Relationship between Travel Time from Home to a Regional Sleep Apnea Clinic in British Columbia, Canada, and the Severity of Obstructive Sleep

Transcription

1 Relationship between Travel Time from Home to a Regional Sleep Apnea Clinic in British Columbia, Canada, and the Severity of Obstructive Sleep A. J. M. Hirsch Allen 1, Ofer Amram 2, Hamid Tavakoli 1, Fernanda R. Almeida 3, Mona Hamoda 3, and Najib T. Ayas 1,4 1 Division of Respiratory Medicine, and 3 Department of Dentistry, University of British Columbia, Vancouver, British Columbia, Canada; 2 Department of Health Geography, Simon Fraser University, Vancouver, British Columbia, Canada; and 4 Sleep Disorders Program, University Hospital, Vancouver, British Columbia, Canada Abstract Rationale: In the majority of people with obstructive sleep apnea, the disorder remains undiagnosed. This may be partly a result of inadequate access to diagnostic sleep services. We thus hypothesized that even modest travel times to a sleep clinic may delay diagnosis and reduce detection of milder disease. Objectives: We sought to determine whether travel time between an individual s home and a sleep clinic is associated with sleep apnea severity at presentation. Methods: We recruited patients referred for suspected sleep apnea to the University of British Columbia Hospital Sleep Clinic between May 2003 and July The patient s place of residence was geocoded at the postal code level. Travel times between the population-weighted dissemination areas for each patient and the sleep clinic were calculated using ArcGIS (ESRI, Redlands, CA) network analyst and the Origin Destination matrix function. All patients underwent full polysomnography. Measurements and Main Results: There were 1,275 patients; 69% were male, the mean age was 58 years. (SD = 11.9), and the mean apnea hypopnea index was 22 per hour (SD = 21.6). In the univariate model, travel time was a significant predictor of obstructive sleep apnea severity (P = 0.02). After controlling for confounders including sex, age, obesity, and education, travel time remained a significant predictor of sleep apnea severity (P, 0.01). In the multivariate model, each increase in 10 minutes of travel time was associated with an increase in the apnea-hypopnea index of 1.4 events per hour. Conclusions: For reasons that remain to be determined, travel times are associated with the severity of obstructive sleep apnea at presentation to a sleep clinic. If the results can be verified at other centers, this may help guide the geographic distribution of sleep centers within a health care system. Keywords: epidemiology; health services accessibility; sleep; geographical information systems (Received in original form September 18, 2015; accepted in final form January 5, 2016 ) Supported by Canadian Institutes of Health Research (a Sleep Disordered Breathing Team grant), a Vancouver Coastal Health Research Institute Scientist Award, and a BC Lung Association Operating grant. Author Contributions: A.J.M.H.A. contributed to the design of the study and drafting of work; O.A. contributed to the design of the study and analysis and interpretation of data; H.T. contributed to the analysis and interpretation of data; F.R.A. contributed to critical revision of the work for intellectual content; M.H. contributed to critical revision of the work for intellectual content; and N.T.A. contributed to the drafting of the work. Correspondence and requests for reprints should be addressed to Najib T. Ayas, M.D., M.P.H., University of British Columbia. Room 224, St. Paul s Hospital, 1080 Burrard Street, Vancouver, BC, V6Z 1Y6 Canada. nayas@providencehealth.bc.ca Ann Am Thorac Soc Vol 13, No 5, pp , May 2016 Copyright 2016 by the American Thoracic Society DOI: /AnnalsATS BC Internet address: Obstructive sleep apnea (OSA) is a common yet underdiagnosed respiratory disorder characterized by recurrent upper-airway obstruction during sleep (1, 2). OSA results in sleep fragmentation and repetitive hypoxemia and it is associated with a variety of adverse consequences that affect virtually every organ system (3). These consequences include excessive daytime sleepiness, reduced quality of life, neurocognitive impairment, motor vehicle crashes, occupational injuries, and cardiovascular disease (4). Treatment of OSA is highly effective in mitigating many of these adverse effects and improving health; OSA therapy is also an extremely cost-effective use of health care resources (5). Allen, Amram, Tavakoli, et al.: OSA Severity and Travel Time to the Clinic 719

