NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Lin JS, Webber EM, Thomas RG. Screening for Chronic Obstructive Pulmonary Disease: A Targeted Evidence Update for the U.S. Preventive Services Task Force [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2022 May. (Evidence Synthesis, No. 215.)
Screening for Chronic Obstructive Pulmonary Disease: A Targeted Evidence Update for the U.S. Preventive Services Task Force [Internet].
Show detailsLiterature Search
Results of this search represent literature published since the previous systematic review on this topic. We screened 6387 abstracts and assessed 229 full-text articles for inclusion (Appendix B Figure 1). After screening the full-text articles, no studies were included for KQ1, 16 studies (in 28 articles)66–80 were included for KQ2, and eight studies (in 20 articles) for KQ3.69, 72, 75, 76, 78, 81, 82 The full list of included studies and their ancillary articles is available in Appendix C. The list of excluded studies with reasons for their exclusion is available in Appendix D.
KQ1. Does Screening for COPD Improve Health-Related Quality of Life or Reduce Morbidity or Mortality?
Summary of Results
We found no trials examining the effectiveness of screening or active case finding for COPD on health outcomes. Screening or active case-finding studies to date are limited to describing the yield of COPD cases (Chapter 4: Discussion).
KQ2. Does Treatment of Screen Detected or Mild to Moderate COPD Improve Health-Related Quality of Life or Reduce Morbidity or Mortality?
Summary of Results
We included 16 trials evaluating the treatment of mild to moderate, or minimally symptomatic, COPD.66–80, 82 Three of these trials (n=20,058) evaluated inhaled bronchodilators (LAMA, LABA) and/or ICS,72, 75, 76 and 13 trials (n=3,658).66–74, 77–80 evaluated non-pharmacologic interventions (i.e., self-management interventions, exercise counseling interventions, supervised exercise and pulmonary rehabilitation interventions, and clinician education interventions).66–71, 73, 74, 77–80
One large RCT (SUMMIT) (n=16,590) demonstrated that LABA, ICS or LABA/ICS could reduce the annual rate of exacerbations in adults with fairly symptomatic (i.e., mean SGRQ score 45-46 or CAT score 18-19) moderate COPD (i.e., mean FEV1 59.7% predicted) with a median followup of 1.8 years.76, 83 Subgroup analyses of another RCT (UPLIFT) demonstrated that LAMA could reduce clinically important deterioration (including clinically significant difference in SGRQ and exacerbations) in persons with moderate COPD (i.e., mean FEV1 59% predicted) (n=2603)84 and reduce exacerbations in patients with minimal symptoms (i.e., GOLD category A) (n=357) at 48 months.85
Although a large variety of non-pharmacologic interventions were studied, overall, there was no consistent benefit observed across a range of outcomes (i.e., exacerbations, HRQoL, dyspnea, exercise or physical performance measures, mental health, smoking cessation) at 26 to 104 weeks. Two RCTs (n=114 and 239) evaluated the same exercise-focused web-based intervention in a VA population.78, 86 Only one of these two trials (n=114) demonstrated a reduction in COPD exacerbations at 65 weeks,78 the other trial (n=239) found no reduction in a composite outcome of exacerbations and pneumonia at 52 weeks.86 Other trials, not conducted in the US, evaluating more intensive self-management interventions, supervised exercise, and pulmonary rehabilitation interventions in persons with mild to moderate COPD, or minimal symptoms, did not demonstrate a reduction in exacerbations or other outcomes. One cluster RCT (n=254 patients) evaluating clinician education/training demonstrated an increase in influenza vaccination, but not pneumococcal vaccination or other outcomes at 52 weeks (Chapter 4: Discussion).80 It is unclear if and how small sample sizes, usual care comparators in trials conducted outside the US, and/or poor adherence to the interventions contributed to the largely null findings of these trials.
Included Studies
Pharmacologic Interventions
We found three studies with newly published analyses in mild to moderate COPD, or minimally symptomatic persons (i.e., GOLD category A) since the 2016 recommendation (Tables 3 and 4). Two studies (PINNACLE and SUMMIT) evaluated LABAs, two studies (PINNACLE and UPLIFT) evaluated LAMAs, one study (PINNACLE) evaluated LABA/LAMA, and one study (SUMMIT) evaluated ICS with or without LABA.
