Abstract
Study Objectives:
Excessive daytime sleepiness is an important public health concern associated with increased morbidity and mortality. However, in the absence of sleep diagnostic testing, it is difficult to separate the independent effects of sleepiness from those of intrinsic sleep disorders such as obstructive sleep apnea (OSA). The objective of this study was to determine if excessive daytime sleepiness was independently associated with increased health care utilization among patients referred for assessment of OSA.
Design:
Cross-sectional study.
Setting/Participants:
2149 adults referred for sleep diagnostic testing between July 2005 and August 2007.
Interventions:
N/A
Measurements:
Subjective daytime sleepiness was defined as an Epworth Sleepiness Scale score ≥10. Health care use (outpatient physician visits, all-cause hospitalizations, and emergency department visits) was determined from Alberta Health and Wellness administrative databases for the 18-month period preceding their sleep study. Rates of health resource use were analyzed using negative binomial regression, with predictors of increased health care use determined using logistic regression.
Results:
Excessive daytime sleepiness was associated with an increased rate of outpatient physician visits after adjustment for demographic variables, sleep medication use, hypertension, diabetes, depression, and OSA severity (rate ratio [RR]: 1.09 (95% confidence interval [CI]: 1.01, 1.18, P = 0.02) compared to non-sleepy subjects. There was an interaction between severe OSA and sleepiness (RR: 1.22 [95% CI: 1.06, 1.41]), although OSA was not an independent predictor of health care use. Also, sleepy patients with treated depression had a lower likelihood of outpatient visits (RR: 0.95 [95% CI: 0.86, 1.05]). Finally, sleepiness was an independent predictor of increased health care use for outpatient physician visits (odds ratio [OR]: 1.25 [95% CI: 1.00, 1.57; P = 0.048]) and all-cause hospitalizations (OR: 3.94 [95% CI: 1.03, 15.04; P = 0.046]).
Conclusions:
Excessive daytime sleepiness is associated with increased health care utilization among patients referred for assessment of OSA. Further investigation is required to determine whether the findings are related to direct effects of sleepiness, or in part, to interactions with other comorbidity such as OSA.
Citation:
Ronksley PE; Hemmelgarn BR; Heitman SJ; Flemons WW; Ghali WA; Manns B; Faris P; Tsai WH. Excessive daytime sleepiness is associated with increased health care utilization among patients referred for assessment of OSA. SLEEP 2011;34(3):363-370.
Keywords: Sleep, sleepiness, obstructive sleep apnea, chronic disease
INTRODUCTION
Excessive daytime sleepiness is an important public health concern and has been associated with increased morbidity and mortality.1–3 Sleepiness is also commonly observed among subjects with obstructive sleep apnea (OSA).4,5 Moreover, in the setting of sleep disordered breathing, hypersomnolence may identify specific subgroups at risk for poor health outcomes.6–9 However, in the absence of sleep diagnostic testing, it is difficult to separate the independent effects of sleepiness from those of intrinsic sleep disorders such as OSA; the latter may be associated with increased cardiovascular morbidity, metabolic dysfunction, and increased health care utilization.10–14
A limited number of prior studies have shown an association between sleep disorders, other than OSA, and increased healthcare utilization.15–18 However, many of these cross-sectional studies are based on survey response and were unable to exclude confounding by coexisting OSA. The Sleep Heart Health Study evaluated 6,440 patients in a community-based sample and reported a relationship between subjective daytime sleepiness, inadequate sleep time, and objective measures of sleep disordered breathing and health care utilization.15 However, this study employed an indirect measure of health care utilization through the use of a modified chronic disease score, which has been associated with increased health care use.
To our knowledge, no study has assessed the relationship between excessive daytime sleepiness and health care utilization using validated and objective measures of both the exposure (excessive daytime sleepiness) and outcome (health care use from administrative data sources). The objectives of this study therefore were to: (1) determine if excessive daytime sleepiness is associated with increased health care utilization, and (2) identify determinants of increased health care utilization among patients referred for assessment of OSA.
METHODS
Study Population
This project was part of a related study investigating the effect of continuous positive airway pressure (CPAP) on health care utilization among patients with obstructive sleep apnea (OSA). We included all patients 18 years of age and older who were referred for sleep diagnostic testing at either the Foothills Medical Centre or private respiratory care companies within the Calgary Health Region, Alberta, Canada, (catchment population ˜1.3 million) between July 2005 and August 2007. These patients were referred by their family doctor because of a suspected sleep disorder, including OSA. Over the recruitment period, nearly all sleep diagnostic testing was conducted in these facilities. We excluded non-Alberta residents, patients with previously diagnosed OSA, and those who had prior sleep diagnostic testing (polysomnography or ambulatory monitoring).
Study Variables
Epworth Sleepiness Scale
Excessive daytime sleepiness was determined using the Epworth Sleepiness Scale (ESS), an 8-item self-administered questionnaire validated in obstructive sleep apnea, narcolepsy, and idiopathic hypersomnia.19,20 Excessive daytime sleepiness was defined using a standard cut-point of ESS ≥ 10.21 An elevated ESS has been associated with diabetes in patients with severe OSA8 and has also been identified as a predictor of CPAP compliance and treatment response to CPAP in patients with OSA and hypertension.22–24
Health care utilization
Using the patient's unique Provincial Health Number (PHN), the study population was linked to 3 government health administrative databases in Alberta: the hospitalization discharge database, physician claims database, and the Ambulatory Care Classification System (ACCS) database to identify emergency department visits. For all subjects, hospitalization, physician claim, and emergency department visit information was obtained for an 18-month period prior to their sleep diagnostic testing. The hospital inpatient data source contains details regarding hospitalizations including admission date, discharge date, length of stay, 25 diagnostic codes (ICD-10), and 10 procedure codes for each admission. The physician claims registry contains information on physician services including dates and location of the visits, diagnostic codes (ICD-9-CM), and provider specialty. The ACCS database contains records on day surgeries, day procedures, and emergency room visits, including 16 diagnostic codes (ICD-10) and 10 procedure codes. All 3 diagnostic coding fields were used within the physician claims data, and all 25 and 16 diagnostic codes within inpatient hospitalization data and ACCS, respectively. The date of the sleep study was used to define the index date.