2 An important issue is that less than 10 20% of individuals with OSA have received a clinical diagnosis (6), in part because access to sleep-related diagnostic services is limited in many geographic areas (7,8). In a recent study by Evans and colleagues, there were only 0.7 diagnostic sleep beds per 100,000 people in the province of British Columbia, a number that was well below the national average of two beds per 100,000 people (7). Evans and colleagues also found that geographic region was a significant predictor of sleep laboratory testing referral in Canada (7). In addition, if there are few sleep-related diagnostic centers in a region, the travel time from the patient s residence to a center may be relatively long and may reflect a barrier for patients seeking care (9). With a limited number of sleep laboratories, long travel times from a patient s home to the clinic become more likely, and the geographic location of a patient s home becomes a potential additional barrier to diagnosis and treatment. We hypothesized that even modest travel times would act as an additional barrier to polysomnography (PSG) access, further limiting the diagnosis of cases of OSA. Specifically, we hypothesized that patients who lived farther from the clinic would present with increased symptoms and more severe OSA. Methods Adult patients who were referred to the University of British Columbia (UBC) Hospital Sleep Disorders Clinic and who had a full PSG for suspected OSA between May 2003 and July 2011 were recruited to participate in this study. Ethics approval for the recruitment of patients was provided by the UBC office of research services clinical research ethics board. Consenting patients were given extensive questionnaires regarding their demographics, sleep habits, and subjective sleepiness as measured by the Epworth Sleepiness Scale (ESS), as well as an overnight PSG to objectively identify OSA and its severity according to the apnea hypopnea index (AHI). PSG is the gold standard for objectively analyzing sleep and is recommended by both the American Thoracic Society and the American Academy of Sleep Medicine. PSG was performed using conventional instrumentation and analysis according to the recommendations on syndrome definition and measurement techniques published by the American Academy of Sleep Medicine (10). Because of regionalization of resources, the UBC sleep clinic is the only sleep clinic in Vancouver. The next closest sleep clinic is located in another municipality (Richmond, BC), approximately 14 km away. Patients were excluded if they were medically unstable or had active psychiatric disease. We included only adult patients whose travel times were less than1 hour because patients who live more than 1 hour away are usually referred to the UBC sleep clinic for an unusually complex sleep problem. Each patient s place of residence was geocoded at the postal code level using Digital Mapping Technologies Inc. (DMTI, Huntington Beach, CA) geopinpoint software. DMTI is software that positions data using a geolocation process called geocoding. Geocoding is the process of converting addresses into coordinates, which can then be used to place markers on a map. DMTI route logistic road data were used, together with ArcGIS network analyst function, to calculate travel time between the postal code and the sleep clinic. ArcGIS is a geographic information system (GIS) software that is used for the purpose of spatial access analysis in health research in a multitude of contexts (11). Route logistic provides a means to calculate turn-by-turn travel time over the road network. Statistical Analysis Baseline characteristics (e.g., sex, age, education, body mass index [BMI], travel time from home to the sleep clinic, subjective sleepiness as measured by the ESS, and severity of OSA as measured by AHI) were summarized as means and proportions as appropriate. For our analysis, we used education level rather than income as a marker of socioeconomic