Table 3
Key Question 2 Results: Study Characteristics for Included RCTs of Pharmacologic Treatment.
Table 4
Key Question 2 Results: Population Characteristics for Included RCTs of Pharmacologic Treatment.
The included PINNACLE study was actually a post-hoc analysis of GOLD category A patients (n=729) in three PINNACLE trials (PINNACLE 1, PINNACLE 2, PINNACLE 4) (n=4983).72 These trials evaluated glycopyrrolate (LAMA), formoterol fumarate (LABA), and glycopyrrolate/formoterol fumarate (LAMA/LABA) versus placebo. Of note, participants on a stable dose of ICS were allowed to continue this therapy. In this subgroup analysis, the mostly male population had a mean age of 65.5 years, and a mean pack year history of 45.6 years, but a low symptom score (i.e., CAT mean score 6.5) consistent with GOLD category A classification.
The SUMMIT trial (n=16,590) evaluated vilanterol (LABA), fluticasone furoate (ICS), and vilanterol/fluticasone furoate (LABA/ICS) versus placebo.76, 83 All background LABA, LAMA, and ICS were discontinued prior to the trial. In this trial, all participants had to have known cardiovascular disease (CVD) or be at increased risk for CVD (as defined by taking medication for more than 2 cardiovascular risk factors). In this trial, the mostly male population had a mean age of 65 years, mean pack year history of 41 years, mean FEV1 of about 60 percent predicted, and a fairly high symptom score (i.e., SGRQ 45-56, CAT 18-19).
The UPLIFT trial (n=5993)75 was included in the previous systematic review that supported the 2016 recommendation, but had 2 newly published post-hoc subgroup analyses in adults with moderate COPD (Stage II) (n=2603)84 and GOLD category A patients (n=357).85 UPLIFT evaluated tiotropium (LAMA) versus placebo. Patients were allowed to continue all background medication except for other muscarinic antagonists. In the included subgroup analyses for this trial, the mostly male population had a mean age of 65 years, and a mean pack year history of about 47 years. In the subgroup analysis in adults with moderate COPD, the mean FEV1 was 59 percent predicted with a fairly high symptom score (i.e., SGRQ 41.5).84, 87 In the GOLD category A subgroup analysis, the mean FEV1 was 60.4 percent predicted but the mean symptom score was lower (i.e., SGRQ 16.8).85
Although these trials were well conducted multisite RCTs, a few limitations are worth noting. The population studied in SUMMIT is less generalizable to screen-detected populations (i.e., greater degree of airflow obstruction and symptoms), and this trial did not include subgroup analyses of less symptomatic participants as the trial’s inclusion criteria specified a symptom score of greater than or equal to 2 on the mMRC. While both PINNACLE and UPLIFT include subgroup analyses or populations more generalizable to screen-detected persons, all were post hoc analyses and only the subgroup analysis of GOLD category A participants in the UPLIFT trial reported interaction testing and controlled for potential confounders (Table 5).85 However, this subgroup analysis of GOLD category A participants only included 357 participants. Last, the PINNACLE trials were limited to 24-week followup.
Table 5
Key Question 2 Results: Subgroup Credibility Ratings for RCTs of Pharmacologic Treatment.
Non-Pharmacologic Interventions
We found 13 trials evaluating non-pharmacologic interventions used in the management of mild to moderate COPD, or minimally symptomatic persons (i.e., GOLD category A): seven trials of self-management interventions, one trial of exercise-only counseling, three trials of intensive supervised exercise or pulmonary rehabilitation, and two trials of clinician education/training on COPD care (Tables 6 and 7).
Table 6
Key Question 2 Results: Study Characteristics for Included Trials of Non-Pharmacologic Treatments.
Table 7
Key Question 2 Results: Intervention Characteristics for Included Trials of Non-Pharmacologic Treatments.