We assessed cumulative health care use incurred during the 18-month period prior to sleep diagnostic testing. As many sleep related disorders are chronic, any associated change in health care use was likely present many years prior to actual diagnosis. Therefore, assessing health care use during this defined period is reasonable. Health care use was defined as the cumulative sum (counts) of each of outpatient physician visits, all-cause hospitalizations, and emergency department visits within the defined period of 18 months. Rates and rate ratios were calculated for each outcome variable to account for person-time, as not all patients had the same duration of exposure prior to sleep testing.
Clinical characteristics and comorbidity
Baseline clinical and demographic information was collected for all participants prior to sleep diagnostic testing including age, gender, body mass index (BMI), neck circumference, and smoking status. Comorbidity was determined through the use of a questionnaire administered by trained personnel. Patients were asked to report the presence of specific comorbidities including hypertension, asthma, depression, cardiac arrhythmia, myocardial infarction, chronic obstructive pulmonary disease (COPD), diabetes, heart failure, and stroke. OSA was diagnosed by ambulatory monitoring or polysomnography.
Validated algorithms using administrative data were also employed to define diabetes, asthma, stroke, myocardial infarction, heart failure, and hypertension.25–30 Within the administrative datasets, the condition was considered present if the algorithm defining the condition was satisfied. Comorbidities that did not have a validated algorithm (COPD, depression, cardiac arrhythmia) were considered present if at least one diagnostic code for the condition within either the physician claims data or hospitalization data was recorded within the 2-year period prior to sleep diagnostic testing. Finally, using either self-report or administrative data, an enhanced measure of comorbidity was developed. This method has been shown to have face validity and provide clinically meaningful trends in the prevalence of comorbidity among this population (Table 1).31
Table 1.
ICD-9-CM and ICD-10 codes to define comorbidity among patients referred for assessment of OSA
| Comorbidities | Authors | Algorithm | ICD-10 diagnostic codes | ICD-9-CM diagnostic codes | Sensitivity | Specificity | PPV |
|---|---|---|---|---|---|---|---|
| Hypertension (with and without complication) | Tu et al.30 | 2 physician claims in 3 years | I10.x, I11.x-I13.x, I15.x | 401.x, 402.x-405.x | 73% | 95% | 87% |
| Diabetes (with and without complication) | Hux et al.26 | 1 hospitalization or 2 physician claims in 2 years | E10.0-E10.9, E11.0-E11.9, E12.0-E12.9, E13.0-E13.9, E14.0-E14.9 | 250.0-250.9 | 86% | 97% | 80% |
| Asthma | Huzel et al.27 | 1 or more physician claims in 2 years | J45.0, J45.1, J45.8, J45.9 | 490.0, 491.0, 492.0, 493.0 | 50.9% | 98.1% | NR |
| Myocardial infarction | Austin et al.25 | Primary discharge diagnosis of AMI in hospitalization database | I21.x, I22.x, I25.2 | 410.x | 88.8% | 92.8% | 88.5% |
| Congestive heart failure | Lee et al.29 | Primary discharge diagnosis of CHF in hospitalization database | I09.9, I11.0, I13.0, I13.2, I25.5, I42.0, I42.5-42.9, I43.x, I50.x, P29.0 | 428.x | NR | NR | 94.3% |
| Cerebrovascular accident/transient ischemic attack | Kokotailo and Hill28 | Primary discharge diagnosis of stroke in hospitalization database | H34.1, I63.x, I64.x, I61.x, I60.x, G45.x | 362.3, 430.x, 431.x, 433.x1, 434.x1, 435.x, 436 | 67% | 97% | 84% |
| COPD | No validated algorithm | J44 | 491.21, 493.2, 496 | NR | NR | NR | |
| Depression | No validated algorithm | F20.4, F31.3-F31.5, F32.x, F33.x, F34.1, F41.2, F43.2 | 296.2, 296.3, 296.5, 300.4, 309.x, 311 | NR | NR | NR | |
| Cardiac arrhythmia | No validated algorithm | I44.1-I44.3, I45.6, I45.9, I47.x-I49.x, R00.0, R00.1, R00.8, T82.1, Z45.0, Z95.0 | 426.0, 426.1, 426.7, 426.9, 426.10, 426.12, 427.0-427.4, 427.6-427.9, 785.0, 996.01, 996.04, V45.0, V53.3 | NR | NR | NR | |
NR, Not reported
Obstructive sleep apnea
OSA was diagnosed by ambulatory monitoring in the patients' home using the Remmers Sleep Recorder, or by polysomnography in the sleep laboratory at Foothills Medical Centre. Within the Calgary Health Region, virtually all patients undergo ambulatory monitoring as their diagnostic test to investigate possible OSA. Polysomnography at the Foothill Medical Sleep Centre is used primarily in patients with complex breathing disorders or for those patients suspected to have a non-breathing related cause for excessive daytime sleepiness. Any patient who underwent polysomnography following an ambulatory test was classified in the ambulatory monitoring group.
