Introduction
Chronic obstructive pulmonary disease (COPD) is a highly prevalent condition, with 6.2% of adults reporting being diagnosed with COPD in 2017, and its burden is anticipated to increase as the population continues to age [
1,
2]. Acute exacerbations of COPD lead to deterioration in lung function and quality of life, increased risk for mortality, and frequently require emergency department (ED) visits and hospitalization, therefore contributing to excess healthcare use [
3‐
5]. Exacerbations along with complex care needed to address multiple comorbidities commonly present among patients with COPD imparts a great economic burden to healthcare systems [
6]. Identifying modifiable risk factors is critical for the prevention of COPD exacerbations and the consequent reduction in healthcare utilization and costs.
Insomnia is a common complaint among patients with COPD [
7‐
10]. Sleep difficulties in COPD may plausibly arise from smoking, psychiatric and medical comorbidities including depression, anxiety, obstructive sleep apnea, restless legs syndrome, and pain, supplemental oxygen use and medications for the treatment of COPD, and nocturnal awakenings due to nighttime respiratory symptoms such as cough and dyspnea [
11]. Irrespective of etiology, insomnia has been linked to adverse outcomes in COPD including reductions in quality of life and daytime function, COPD-related symptoms and incident exacerbations, and increased risk for mortality [
12‐
15].
Untreated insomnia is associated with substantial healthcare utilization and costs, particularly among older adults and those with comorbidities [
16‐
22]. Relative to individuals without insomnia, rates of inpatient, ED, and outpatient care and healthcare costs, primarily driven by inpatient costs, are significantly higher among individuals with insomnia [
18,
20]. Even after controlling for comorbidities, individuals with insomnia continue to demonstrate greater healthcare utilization and costs [
16,
20]. When coupled with comorbidities, healthcare costs are as much as 80% higher in the 12 months after insomnia diagnosis [
16]. To our knowledge, no prior studies have examined the impact of insomnia on healthcare utilization and costs in patients with COPD.
This study investigated the associations between insomnia and COPD-related healthcare utilization and costs utilizing a large cohort of patients with COPD receiving care within the national Veterans Health Administration (VHA). We hypothesized that patients with COPD and insomnia would have greater rates of COPD-related healthcare utilization including outpatient visits, hospitalizations, and number of prescription fills of corticosteroids and/or antibiotics and higher COPD-specific outpatient visit and hospitalization costs compared to those with COPD only.
Discussion
To the best of our knowledge, this study presents for the first time the prevalence of insomnia and its associated impact on utilization of healthcare services and associated costs in a large national cohort with COPD. Results revealed an approximately 4-fold higher prevalence of insomnia in patients with COPD compared to rates reported in the general population [
37]. Insomnia was associated with increased COPD-related healthcare utilization and costs. Patients with insomnia had hospital stays that were 38% longer than patients without insomnia, which likely contributed to greater hospitalization-related costs among patient with insomnia.
Prior studies of COPD patients reported the prevalence of insomnia disorder to be between 25% and 47.2% [
7,
10,
38] and utilized generally small sample sizes and non-diagnostic insomnia criteria such as questionnaires or specific scales. In contrast, our cohort included over one million patients with COPD and utilized ICD codes and sedative-hypnotic prescription as indicators of insomnia. The high prevalence of insomnia (37%) found in our study aligns with prevalence rates in previous reports that utilized diagnostic insomnia criteria [
39,
40]. Age, sex, and co-existing medical and psychiatric conditions have been identified as risk factors for insomnia [
41]. In our study and prior studies in patients with COPD, younger age, female sex, current smoking, and physical and mental disorders were found to be associated with insomnia [
7,
10]. Of particular note, obstructive sleep apnea, depression, and post-traumatic stress disorder were approximately 3 to 4-times more likely in those with insomnia compared to those without insomnia in our COPD cohort which was comprised of Veterans suggesting a potential critical role of comorbid sleep and psychiatric disorders in the manifestation of insomnia.
Comorbidities among patients with COPD increase the rates of all-cause and COPD-related hospitalizations, length of stay, and in-hospital costs [
42‐
46]. Our study builds upon the current literature by investigating the impact of comorbid insomnia on COPD-related healthcare utilization and costs, showing that insomnia is longitudinally predictive of higher rates of outpatient visits and hospitalizations, longer hospital length of stay, and hospital-related costs even after controlling for other comorbidities. Our results confirm an earlier report demonstrating baseline sleep disturbance suggestive of insomnia as a predictor of COPD-related emergency utilization (hospitalizations or ED visits) over the ensuing year [
12]. These findings suggest that insomnia is a strong contributor of healthcare outcomes and costs in patients with COPD. The connection between insomnia and healthcare use and costs may be partially explained by evidence indicating that sleep difficulties independently predict incident COPD exacerbations [
47,
48]. In our study, we examined the number of prescription fills for steroids and/or antibiotics, which can be indicative of a COPD exacerbation. Insomnia was longitudinally associated with greater prescription fills for corticosteroids and/or antibiotics, thus providing further support for insomnia being a risk factor for worse COPD outcomes.