Self-Management Interventions
Three trials (n=165 to 577) evaluated in-person or phone-based (as opposed to web-only) self-management interventions.67, 69, 79 Although each of the interventions was different, self-management generally included education on COPD, medications, healthy lifestyle (including but not limited to exercise), tobacco cessation, and an exacerbation management/action plan. These interventions ranged from two to four in-person or phone sessions over several weeks with or without followup calls. The comparator groups received usual care or non-tailored written education materials. These trials included both male and female populations with a mean age ranging from 64.9 to 70.4 years, and 23 to 30 percent of the populations included were active smokers. The trial by Jolly and colleagues (n=577),69 conducted in the UK, explicitly recruited minimally symptomatic persons with COPD (MRC score 1 or 2), and the mean FEV1 was 71.7 percent predicted, with a mean SGRQ of 28.7. Nonetheless, almost half of the included population had one or more exacerbations in the previous year. In the other two trials, conducted in the Netherlands, the mean FEV1s were lower (60.6 and 65.4 percent).67, 79
Another four trials (n=83 to 1325) evaluated a web-only self-management intervention.73, 77, 78, 86 Two trials evaluated interactive web-based self-management interventions broadly inclusive of topics addressed in the in-person self-management interventions; however, two trials conducted in the VA setting in the United States focused primarily on exercise (‘step count’) although included some information on disease self-management, self-efficacy, and an online community forum for support.78, 86 The comparator groups received usual care and/or a pedometer with exercise information. These trials included both male and female populations with a mean age ranging from 57.6 to 68.6 years, and 19 to 37 percent of the populations included were active smokers. One trial by Voncken-Brewster and colleagues (n=1325), conducted in the Netherlands, recruited both person with COPD and at risk for COPD, 67.6 percent of whom had an MRC score of 1 or 2; however, the mean FEV1 was not reported.77 In the other two trials the mean FEV1 was 60.0 to 62.6 percent.73, 78
Exercise Only Counseling
One trial (n=48 in primary care arm) by Altenburg and colleagues, conducted in the Netherlands, evaluated an in-person counseling intervention in which participants were randomized to receive five 30-minute sessions on improving physical activity or usual care. In this trial, the mostly male population had a median age of 65 years, a median pack year history of 30 years, and median FEV1 of 78 percent predicted.66
Supervised Exercise or Pulmonary Rehabilitation
One trial (n=90) by Fastenau and colleagues, conducted in the Netherlands, evaluated an intensive in-person supervised exercise intervention in which participants were randomized to receive twice weekly sessions (60 to 90 min) for 17 weeks or a low intensity session (30 min) once weekly. In this trial, the population was about half male and had a mean age of 62.5 years; 38 percent were current smokers. The mean FEV1 was 74.2 percent predicted.68
Two trials evaluated pulmonary rehabilitation programs; one trial (n=272) evaluated a home-based program,70 while the other trial (n=71) evaluated an intensive in person program.74 The trial by Liang and colleagues, conducted in Australia, evaluated an 8-week home-based pulmonary rehabilitation program involving one home visit and followup calls supplemented by a one-time in home pharmacist medication review and smoking cessation session (90 minutes) versus usual care with non-tailored written educational material. In this mostly male population, the mean age was 64.5 years, 61 percent were current smokers, and the mean FEV1 was 70.0 percent predicted.70 The other trial by Roman and colleagues, conducted in Spain, evaluated an intensive in-person pulmonary rehabilitation program consisting of three 60 minute sessions a week for 3 months with or without weekly sessions during months 4 through 12, versus usual care. This program also included counseling on medication and smoking cessation. In this mostly male population, the mean age was 64.2 years, and 34 percent were current smokers. This trial recruited only patients with moderate COPD, and the mean FEV1 was 60.3 percent predicted.74
Clinician Education Only Interventions
While some of the included trials implicitly or explicitly provided clinician education as part of the execution of the trial, for most trials, the primary intervention was directed at the patient, and not the clinician. However, two cluster randomized trials (n=216 to 254) conducted outside the United States evaluated clinician education and training for COPD, primarily aimed at general practitioners.71, 80 One trial by Zwar and colleagues, conducted in Australia, explicitly included persons with COPD identified via active case finding, and therefore also included training aimed at practice nurses to conduct case finding.80 As such, this trial tended to include persons with less severe symptoms (mean SGQR 19.5, CAT 10.2) and airflow obstruction (mean FEV1 74.6 percent predicted). The other trial by Markun and colleagues also included persons with less severe symptoms (69 percent with GOLD category A or B, mean CAT 10.6) and airflow obstruction (mean FEV1 67.3 percent predicted).71
All the included trials were fair quality without significant threats to validity. However, many of the trials had characteristics that may limit the ability to detect a true benefit (type II error): 1) small trials with a limited number of participants per intervention arm, 2) usual care (most trials were conducted outside the US) or control groups receiving more care than may typically be delivered in the US, and 3) poor adherence to or uptake of the intervention itself.