OSA severity was defined by the respiratory disturbance index (RDI). The RDI was defined as the number of apneas and hypopneas per hour of sleep. Apnea was defined as a cessation of airflow ≥ 10 sec. Hypopnea was defined as an abnormal respiratory event ≥ 10 sec, with ≥ 30% reduction in thorocoabdominal movement or airflow compared to baseline, and associated with ≥ 4% oxygen desaturation. OSA severity categories included: no OSA (RDI < 5/h), mild OSA (RDI = 5–14.9/h), moderate OSA (RDI = 15–29.9/h), and severe OSA (RDI ≥ 30/h). This classification system is well accepted in clinical practice and in medical literature.32,33
Polysomnography
Polysomnographic data were recorded by a computerized system (Sandman Elite Version 8.0, Nellcor Puritan Bennett (Melville) Ltd, Kanata, Ontario, Canada). This included a standardized montage: 3 electroencephalographic channels (C4/A1, C3/A2, O1/A2), bilateral electro-oculograms (EOG), submental electromyogram (EMG), bilateral leg EMGs, and electrocardiography (ECG). Airflow was measured using a nasal pressure transducer (Braebon Medical Corp, Ontario, Canada). Chest and abdominal movement were assessed by inductance plethysmography (Respitrace Ambulatory Monitoring, Ardsley, NY, USA), and oxygen saturation was recorded by oximetry (Nellcor Pulse Oximeter, Nellcor Puritan Bennett, [Melville] Ltd, Kanata, Ontario, Canada).
Ambulatory monitoring
The Remmers Sleep Recorder (SagaTech Electronics Ltd, Calgary, Canada) is an ambulatory monitor that records snoring, oxygen saturation, respiratory airflow (by monitoring nasal pressure), and body position. The RDI is derived from automated analysis of the oximetry signal, using a 4% desaturation threshold. This algorithm uses both shape and magnitude of oxygen desaturation to score respiratory events.34 Ambulatory studies are manually reviewed by the interpreting physician with the flow signal being used for quality assurance purposes. Ambulatory studies are repeated if there are discrepancies between the automatically scored respiratory events and the airflow channel. This monitor has excellent agreement with the polysomnographically determined apnea-hypopnea index (AHI).35 It has also been validated as a clinical management tool.34,36
Statistical Analysis
Patient characteristics were described using mean and standard deviation (SD) as well as median and interquartile range (IQR) as appropriate. Means and proportions were compared using unpaired t-tests and χ2 tests, respectively. Absolute measures of health care utilization were reported as median values per year. The Wilcoxon rank sum test was used to compare skewed variables. Baseline characteristics were stratified by daytime sleepiness (ESS < 10 and ESS ≥ 10).
The association between daytime sleepiness and health care use was determined by Poisson regression. Absolute rates and rate ratios (RR) were calculated for each measure of health care utilization using patients with no daytime sleepiness (ESS < 10) as the reference group. Daytime sleepiness was modeled as a dichotomous rather than continuous variable based on available clinical criteria to define excessive daytime sleepiness and for ease of interpretation.21 Initially, a Poisson Log-Linear regression model was fit for each dependent variable of healthcare utilization: outpatient physician visits, all-cause hospitalizations, and emergency department visits. To determine if over-dispersion was present, the deviance goodness-of-fit test was performed. If present, a negative binomial regression model was used to determine the RRs. Regression models were derived using backward selection techniques, and included relevant interaction terms (ESS × depression, ESS × diabetes, and ESS × OSA categories [mild, moderate, severe]). The fit of each model was also assessed by the likelihood ratio test. A similar analysis was performed to assess the association between OSA severity and health care use. OSA was modelled by severity category (mild, moderate, severe) using patients with no OSA as the referent group.
For the outcome “increased health care use”, we defined increased use based on the distribution of the count data for each dependent variable, with the upper quartile of use employed to define the outcome. Logistic regression was used to determine predictors of increased health care use, including ESS (score ≥ 10), OSA presence (≥ 5 events/h) or OSA severity (mild, moderate, severe), age (≥ 65 years), sex, BMI (≥ 30 kg/m2), comorbidities (present or absent), and nocturnal oxygen saturation profile (≥ 12% of the total sleep time spent below 90% oxygen saturation). For all statistical tests, P < 0.05 was considered statistically significant. All analyses were performed using STATA 10.0 (Statacorp, College Station, TX). This study was approved by the Ethics Review Board of the University of Calgary.
RESULTS
From July 2005 to August 2007, 2295 patients were referred for sleep diagnostic testing, of whom 78 (3.4%) patients refused to participate. An additional 42 (1.8%) patients lived outside Alberta and were therefore excluded from the study because of our inability to study the health care utilization outcome. Of the remaining 2175 patients, 26 (1.2%) were excluded because they were not present in the Alberta Health and Wellness registry, for a final study population of 2149 (Figure 1). Baseline subject characteristics by presence of excessive daytime sleepiness are presented in Table 2. Overall, 66.0% of subjects had excessive daytime sleepiness (ESS ≥ 10); these patients had a higher RDI, BMI, greater neck circumference, and were more likely to have diabetes and depression than subjects without excessive daytime sleepiness. Based on RDI criteria, 432 (20.1%) patients did not have OSA, 738 (34.3%) had mild OSA, 443 (20.6%) had moderate OSA, and 536 (24.9%) had severe OSA.
Figure 1.
Patient flow diagram
Table 2.
Baseline participant characteristics
| Variable | All (n = 2149) | ESS < 10 (n = 731) | ESS ≥ 10 (n = 1418) | P value* |
|---|---|---|---|---|
| Male, n (%) | 1346 (62.6) | 456 (62.4) | 890 (62.8) | 0.86 |
| Age (y), mean (SD) | 50.1 (12.9) | 50.3 (13.1) | 50.1 (12.8) | 0.72 |
| Body mass index, kg/m2 median (IQR) | 31.3 (27.3-36.6) | 30.1 (26.7-35.5) | 31.8 (27.5-37.1) | < 0.001 |
| Current smoker, n (%) | 354 (16.5) | 114 (15.6) | 240 (16.9) | 0.43 |
| Neck circumference, inches mean (SD) | 16.0 (1.9) | 15.8 (1.8) | 16.1 (1.9) | 0.003 |
| ESS, mean (SD) | 11.3 (5.4) | 5.9 (2.3) | 14.8 (3.5) | < 0.001 |
| RDI, events/hmedian (IQR) | 13.2 (6.0-29.9) | 11.7 (6.0-24.6) | 14.0 (6.0-32.9) | 0.006 |
| Use of sleep medication, n (%) | 191 (8.9) | 63 (8.6) | 128 (9.0) | 0.75 |
| Diabetes, n (%) | 313 (14.6) | 76 (10.4) | 237 (16.7) | < 0.001 |
| Depression, n (%) | 790 (36.8) | 235 (32.1) | 555 (39.1) | 0.0014 |
| Hypertension, n (%) | 929 (43.2) | 312 (42.7) | 617 (43.5) | 0.71 |
ESS (Epworth Sleepiness Scale)
P values for comparison of ESS categories.