There are several potential explanations for the association between insomnia and greater healthcare usage. Sleep disturbance is associated with an elevated systemic inflammatory response, increased C-reactive protein and interleukin-6, which could instigate COPD exacerbations due to increased systemic and airway inflammation [
49]. Sleep insufficiency compromises immune function, thus increasing susceptibility to upper and lower respiratory tract infections that can trigger COPD exacerbations [
50,
51]. Impairments in memory and attention is common in insomnia and cognitive dysfunction can lead to poor adherence to COPD medications and improper inhaler use [
52,
53]. Conversely, insomnia could be an indicator of more severe disease, as increased and unstable COPD symptoms and use of medications such as β-agonists and corticosteroids could lead to sleep disturbances [
11]. In addition, insomnia frequently co-occurs with common comorbidities in COPD, such as depression, anxiety, and obstructive sleep apnea, which have been associated with increased risk of COPD exacerbations [
54‐
57]. The relationship between insomnia and COPD is complex and likely multiple mechanisms are in effect at once. Management of multimorbidity in patients with COPD involves identification and treatment of comorbidities [
58]. Although treatment of COPD and other comorbid conditions may resolve insomnia symptoms, in some cases insomnia may become self-sustaining and consequently an independent disease process which requires targeted treatment. Cognitive-behavioral therapy for insomnia (CBT-I) is the first-line treatment for insomnia and is associated with reduced healthcare utilization and costs [
59,
60]. A recent clinical trial of CBT-I in patients with COPD reported decreases in insomnia and improvements in fatigue and dyspnea [
61]. Hypnotics, including benzodiazepine receptor agonists and sedating antidepressants, should be used with caution in patients with severe COPD and used short-term or intermittently in more stable patients as long-term use may be associated with adverse respiratory outcomes [
47]. Despite these risks, benzodiazepine receptor agonists are frequently prescribed among patients with COPD [
62,
63]. Non-benzodiazepine receptor agonists may have fewer respiratory depressant effects in COPD patients than benzodiazepine receptor agonists [
64]. Future studies are needed to examine whether treatment-related reductions in insomnia via CBT-I or pharmacological treatments are associated with reduced healthcare usage and costs in patients with COPD.
Strengths of the study include a large sample size inclusive of all COPD users of the largest integrated healthcare system in the United States, use of electronic medical records data rather than patient self-report for determination of healthcare utilization, and prospective, detailed utilization and cost data. The study also has several limitations. First, insomnia is frequently treated but not often diagnosed [
26]. We attempted to address this discrepancy by identifying insomnia through prescription fills for sedative-hypnotics; however, the indication for the prescribed medications could not be determined. Diagnosis of insomnia was based on ICD-9-CM and ICD-10-CM codes, yet the criteria used by providers for determining diagnosis is unclear. Therefore, the number of patients with insomnia in our study may over- or under-represent the true prevalence. Second, the use of ICD codes for the identification of COPD diagnosis could have led to misdiagnosis or underdiagnosis of COPD in our study. Additionally, the severity of COPD was unable to be reported and may be an important factor influencing healthcare use and costs. Unfortunately, availability of spirometry and Global Initiative for Chronic Obstructive Lung Disease data to confirm ICD diagnosis and severity of COPD was very limited in our study cohort due to variability in data collection, often which requires natural language processing of notes to extract relevant data. Prior research reported that sleep disturbance remained longitudinally associated with respiratory-related emergency utilization, even after controlling for forced expiratory volume in 1 s and COPD severity based on a validated survey [
12]. Third, this study included veterans within the VA, thus studies within other healthcare systems are needed to confirm associations between insomnia and healthcare utilization and costs among patients with COPD. Finally, because utilization outcomes were extracted from VHA administrative data, we were unable to capture utilization that occurred outside of the VA. Furthermore, given that the study was focused on the VHA system, costs for healthcare delivered outside of the VHA system was not captured and thus the findings may not generalize to other healthcare delivery systems. Future studies should evaluate healthcare utilization across multiple healthcare systems or take into account provider visits that occurred outside on patients’ primary systems.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.