Detailed Results
Pharmacologic Interventions
Two of the three included trials demonstrated that LABA, ICS or LABA/ICS could reduce exacerbations in persons with fairly symptomatic moderate COPD and that LAMA could reduce exacerbations in patients with minimal symptoms (i.e., GOLD category A) (Table 9).
Table 8
Key Question 2 Results: Population Characteristics for Included Trials of Non-Pharmacologic Treatments.
Table 9
Key Question 2 Results: Outcomes for Included RCTs of Pharmacologic Treatment.
The analysis from the PINNACLE trials evaluating glycopyrrolate (LAMA), formoterol fumarate (LABA), and glycopyrrolate/formoterol fumarate (LAMA/LABA) versus placebo in GOLD category A patients (n=729) only reported mortality outcomes and adverse outcomes at 24 weeks. This pooled analysis found no difference in mortality outcomes.72 Harms are discussed in Key Question 3.
The SUMMIT trial (n=16,590) in persons with mild to moderate COPD (mean FEV1 59.7 percent, SGRQ 45-46) found that vilanterol (LABA), fluticasone furoate (ICS), and vilanterol/fluticasone furoate (LABA/ICS) all reduced moderate to severe COPD exacerbations and hospitalizations for exacerbations compared to placebo at a median followup of 1.8 years.76, 83 In this population, 39.2 percent of persons had a least one exacerbation in the year prior to trial recruitment. The percent reduction in the annual rate of moderate to severe exacerbations was higher for LABA/ICS (29 percent [95% CI, 22 to 35%]) than for LABA (10 percent [95 % CI, 2 to 18 %) or ICS (12 percent [95% CI, 4 to 19%) alone. This trial also found that LABA alone reduced pneumonias compared placebo at a median of 1.8 years, discussed in Key Question 3.
The UPLIFT trial subgroup analysis in moderate (stage II) COPD (n=2603) found that tiotropium (LAMA) reduced clinically important deterioration at 48 months compared to placebo (deterioration was a composite outcome of a decrease in trough FEV1 of ≥100 mL, increase in SGRQ total score from baseline ≥4 units [clinically meaningful difference], or moderate/severe exacerbation).84 Of note, these results are consistent with prior subgroup analyses by stage and FEV1 included in the previous review and do not add any new understanding of the benefits of tiotropium over placebo in this population.87, 88 Another subgroup analysis in GOLD category A patients (n=357) did demonstrate that LAMA reduced the proportion of persons with exacerbations compared to placebo at 48 months, which decreased to 48.4 percent compared to 54.4 percent (RR 0.64 [95% CI, 0.47 to 0.89]).85 The overall symptom score in these patients was low (i.e., SGRQ 16.8), although the severity of airflow obstruction (FEV1 60.4 percent) was similar to those in the moderate (stage II) COPD subgroup analyses.
Non-Pharmacologic Interventions
Self-Management Interventions
Overall, there were no consistent and significant benefits in outcomes across the four trials evaluating in person or phone-based self-management interventions. However, intervention intensity, outcomes reported, and duration of followup varied across the trials. Neither of the two trials evaluating exacerbation outcomes found a statistically significant difference in exacerbations or hospitalizations at 52 or 104 weeks 67, 69 (Table 10). All three trials reported disease specific or general HRQoL outcomes using different instruments (i.e., CRQ, SGRQ, CCQ, and EQ-5D); nonetheless, no trial demonstrated a clinically meaningful or statistically significant benefit in HRQoL at 26, 39 or 104 weeks (Table 11). And all three trials reported dyspnea outcomes using either the CRQ (dyspnea domain) or MRC instrument (Table 12). Across the trials there were no statistically significant improvements in dyspnea at 26, 39 or 52 weeks; however, in the trial by Bischoff and colleagues (n analyzed=110), the dyspnea outcome improved in the intervention group that received a non-tailored self-management support at 104 weeks (mean difference 0.40 [95% CI, 0.041 to 0.78]).67 The clinical significance of this finding is unclear, given the small change on a continuous measure and in the setting of no impact on the overall CRQ instrument. In addition, the trial by Jolly and colleagues reported other mental health, exercise, and smoking cessation outcomes, finding no statistically significant improvements in anxiety/depression (data not shown), objective measures of exercise, or smoking cessation at 26 or 52 weeks (Tables 13 and 14).69
Table 10
Key Question 2 Results: Exacerbation and Hospitalization Outcomes for Included Non-Pharmacologic Trials.