Among the 2149 patients, the majority (98.7%) had at least one outpatient physician visit (regardless of specialty) within the 18-month period prior to sleep diagnostic testing. After accounting for duration of follow-up, the median number of visits per year (IQR) was 10.67 (6.00–18.01) (Table 3). There was a significant difference in the median number of visits per year (IQR) for patients with excessive daytime sleepiness compared to those without (10.67 [IQR: 6.00–18.68] vs. 10.01 [IQR: 5.34–16.68]; P = 0.005). Only 296 (13.8%) subjects had one or more hospitalizations within the 18-month period prior to sleep testing, and 768 (35.7%) had at least one emergency department visit within the same time frame. Among subjects with at least one hospitalization or emergency department visit, there was no difference in the median number of hospitalizations or emergency department visits per year by excessive daytime sleepiness status (P = 0.83, P = 0.57, respectively).
Table 3.
Absolute rates of health care utilization by level of sleepiness
| All (n = 2149) | ESS < 10 (n = 731) | ESS ≥ 10 (n = 1418) | P value†† | |
|---|---|---|---|---|
| Total outpatient physician visits | 10.67 | 10.01 | 10.67 | 0.005 |
| Median visits per year (IQR) | (6.00–18.01) | (5.34–16.68) | (6.00–18.68) | |
| Hospitalizations* | 0.67 | 0.67 | 0.67 | 0.83 |
| Median visits per year (IQR) | (0.67–1.33) | (0.67–1.33) | (0.67–1.33) | |
| Emergency room visits† | 0.67 | 0.67 | 0.67 | 0.57 |
| Median visits per year (IQR) | (0.67–2.00) | (0.67–1.33) | (0.67–2.00) |
Analysis limited to those with ≥ one hospitalization within the 18-month period prior to sleep testing (n = 296)
Analysis limited to those with ≥ one emergency room visit within 18-month period prior to sleep testing (n = 768)
P values for comparison of ESS categories
The association between excessive daytime sleepiness and various measures of health care utilization are presented in Table 4. Using a negative binomial regression model, the unadjusted model showed that the rate of total outpatient visits was higher for sleepy patients than non-sleepy patients (RR: 1.08 [95% CI: 1.01, 1.16, P = 0.02]). After adjustment for age, sex, BMI, OSA severity, level of hypoxemia, use of sleeping medications, hypertension, diabetes, depression, and relevant interactions, this relationship remained (RR: 1.09 [95% CI: 1.01, 1.18, P = 0.02]). A significant interaction between sleepiness and depression was observed (P = 0.013). When separate models were developed for each stratum, sleepy patients with associated depression had a lower likelihood of outpatient visits (RR: 0.95 [95% CI: 0.86, 1.05]), whereas those with no underlying depression had an increased risk of outpatient visits (RR: 1.11 [95% CI: 1.02, 1.19]).
Table 4.
Association between daytime sleepiness and measures of health care utilization
| Unadjusted Model |
Multivariate Adjusted Model* |
|||
|---|---|---|---|---|
| RR (95% CI) | P value | RR (95% CI) | P value | |
| Total outpatient physician visits | 1.08 (1.01, 1.16) | 0.02 | RDI ≥ 30 events/h: 1.22 (1.06, 1.41) | 0.01 |
| RDI < 30 events/h: 1.02 (0.91, 1.14) | 0.77 | |||
| Depression: 0.95 (0.86, 1.05) | 0.34 | |||
| No depression: 1.11 (1.02, 1.19) | 0.01 | |||
| Hospitalizations | 1.12 (0.90, 1.39) | 0.31 | 1.13 (0.90, 1.41) | 0.29 |
| Emergency room visits | 1.08 (0.93, 1.25) | 0.30 | 1.04 (0.90, 1.21) | 0.59 |
Negative binomial regression models adjusted for age, sex, BMI, OSA severity, level of hypoxemia, use of sleeping medications, hypertension, diabetes, depression, depression × sleepiness, OSA severity × sleepiness
Similar negative binomial regression models were developed for all-cause hospitalizations and emergency room visits. Although median values were no different between sleepy and non-sleepy subjects (Table 3), when modeled using negative binomial regression, patients with excessive daytime sleepiness had slightly higher unadjusted rates of hospitalizations and emergency department visits than those with no sleepiness (RR: 1.12 [95% CI: 0.90, 1.39, P = 0.31] and RR: 1.08 [95% CI: 0.93, 1.25, P = 0.30], respectively) (Table 4). These estimates remained nonsignificant after adjustment for demographics and comorbidity.
A similar analysis was performed to assess the association between OSA severity and health care use. Table 5 shows the absolute rates of outpatient visits for mild, moderate, and severe OSA, stratified by sleepiness status. Using a fully adjusted negative binomial regression model, there was no association between OSA severity and likelihood of outpatient visits (RR [95% CI]) for mild OSA: 1.00 [0.92, 1.09], moderate OSA: 1.06 [0.96, 1.16], severe OSA: 1.04 [0.95, 1.14]). However, a significant interaction was observed between sleepiness and severe OSA (P = 0.03). Sleepy subjects with severe OSA (RDI ≥ 30 events/h) used more outpatient visits (RR: 1.22 [95% CI: 1.06, 1.41]) than sleepy subjects without severe OSA (RR: 1.02 [95% CI: 0.91, 1.14]) (Table 4). Finally, OSA severity was not associated with higher rates of all-cause hospitalizations or emergency department visits.
