
ICS → ICS+LABA	0.88 (0.82–0.95)	0.002
ICS	0.82 (0.73–0.91)	<0.001
ICS+LABA → LAMA	0.86 (0.77–0.96)	0.009
LABA → ICS+LABA	0.93 (0.83–1.04)	0.192
ICS → ICS+LAMA	0.87 (0.78–0.98)	0.022
LAMA → ICS+LABA	1.14 (0.99–1.30)	0.068
ICS → ICS+LABA → ICS → ICS+LABA	0.80 (0.69–0.92)	0.003
LAMA → LABA+LAMA	1.03 (0.88–1.20)	0.728
ICS+LAMA	0.93 (0.80–1.08)	0.352
Other	0.92 (0.86–0.98)	0.013
GOLD 2011 category pre-triple therapy (considering 12 months prior to, or on, index date, and post-COPD diagnosis) (C/D [severe] vs A/B [non-severe])
ICS+LABA	2.86 (2.67–3.06)	<0.001
LAMA	2.26 (2.09–2.43)	<0.001
ICS → ICS+LABA	5.03 (4.62–5.48)	<0.001
ICS	2.82 (2.50–3.18)	<0.001
ICS+LABA → LAMA	3.28 (2.88–3.74)	<0.001
LABA → ICS+LABA	4.29 (3.73–4.93)	<0.001
ICS → ICS+LAMA	3.92 (3.42–4.50)	<0.001
LAMA → ICS+LABA	3.14 (2.68–3.67)	<0.001
ICS → ICS+LABA → ICS → ICS+LABA	6.58 (5.36–8.07)	<0.001
LAMA → LABA+LAMA	3.74 (3.11–4.51)	<0.001
ICS+LAMA	1.95 (1.66–2.30)	<0.001
Other	4.77 (4.44–5.13)	<0.001
Age at COPD diagnosis (per 1-year increase)
ICS+LABA	0.98 (0.97–0.98)	<0.001
LAMA	0.99 (0.99–0.99)	<0.001
ICS → ICS+LABA	0.97 (0.96–0.97)	<0.001
ICS	0.98 (0.98–0.99)	<0.001
ICS+LABA → LAMA	0.98 (0.97–0.98)	<0.001
LABA → ICS+LABA	0.98 (0.97–0.98)	<0.001
ICS → ICS+LAMA	0.98 (0.97–0.98)	<0.001
LAMA → ICS+LABA	0.98 (0.98–0.99)	<0.001
ICS → ICS+LABA → ICS → ICS+LABA	0.95 (0.95–0.96)	<0.001
LAMA → LABA+LAMA	0.98 (0.97–0.99)	<0.001
ICS+LAMA	0.98 (0.98–0.99)	<0.001
Other	0.97 (0.97–0.97)	<0.001
Pre-triple therapy exacerbation rate (per additional exacerbation)
ICS+LABA	0.97 (0.97–0.98)	<0.001
LAMA	0.98 (0.97–0.98)	<0.001
ICS → ICS+LABA	0.94 (0.92–0.95)	<0.001
ICS	0.96 (0.94–0.97)	<0.001
ICS+LABA → LAMA	0.96 (0.94–0.97)	<0.001
LABA → ICS+LABA	0.93 (0.90–0.95)	<0.001
ICS → ICS+LAMA	0.91 (0.89–0.94)	<0.001
LAMA → ICS+LABA	0.93 (0.90–0.96)	<0.001
ICS → ICS+LABA → ICS → ICS+LABA	0.96 (0.94–0.99)	<0.001
LAMA → LABA+LAMA	0.95 (0.92–0.98)	<0.001
ICS+LAMA	0.99 (0.98–1.00)	0.078
Other	0.95 (0.94–0.95)	<0.001

Notes: ^aOnly includes patients with an index date after COPD diagnosis; ^bReference category: no therapy prior to triple therapy.