Table 11
Key Question 2 Results: Quality of Life Outcomes for Included Trials of Non-Pharmacologic Treatment.
Table 12
Key Question 2 Results: Dyspnea Outcomes for Included Trials of Non-Pharmacologic Treatment.
Table 13
Key Question 2 Results: Exercise and Physical Functioning Outcomes for Included Trials of Non-Pharmacologic Treatment.
Table 14
Key Question 2 Results: Smoking Cessation and Vaccination Outcomes for Included Trials of Non-Pharmacologic Treatment.
Overall, there were no consistent benefits seen in outcomes across the four RCTs evaluating web-based self-management interventions. Only one exercise-focused web-based intervention did demonstrate a reduction in exacerbations (Table 10). Two trials evaluated the same web-based self-management intervention which focused primarily on exercise (‘step count’).78, 86 The initial trial by Moy and colleagues (n analyzed=238) found no difference in exacerbations or pneumonia at 52 weeks;86 whereas the second trial by Wan and colleagues (n analyzed=109)78 reported a reduction in exacerbations at 65 weeks in participants randomized to the web-based intervention versus the pedometer only group (relative risk [RR] 0.51 [95% CI, 0.31 to 0.85]).78 The absolute reduction in exacerbations is not reported, but at baseline, the mean (standard deviation) rate of COPD exacerbations in the prior year was 0.42 (0.98) in the intervention group and 0.27 (0.63) in the pedometer only group. All four trials reported disease-specific or general HRQoL outcomes using different instruments (i.e., CAT, CCQ, SGRQ, and EQ-5D); nonetheless, no trial demonstrated a clinically meaningful or statistically significant benefit in HRQoL at 26 or 52 weeks (Table 11).73, 77, 78, 86 Two trials reporting dyspnea outcomes using either the MRC or modified version of the instrument found no statistically significant benefit at 26 or 52 weeks (Table 12).73, 77 All four trials reported objective measures of exercise and found no improvement in exercise outcomes at 26 or 52 weeks, including both trials evaluating the exercise-focused web-based intervention (Table 13).73, 77, 78, 86 In addition, one trial found no benefit in smoking cessation at 26 weeks, using a variety of measures (Table 14).77
Exercise Only Counseling
One trial by Altenburg and colleagues found no statistically significant benefit on HRQoL (using the CCQ, CRQ), or on objective measures of global physical performance (using the 6MWT) at 65 weeks for an in-person exercise counseling intervention (Tables 11 and 13). This study did not report any exacerbation or dyspnea outcomes.66
Supervised Exercise or Pulmonary Rehabilitation
Overall, there were no benefits in reported outcomes across the three trials evaluating supervised exercise or pulmonary rehabilitation interventions. Only one of these trials reported on exacerbations and found no reduction in exacerbations at 52 weeks (Table 10).74 All three trials reported disease specific HRQoL outcomes using different instruments (i.e., CAT, CCQ, SGRQ); none of the trials demonstrated a clinically meaningful or statistically significant benefit in HRQoL at 26 or 52 weeks (Table 11).68, 70, 74 None of the three trials reporting dyspnea outcomes using the CRQ (dyspnea domain), MRC or modified MRC instrument found a statistically significant benefit at 26 or 52 weeks (Table 12).68, 70, 74 One trial by Liang and colleagues reporting mental health outcomes found no improvement in depression and anxiety scores at 26 and 52 weeks (data not shown).70 Two trials reporting objective measures of physical performance found no improvement in the 6MWD at 26 and 52 weeks (Table 13).68, 74
Clinician Education Only Interventions
Overall, there were no benefits in reported outcomes in the two cluster RCTs evaluating clinician education and training interventions; however, one trial by Zwar and colleagues (n analyzed= 222) did demonstrate a statistically significant increase in influenza but not pneumococcal vaccinations or smoking cessation at 52 weeks (Chapter 4: Discussion).80 Reporting of patient level outcomes was more limited for these trials as compared to the other included intervention trials. Neither trial found a benefit on disease specific HRQoL outcomes as measured by the CAT at 52 weeks (Table 11).