Table 5.
Absolute rates of outpatient physician visits* by OSA severity (stratified by sleepiness status)
| Total (n = 2149) | No OSA (n = 432) | Mild OSA (n = 738) | Moderate OSA (n = 443) | Severe OSA (n = 536) | |
|---|---|---|---|---|---|
| Daytime sleepiness (ESS ≥ 10) | 10.67 (6.00–18.68) | 9.34 (6.00–17.35) | 10.67 (5.67–18.01) | 11.34 (6.67–19.35) | 11.34 (7.00–19.35) |
| No daytime sleepiness (ESS < 10) | 10.01 (5.34–16.68) | 9.34 (4.67–16.68) | 9.34 (4.67–16.68) | 11.01 (6.00–17.35) | 9.34 (6.00–15.68) |
Median visits per year (IQR)
Logistic regression was used to identify independent predictors of increased outpatient physician visits. Based on the distribution of the data, the upper quartile (≥ 18 visits/y) was used as the cut-point to define increased outpatient use. A crude logistic model showed that excessive daytime sleepiness was a predictor of increased outpatient physician visits (OR: 1.28 [95% CI: 1.04, 1.58; P = 0.021]). After adjusting for age, BMI, sex, OSA presence, nocturnal oxygen saturation profile, and comorbidity, sleepiness was still an independent predictor of increased outpatient use (OR: 1.25 [95% CI: 1.00, 1.57; P = 0.048]). A number of other predictors of increased outpatient visits were identified including female sex, older age, hypoxemia, and comorbidities (including hypertension, depression, and diabetes) (Table 6).
Table 6.
Predictors of increased outpatient physician visits using multivariate logistic regression
| Predictors | Odds Ratio (95% CI) | P-value |
|---|---|---|
| Excessive daytime sleepiness (ESS ≥ 10) | 1.25 (1.00, 1.57) | 0.048 |
| Age (≥ 65) | 1.82 (1.36, 2.43) | < 0.001 |
| BMI (≥ 30 kg/m2) | 0.96 (0.76, 1.21) | 0.74 |
| Female sex | 2.25 (1.82, 2.78) | < 0.001 |
| *OSA (RDI ≥ 5 events/h) | 1.01 (0.71, 1.44) | 0.95 |
| †Mild | 1.08 (0.81, 1.45) | 0.60 |
| †Moderate | 1.05 (0.75, 1.47) | 0.79 |
| †Severe | 0.94 (0.66, 1.34) | 0.72 |
| % TST spent < 90% (≥ 12%) | 1.33 (1.06, 1.67) | 0.014 |
| Hypertension | 1.58 (1.24, 2.02) | < 0.001 |
| Diabetes | 2.14 (1.57, 2.93) | < 0.001 |
| Depression | 2.75 (2.17, 3.48) | < 0.001 |
Results for OSA when modeled as a dichotomous outcome (Covariate estimates presented are based on the model with OSA as a dichotomous outcome).
Results for OSA when modeled as a categorical outcome (Covariate estimates within the multivariate model did not change appreciably from those obtained when OSA was modeled as a dichotomous variable, and were therefore not reported).
Similar models were developed to identify predictors of increased hospitalizations (≥ 1.3 visits/y) and emergency department visits (≥ 2 visits/y). Excessive daytime sleepiness was also found to be a significant independent predictor of increased hospitalizations in multivariate logistic models (OR: 3.94 [95% CI: 1.03, 15.04; P = 0.046]) but not emergency department visits (OR: 0.83 [95% CI: 0.52, 1.34; P = 0.45]).
DISCUSSION
In this large clinic-based cohort of subjects referred for assessment of OSA, we found that subjective daytime sleepiness, as defined by an ESS ≥ 10, was associated with increased rates of outpatient physician visits. This association remained significant even after adjustment for age, BMI, sex, comorbidity, sleep medication use, and OSA severity. Furthermore, when analyzed using logistic regression, sleepiness was an independent predictor of increased physician visits and all-cause hospitalizations, as defined by the upper quartile of health care utilization.
We observed a 9% increase in outpatient physician utilization among sleepy subjects. These findings complement those reported from other large cross-sectional studies. Previous studies have demonstrated increased health care utilization in the setting of insomnia, sleep disruption, and sleep complaints.15–18 However, many of these cross-sectional studies are based on survey response and were unable to adjust for the confounding effects of comorbidity and underlying sleep disorders.
In contrast, the Sleep Heart Health Study evaluated 6,440 patients in a community-based cross-sectional sample, and reported a relationship between subjective daytime sleepiness and health care utilization using polysomnographic variables.15 Specifically, patients in the highest quartile of the ESS (≥ 11) used approximately 11% more health care services. This study employed an indirect measure of health care utilization through the use of a modified chronic disease score calculated from medication data. The modified chronic disease score has been shown to correlate with ambulatory visits, hospitalizations, and mortality, and represents the expected health care use within a defined period. However, this may underestimate the true magnitude of the relationship between sleepiness and health care use. To our knowledge, our study is the first to show an association between sleepiness and a direct measure of health care use through administrative data.
Though previous studies have reported health resource use to be high among patients with underlying sleep disorders, such as OSA,10,11,37–39 we found no such association. However, in keeping with previous studies, predictors of increased health care use such as older age, female sex, depression, and diabetes were significant; and thus provide face validity for the potential association between sleepiness and higher resource use.40–45
The lack of association of OSA with health care utilization in this study does not negate an association in specific subgroups. Severe OSA and significant comorbidity has been associated with higher rates of physician visits and hospitalizations when compared to age- and gender-matched controls.11,46 In this study, a significant interaction was observed between sleepiness and severe OSA (P = 0.03). Specifically, sleepy subject with severe OSA (RDI ≥ 30) had increased outpatient physician visits (RR: 1.22 [95% CI: 1.06, 1.41]). It is therefore possible that an elevated ESS may indicate differences in individual susceptibility to the effects of OSA.