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; GOLD, Global Initiative for Chronic Obstructive Lung Disease; ICS, inhaled corticosteroid; LABA, long-acting β₂-agonist; LAMA, long-acting muscarinic antagonist; OR, odds ratio.

Table 5 Association Between Non-Adherence, Triple Therapy Exacerbation Rate and Healthcare Resource Utilization (Adjusted Analyses)

Parameters	Means Ratio (95% CI)	Wald Chi-Square, DF	p value
Non-adherence and triple therapy exacerbation rate (n = 41,111)^a
Adherence (adherent vs non-adherent)	1.05 (1.03–1.07)	18.65, 1	<0.001
Phenotype (asthmatic vs non-asthmatic)	1.33 (1.30–1.36)	582.44, 1	<0.001
Gender (male vs female)	0.86 (0.85–0.88)	193.14, 1	<0.001
GOLD 2011 pre-triple therapy (C/D vs A/B)^b	1.86 (1.82–1.91)	2466.06, 1	<0.001
Age at index date (1-year increase)	1.00 (1.00–1.00)	3.14, 1	0.076
Non-adherence and healthcare resource utilization (n = 65,578)^a
Adherence (adherent vs non-adherent)	1.15 (1.14–1.16)	1840.98, 1	<0.001
Phenotype (asthmatic vs non-asthmatic)	1.05 (1.04–1.05)	164.22, 1	<0.001
Gender (male vs female)	0.97 (0.97–0.98)	78.04, 1	<0.001
GOLD 2011 pre-triple therapy (C/D vs A/B)^b	1.21 (1.20–1.22)	2612.04, 1	<0.001
Age at index date (1-year increase)	1.00 (1.00–1.00)	480.59, 1	<0.001

Notes: ^aOnly includes patients with index date after COPD diagnosis, and with at least 6 months of follow-up; ^bConsidering 12 months prior to or on index date, and post-COPD diagnosis.

Abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; DF, degrees of freedom; GOLD, Global Initiative for Chronic Obstructive Lung Disease.

Table 6 Changes in Therapy After Exacerbations

Variables	Study Population (N = 82,300)
Patients with ≥1 annual review,^a n (%)	65,129 (79.1)
Annual reviews per patient^a,b
Mean (SD)	2.9 (1.9)
Median (IQR)	2.0 (1.0–4.0)
Outcome of review, n (%) [N = 189,465]
Therapy remained the same	135,551 (71.5)
Therapy step-up	44,665 (23.6)
Therapy changed	4442 (2.3)
Therapy step-down	2963 (1.6)
Therapy discontinued	1844 (1.0)

Notes: ^aOnly patients with at least 90 days of data after the annual review date were included in this analysis; ^bConsidering only annual reviews for patients with at least 1 exacerbation in the previous 12 months.

Abbreviations: IQR, interquartile range; SD, standard deviation.

Table 7 Time to Step-Down from Initiation of Triple Therapy, and Most Frequent Treatments Following Step-Down from Triple Therapy

Time to Stepping Down from Start of Triple Therapy, Months	Study Population (N = 82,300)
Patients who discontinued or stepped down, n (%)	48,616 (59.1)
Median (IQR)	5.2 (1.2–15.2)
Pathway, n (%)
ICS+LABA+SABA	14,785 (18.0)
LAMA+SABA	10,677 (13.0)
LAMA	9301 (11.3)
ICS+LABA	5120 (6.2)
ICS+LABA+SABA+SAMA	3144 (3.8)
LABA+LAMA+SABA	1403 (1.7)
ICS+LAMA+SABA	1154 (1.4)
Discontinued	1107 (1.3)

Note: Pathways with ≥1% of patients are reported.

Abbreviations: ICS, inhaled corticosteroid; IQR, interquartile range; LABA, long-acting β₂-agonist; LAMA, long-acting muscarinic antagonist; SABA, short-acting β₂-agonist; SAMA, short-acting muscarinic antagonist.