KQ3. What Are the Adverse Effects of COPD Treatments in This Population?
Summary of Results
Overall, there was very limited included evidence on the harms of pharmacologic and non-pharmacologic interventions in the treatment of mild to moderate, or minimally symptomatic COPD. None of the included trials (i.e., LAMA, LABA, LABA/LAMA or ICS with or without LABA, self-management intervention) that reported adverse effects (six trials) found significant harms. However, the small number of included participants and limited length of followup in the included trials (or their relevant subgroup analyses) limits the ability to detect uncommon harms or longer-term harms. Given this update’s focus on screen-relevant populations, only two large observational studies addressing harms of pharmacologic treatment met our inclusion criteria. These studies demonstrated that initiation of LAMA or LABA may increase the risk of a serious cardiovascular event in treatment naïve patients and that ICS may increase the risk of developing diabetes. However, these observational studies are a subset of a much larger body of evidence on serious harms of bronchodilators and ICS in COPD (Chapter 4: Discussion).
Included Studies
Pharmacologic Interventions
The three included trials from Key Question 2 also report adverse effects (Tables 3 and 4).72, 76, 83–85, 87, 88 Two trials (PINNACLE and SUMMIT) evaluated LABAs, two trials (PINNACLE and UPLIFT) evaluated LAMAs, one trial (PINNACLE) evaluated LABA/LAMA, and one trial (SUMMIT) evaluated ICS with or without LABA. For a description of these trials, see Key Question 2: Included studies section.
In addition to these three trials, we also found two observational studies evaluating the harms of LABA, LAMA, or ICS in screen-relevant populations (Table 16). One nested case-control study in Taiwan by Wang and colleagues examined the cardiovascular risk for the initiation of LABA and LAMA in treatment naïve COPD patients.82 The other study was a matched cohort study in the UK by Price and colleagues which examined the risk of diabetes and osteoporosis from ICS use in populations in whom the majority of patients were stage I or II and GOLD category A or B.81 The mean age in the observational studies was slightly older than the trial populations. The mean FEV1 or symptom score was not reported, nonetheless the studies provided some indication of severity of COPD. The study by Price and colleagues conducted subgroup analysis for GOLD categories A and B patients.81 The vast majority of studies evaluating harms of pharmacologic treatment were excluded because no description was reported regarding the severity or symptoms of included populations. Both of these studies used large nationally representative databases and tried to adjust for known confounders.
Table 15
Key Question 3 Results: Adverse Events for Included RCTs of Pharmacologic Treatment.
Table 16
Key Question 3 Results: Adverse Events for Included Observational Studies of Pharmacologic Treatment.
Non-Pharmacologic Interventions
Only three of the included non-pharmacologic intervention trials from Key Question 2 also reported on harms.69, 78, 86 We found no additional studies that specifically evaluated the harms of non-pharmacologic interventions in a screen-relevant population. All three trials evaluated self-management interventions. One trial by Jolly and colleagues (n=577) conducted in the UK, explicitly recruited minimally symptomatic persons with COPD (MRC score 1 or 2) and evaluated a phone-based intervention. The other two trials by Moy and colleagues (n=238) and by Wan and colleagues (n=114) were conducted in VA settings and included more symptomatic persons and evaluated an exercise-focused web-based intervention. For additional details of these trials, see Key Question 2: Included studies section.
Detailed Results
Pharmacologic Interventions
Overall, none of the included trials found significant harms for LAMA, LABA, LABA/LAMA, or ICS with or without LABA (Table 15). However, the subgroup analyses from PINNACLE and UPLIFT included a limited number of participants and therefore are quite limited in their ability to detect less common harms, and both PINNACLE and SUMMIT have limited length of followup and therefore are limited in their ability to detect longer term harms.