The notion that OSA patients who are sleepy have a different risk profile than non-sleepy patients has been raised in the context of long-term cardiovascular disease, metabolic dysfunction, and response to CPAP therapy.6–9;22–24 Kapur et al. showed that daytime sleepiness modified the association between OSA and hypertension among subjects in the Sleep Heart Heath Study. Specifically, the odds of hypertension was 2.83 among subjects with severe OSA (AHI ≥ 30) who reported daytime sleepiness, compared to 1.22 for subjects with severe OSA who did not report daytime sleepiness.7 Severe OSA has also been independently associated with diabetes mellitus exclusively in patients who report excessive sleepiness.8 This evidence provides plausibility for the apparent association between excessive daytime sleepiness and increased health care use among OSA patients.
A significant interaction between sleepiness and depression was also identified. Depression was defined using self-reported information and physician claims and hospitalizations within administrative data. Unfortunately, it is difficult to determine if we are simply capturing diagnosed depression as opposed to treated depression in this cohort of referred subjects. We also assessed the relationship between sleepiness and health care utilization among depressed individuals on medications (defined as treated depression) compared to depressed individuals not on medications (untreated depression). Again, we found higher health care use among sleepy patients with untreated depression. Though we could speculate that sleepy subjects with undiagnosed or untreated depression may use more health care resources in comparison to those who are treated, it is difficult to determine if sleepiness is a surrogate marker of depression.
Our study has several limitations. Firstly, daytime sleepiness was assessed by the ESS because of its ease of administration and consistency. The ESS is a subjective measure that captures one dimension of daytime sleepiness and is also based on a single measure. Though there is the possibility of misclassification, there is no reason to assume that this would occur differentially with respect to the outcomes of interest. Unfortunately, no single technique for assessing subjective sleepiness has been consistently validated. The best objective evaluation of sleepiness, the multiple sleep latency test, is not routinely used in epidemiologic research because of limited availability and cost.
Secondly, subjects were recruited from patients referred for assessment of OSA, which raises the possibility of referral bias. However, many patients were referred to diagnostic facilities in the community by their primary care physician and were not seen by sleep specialists or by the tertiary sleep disorders center prior to testing. Consequently, our study population may be more representative of patients who present to a primary care setting than a highly specialized sleep clinic. Furthermore, absolute rates of health care utilization are lower than those observed in other sleep cohorts,38,39,47 suggesting that selection was not limited to those with significant underlying sleep disorders and comorbidity. Having said that, we acknowledge that generalizability of these results may be limited to subjects referred for assessment of OSA rather than the general population.
Given the cross-sectional nature of the study, a causal relationship between sleepiness and health care use cannot be determined. Sleepiness is a nonspecific symptom that may be secondary to a variety of etiologies. Whether the findings are a direct effect of sleepiness or an as yet undefined confounder is unclear. There is also potential for residual confounding from the measurement techniques used to defined comorbidity within this cohort. We were unable to adjust for the severity of comorbidity among subjects using administrative data. Furthermore, we were limited in the ability to provide sufficient clinical detail on the type of physician visit or reason for hospitalization that may explain reasons for higher health care use among sleepy subjects.
Excessive daytime sleepiness was associated with increased outpatient physician visits and all-cause hospitalizations in this referred population. This is the first study to find an association between excessive daytime sleepiness and an objective estimate of health care utilization. Though the relative risk of increased health care use was modest, given the high prevalence of sleepiness within this cohort and the absolute risk, the overall impact on cost to the health care system may be substantial. Further investigation is required to determine whether the findings are related to direct effects of sleepiness, or in part, to interactions with other comorbidity such as OSA.
DISCLOSURE STATEMENT
This was not an industry supported study. Dr. Tsai has participated in speaking engagements for GlaxoSmithKline and Boehringer-Ingelheim. The other authors have indicated no financial conflicts of interest.
ACKNOWLEDGMENTS
We thank Respiratory Homecare Solutions, Respiratory Wellness Centre, VitalAire, and Medigas for participation as recruitment sites for this study. We also thank Marie Flexhaug for her diligent work in ensuring medication accuracy.
This study was supported by operating grants from the Alberta Heritage Foundation for Medical Research and the Calgary Health Region. Paul Ronksley is supported by a Frederick Banting and Charles Best Canada Graduate Scholarship from the Canadian Institutes of Health Research. Dr. Hemmelgarn is supported by New Investigator Awards from the Canadian Institutes of Health Research and by Population Health Investigator Awards from the Alberta Heritage Foundation for Medical Research. Dr. William Ghali is supported by a Canada Research Chair in Health Services Research and by a Senior Health Scholar Award from the Alberta Heritage Foundation for Medical Research.
REFERENCES
- 1.Elwood P, Hack M, Pickering J, Hughes J, Gallacher J. Sleep disturbance, stroke, and heart disease events: evidence from the Caerphilly cohort. J Epidemiol Community Health. 2006;60:69–73. doi: 10.1136/jech.2005.039057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Newman AB, Spiekerman CF, Enright P, et al. The Cardiovascular Health Study Research Group. Daytime sleepiness predicts mortality and cardiovascular disease in older adults. J Am Geriatr Soc. 2000;48:115–23. doi: 10.1111/j.1532-5415.2000.tb03901.x. [DOI] [PubMed] [Google Scholar]
- 3.Qureshi AI, Giles WH, Croft JB, Bliwise DL. Habitual sleep patterns and risk for stroke and coronary heart disease: a 10-year follow-up from NHANES I. Neurology. 1997;48:904–11. doi: 10.1212/wnl.48.4.904. [DOI] [PubMed] [Google Scholar]
- 4.Gottlieb DJ, Whitney CW, Bonekat WH, et al. Relation of sleepiness to respiratory disturbance index: the Sleep Heart Health Study. Am J Respir Crit Care Med. 1999;159:502–7. doi: 10.1164/ajrccm.159.2.9804051. [DOI] [PubMed] [Google Scholar]
- 5.Kapur VK, Baldwin CM, Resnick HE, Gottlieb DJ, Nieto FJ. Sleepiness in patients with moderate to severe sleep-disordered breathing. Sleep. 2005;28:472–7. doi: 10.1093/sleep/28.4.472. [DOI] [PubMed] [Google Scholar]
- 6.Choi JB, Nelesen R, Loredo JS, et al. Sleepiness in obstructive sleep apnea: a harbinger of impaired cardiac function? Sleep. 2006;29:1531–6. doi: 10.1093/sleep/29.12.1531. [DOI] [PubMed] [Google Scholar]
- 7.Kapur VK, Resnick HE, Gottlieb DJ. Sleep disordered breathing and hypertension: does self-reported sleepiness modify the association? Sleep. 2008;31:1127–32. [PMC free article] [PubMed] [Google Scholar]
- 8.Ronksley PE, Hemmelgarn BR, Heitman SJ, et al. Obstructive sleep apnoea is associated with diabetes in sleepy subjects. Thorax. 2009;64:834–9. doi: 10.1136/thx.2009.115105. [DOI] [PubMed] [Google Scholar]
- 9.Vgontzas AN, Papanicolaou DA, Bixler EO, Kales A, Tyson K, Chrousos GP. Elevation of plasma cytokines in disorders of excessive daytime sleepiness: role of sleep disturbance and obesity. J Clin Endocrinol Metab. 1997;82:1313–6. doi: 10.1210/jcem.82.5.3950. [DOI] [PubMed] [Google Scholar]
- 10.Banno K, Manfreda J, Walld R, Delaive K, Kryger MH. Healthcare utilization in women with obstructive sleep apnea syndrome 2 years after diagnosis and treatment. Sleep. 2006;29:1307–11. doi: 10.1093/sleep/29.10.1307. [DOI] [PubMed] [Google Scholar]
- 11.Kryger MH, Roos L, Delaive K, Walld R, Horrocks J. Utilization of health care services in patients with severe obstructive sleep apnea. Sleep. 1996;19:S111–6. doi: 10.1093/sleep/19.suppl_9.s111. [DOI] [PubMed] [Google Scholar]
- 12.Nieto FJ, Young TB, Lind BK, et al. Association of sleep-disordered breathing, sleep apnea, and hypertension in a large community-based study. Sleep Heart Health Study. JAMA. 2000;283:1829–36. doi: 10.1001/jama.283.14.1829. [DOI] [PubMed] [Google Scholar]
- 13.Reichmuth KJ, Austin D, Skatrud JB, Young T. Association of sleep apnea and type II diabetes: a population-based study. Am J Respir Crit Care Med. 2005;172:1590–5. doi: 10.1164/rccm.200504-637OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM, Mohsenin V. Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med. 2005;353:2034–41. doi: 10.1056/NEJMoa043104. [DOI] [PubMed] [Google Scholar]
- 15.Kapur VK, Redline S, Nieto FJ, Young TB, Newman AB, Henderson JA. The relationship between chronically disrupted sleep and healthcare use. Sleep. 2002;25:289–96. [PubMed] [Google Scholar]
- 16.Kuppermann M, Lubeck DP, Mazonson PD, et al. Sleep problems and their correlates in a working population. J Gen Intern Med. 1995;10:25–32. doi: 10.1007/BF02599573. [DOI] [PubMed] [Google Scholar]
- 17.Novak M, Mucsi I, Shapiro CM, Rethelyi J, Kopp MS. Increased utilization of health services by insomniacs -an epidemiological perspective. J Psychosom Res. 2004;56:527–36. doi: 10.1016/j.jpsychores.2004.02.007. [DOI] [PubMed] [Google Scholar]
- 18.Simon GE, VonKorff M. Prevalence, burden, and treatment of insomnia in primary care. Am J Psychiatry. 1997;154:1417–23. doi: 10.1176/ajp.154.10.1417. [DOI] [PubMed] [Google Scholar]
- 19.Johns MW. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep. 1991;14:540–5. doi: 10.1093/sleep/14.6.540. [DOI] [PubMed] [Google Scholar]
- 20.Johns MW. Reliability and factor analysis of the Epworth Sleepiness Scale. Sleep. 1992;15:376–81. doi: 10.1093/sleep/15.4.376. [DOI] [PubMed] [Google Scholar]
- 21.Johns M, Hocking B. Daytime sleepiness and sleep habits of Australian workers. Sleep. 1997;20:844–9. doi: 10.1093/sleep/20.10.844. [DOI] [PubMed] [Google Scholar]
- 22.Barbe F, Mayoralas LR, Duran J, et al. Treatment with continuous positive airway pressure is not effective in patients with sleep apnea but no daytime sleepiness. a randomized, controlled trial. Ann Intern Med. 2001;134:1015–23. doi: 10.7326/0003-4819-134-11-200106050-00007. [DOI] [PubMed] [Google Scholar]
- 23.Montserrat JM, Garcia-Rio F, Barbe F. Diagnostic and therapeutic approach to nonsleepy apnea. Am J Respir Crit Care Med. 2007;176:6–9. doi: 10.1164/rccm.200606-795PP. [DOI] [PubMed] [Google Scholar]
- 24.Morrell MJ, Finn L, Kim H, Peppard PE, Badr MS, Young T. Sleep fragmentation, awake blood pressure, and sleep-disordered breathing in a population-based study. Am J Respir Crit Care Med. 2000;162:2091–6. doi: 10.1164/ajrccm.162.6.9904008. [DOI] [PubMed] [Google Scholar]