The analysis from the PINNACLE trials evaluating glycopyrrolate (LAMA), formoterol fumarate (LABA), and glycopyrrolate/formoterol fumarate (LAMA/LABA) versus placebo in GOLD category A patients (n=729) reported adverse outcomes at 24 weeks.72 This pooled analysis across three trials did not report any statistically significant harms. There is a signal of increased serious adverse events related to study treatment for LABA/LAMA compared to placebo (1.8 percent versus 0.9 percent), but this was based on a total of 5 adverse events across both arms.
The SUMMIT trial (n=16,590) in persons with mild to moderate COPD (mean FEV1 59.7 percent, SGRQ 45-46) found that vilanterol (LABA), fluticasone furoate (ICS), and vilanterol/fluticasone furoate (LABA/ICS) did not increase adverse events compared to placebo at a median followup of 1.8 years.76, 83 This trial did not find an increase in pneumonias from ICS with or without LABA. In fact, this study demonstrated that LABA alone reduced pneumonias compared to placebo at a median of 1.8 years, 3.9 versus 5.2 percent (HR 0.72 [95% CI, 0.59 to 0.89]).
The subgroup analyses from the UPLIFT trial did not report any increase in adverse events from tiotropium (LAMA) versus placebo at 48 months.84, 85, 87, 88 However, in the subgroup analysis of GOLD category A patients (n=357), there is a signal of increased major adverse cardiac events from LAMA compared to placebo (5.9 percent versus 1.8 percent), but this was based on a total of 14 adverse events across both arms.85
The nested case-control study by Wang and colleagues (n=37,719 cases and 146,139 controls) observed that the initiation of both LABA and LAMA was associated with an increase in cardiovascular events within 30 days of its initiation (OR 1.50 [95% CI, 1.35 to 1.67 and adj OR 1.52 [95% CI, 1.28 to 1.80], respectively). However, notably this risk association was absent, or even reduced, with prevalent use. Risk of cardiovascular events did not seem to vary by different LABA or LAMA drugs, concomitant COPD medications, history of CVD, or history of prior COPD exacerbations (Table 16).82
The matched cohort study by Price and colleagues (n= 17,970 for diabetes onset and n= 19,898 for osteoporosis onset) observed that long-term ICS use, compared with long-acting bronchodilator use, was associated with an increased risk of developing diabetes (HR 1.27 [95% CI, 1.07 to 1.50]). The subgroup analysis conducted in GOLD categories A and B patients combined had similar findings (Table 16). The study also noted a dose–response relationships at mean ICS exposures of 500 µg/day or greater versus less than 250 µg/day, fluticasone propionate–equivalent.81 The association with an increased risk of developing osteoporosis was not statistically significant and not observed in the subgroup analysis in GOLD categories A and B patients.
Non-Pharmacologic Interventions
The RCT by Jolly and colleagues (n=577) reported no difference in self-reported adverse events at 52 weeks between the phone-based intervention compared to the usual care group (24 versus 20 events, p value NR).69 The RCT by Moy and colleagues (n=238) reported no difference in pulmonary or cardiac adverse events between the web-based intervention and pedometer only groups; however, more minor musculoskeletal adverse events occurred in the intervention group (27.9%) than the pedometer group (10%).86 The RCT by Wan and colleagues (n=114) reported no statistically significant difference in adverse events at 13 weeks between the group receiving the web-based intervention compared to the pedometer only group (14 versus 10 events, p=0.54 respectively).78 In all of the trials, none of the serious adverse events were determined to be related to the study intervention.
- Literature Search
- Does Screening for COPD Improve Health-Related Quality of Life or Reduce Morbidity or Mortality?
- Does Treatment of Screen Detected or Mild to Moderate COPD Improve Health-Related Quality of Life or Reduce Morbidity or Mortality?
- What Are the Adverse Effects of COPD Treatments in This Population?
- Results - Screening for Chronic Obstructive Pulmonary Disease: A Targeted Eviden...Results - Screening for Chronic Obstructive Pulmonary Disease: A Targeted Evidence Update for the U.S. Preventive Services Task Force
Your browsing activity is empty.
Activity recording is turned off.
See more...