- 25.Austin PC, Daly PA, Tu JV. A multicenter study of the coding accuracy of hospital discharge administrative data for patients admitted to cardiac care units in Ontario. Am Heart J. 2002;144:290–6. doi: 10.1067/mhj.2002.123839. [DOI] [PubMed] [Google Scholar]
- 26.Hux JE, Ivis F, Flintoft V, Bica A. Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm. Diabetes Care. 2002;25:512–6. doi: 10.2337/diacare.25.3.512. [DOI] [PubMed] [Google Scholar]
- 27.Huzel L, Roos LL, Anthonisen NR, Manfreda J. Diagnosing asthma: the fit between survey and administrative database. Can Respir J. 2002;9:407–12. doi: 10.1155/2002/921497. [DOI] [PubMed] [Google Scholar]
- 28.Kokotailo RA, Hill MD. Coding of stroke and stroke risk factors using international classification of diseases, revisions 9 and 10. Stroke. 2005;36:1776–81. doi: 10.1161/01.STR.0000174293.17959.a1. [DOI] [PubMed] [Google Scholar]
- 29.Lee DS, Donovan L, Austin PC, et al. Comparison of coding of heart failure and comorbidities in administrative and clinical data for use in outcomes research. Med Care. 2005;43:182–8. doi: 10.1097/00005650-200502000-00012. [DOI] [PubMed] [Google Scholar]
- 30.Tu K, Campbell NRC, Chen ZL, Cauch-Dudek KJ, McAlister FA. Accuracy of administrative databases in identifying patients with hypertension. Open Med. 2007;1:e3–e5. [PMC free article] [PubMed] [Google Scholar]
- 31.Ronksley PE, Tsai WH, Quan H, Faris P, Hemmelgarn BR. Data enhancement for co-morbidity measurement among patients referred for sleep diagnostic testing: an observational study. BMC Med Res Methodol. 2009;9:50. doi: 10.1186/1471-2288-9-50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep. 1999;22:667–89. [PubMed] [Google Scholar]
- 33.American Academy of Sleep Medicine. International classification of sleep disorders. 2nd ed. Westchester, IL: American Academy of Sleep Medicine; 2005. [Google Scholar]
- 34.Mulgrew AT, Fox N, Ayas NT, Ryan CF. Diagnosis and initial management of obstructive sleep apnea without polysomnography: a randomized validation study. Ann Intern Med. 2007;146:157–66. doi: 10.7326/0003-4819-146-3-200702060-00004. [DOI] [PubMed] [Google Scholar]
- 35.Vazquez JC, Tsai WH, Flemons WW, et al. Automated analysis of digital oximetry in the diagnosis of obstructive sleep apnoea. Thorax. 2000;55:302–7. doi: 10.1136/thorax.55.4.302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Whitelaw WA, Brant RF, Flemons WW. Clinical usefulness of home oximetry compared with polysomnography for assessment of sleep apnea. Am J Respir Crit Care Med. 2005;171:188–93. doi: 10.1164/rccm.200310-1360OC. [DOI] [PubMed] [Google Scholar]
- 37.Kapur V, Blough DK, Sandblom RE, et al. The medical cost of undiagnosed sleep apnea. Sleep. 1999;22:749–55. doi: 10.1093/sleep/22.6.749. [DOI] [PubMed] [Google Scholar]
- 38.Tarasiuk A, Greenberg-Dotan S, Brin YS, Simon T, Tal A, Reuveni H. Determinants affecting health-care utilization in obstructive sleep apnea syndrome patients. Chest. 2005;128:1310–4. doi: 10.1378/chest.128.3.1310. [DOI] [PubMed] [Google Scholar]
- 39.Tarasiuk A, Greenberg-Dotan S, Simon-Tuval T, Oksenberg A, Reuveni H. The effect of obstructive sleep apnea on morbidity and health care utilization of middle-aged and older adults. J Am Geriatr Soc. 2008;56:247–54. doi: 10.1111/j.1532-5415.2007.01544.x. [DOI] [PubMed] [Google Scholar]
- 40.Bertakis KD, Azari R, Helms LJ, Callahan EJ, Robbins JA. Gender differences in the utilization of health care services. J Fam Pract. 2000;49:147–52. [PubMed] [Google Scholar]
- 41.Egede LE, Zheng D, Simpson K. Comorbid depression is associated with increased health care use and expenditures in individuals with diabetes. Diabetes Care. 2002;25:464–70. doi: 10.2337/diacare.25.3.464. [DOI] [PubMed] [Google Scholar]
- 42.Koopmans GT, Lamers LM. Gender and health care utilization: the role of mental distress and help-seeking propensity. Soc Sci Med. 2007;64:1216–30. doi: 10.1016/j.socscimed.2006.11.018. [DOI] [PubMed] [Google Scholar]
- 43.Laditka SB, Mastanduno MP, Laditka JN. Health care use of individuals with diabetes in an employer-based insurance population. Arch Intern Med. 2001;161:1301–8. doi: 10.1001/archinte.161.10.1301. [DOI] [PubMed] [Google Scholar]
- 44.Natarajan S, Nietert PJ. Hypertension, diabetes, hypercholesterolemia, and their combinations increased health care utilization and decreased health status. J Clin Epidemiol. 2004;57:954–61. doi: 10.1016/j.jclinepi.2004.01.011. [DOI] [PubMed] [Google Scholar]
- 45.Starfield B, Lemke KW, Herbert R, Pavlovich WD, Anderson G. Comorbidity and the use of primary care and specialist care in the elderly. Ann Fam Med. 2005;3:215–22. doi: 10.1370/afm.307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Ronald J, Delaive K, Roos L, Manfreda J, Bahammam A, Kryger MH. Health care utilization in the 10 years prior to diagnosis in obstructive sleep apnea syndrome patients. Sleep. 1999;22:225–9. doi: 10.1093/sleep/22.2.225. [DOI] [PubMed] [Google Scholar]
- 47.Banno K, Ramsey C, Walld R, Kryger MH. Expenditure on health care in obese women with and without sleep apnea. Sleep. 2009;32:247–52. doi: 10.1093/sleep/32.2.247. [DOI] [PMC free article] [PubMed] [Google Scholar]