status, because many people living within 1 hour of the sleep disorders clinic were retired. We examined the relationship between travel time and OSA severity using linear regression. Thereafter, we constructed a multivariate linear regression model. Variables were included in the final multivariate models on the basis of either consensus from the individual OSA and GIS literature or any variable with a univariate significance of less than 0.1. A P value of,0.05 was considered statistically significant. All statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Inc., Cary, NC). Results There were 1,275 patients, 69% of whom were male. The mean age was 58 years (SD = 11.9), the mean AHI was 22 events per hour (SD = 21.6), and the mean ESS was 11.5 (SD = 7.7). More than 80% of patients had OSA: 30% had mild OSA, 25% had moderate OSA and 26% had severe OSA (Table 1). The participants were generally well educated, with 50% of patients having a college or university degree and an additional 14% having at least a master s degree. Patients were also predominately overweight; more than 85% had a BMI.25 kg/m 2. The mean travel time was 20.8 minutes (SD = 11.9). We divided the patients into two groups according to whether their travel times were longer than or shorter than the mean (Table 1). These groups were then compared with t tests or Chi-squared tests and t tests for categorical and continuous variables, respectively. As can be seen in Table 1, patients who lived closer than the mean travel time to the UBC sleep clinic were more educated, had significantly less severe OSA (as measured by the AHI), were less likely to be overweight, and were slightly older compared with patients who lived farther from the sleep clinic. Patients who lived within 20 minutes of the sleep clinic had an average AHI of 18 compared with patients who lived between 20 and 40 minutes from the clinic, who had an average AHI of 23 (Figure 1). Patients who lived 20 to 40 minutes form the sleep clinic had, on average, an AHI 27% more severe than that of patients who lived within 20 minutes of the clinic. In the univariate model, travel time was a significant predictor of OSA severity (P = 0.023). Each increase of 10 minutes of travel time was associated with an increase in AHI of 1.2 events per hour. After controlling for a number of variables including sex, age, obesity, and education, travel time remained a significant predictor of OSA severity (P, 0.01) (see Table 2). In the 720 AnnalsATS Volume 13 Number 5 May 2016

3 Table 1. Baseline patient characteristics Variable Patients with Below Mean Travel Times Patients with Above Mean Travel Times P Value Total Categorical variables Sex Male 430 (68.7) 331 (68.2) 761 (68.4) Female 196 (31.3) 154 (31.8) 350 (31.6) BMI <25 kg/m (16.9) 52 (11.0) 189 (14.9).25 kg/m (83.1) 414 (89.0) 1,084 (85.1) Education* (6.1) 48 (10.0) 86 (7.7) (26.0) 151 (31.3) 314 (28.2) (52.0) 233 (48.3) 558 (50.1) (16.0) 50 (10.4) 150 (13.5) Continuous variables AHI, events/h Travel time, min N/A Age, yr ESS OSA classification No OSA (AHI, 5/h) 148 (20.4) 97 (17.6) N/A 245 (19.2) Mild OSA (AHI > 5/h and, 15/h) 230 (31.7) 151 (27.4) N/A 381 (29.9) Moderate OSA (AHI > 15/h and, 30/h) 183 (25.2) 133 (24.2) N/A 316 (24.8) Severe OSA (AHI > 30/h) 164 (22.6) 169 (30.7) N/A 333 (26.1) Definition of abbreviations: AHI = apnea hypopnea index; BMI = body mass index; ESS = Epworth Sleepiness Scale; N/A = not applicable; OSA = obstructive sleep apnea. Data are presented as No. (%) or mean 6 SD. *Education variable is categorical: 1 = less than high school diploma, 2 = high school diploma, 3 = college or university degree, 4 = master s or above. multivariate model, each increase of 10 minutes of travel time was associated with an AHI increase of 1.4 events per hour. Both univariate and multivariate models were also analyzed with subjective sleepiness (ESS) as the outcome variable. Travel time was not a significant predictor of ESS (P = 0.20) in the univariate model. After controlling for sex, age, BMI, and education, the effect of travel time on ESS remained nonsignificant (P = 0.37). A model was also produced with OSA as a dichotomous variable (i.e., moderate to severe OSA vs. no OSA to mild OSA) as part of a sensitivity analysis. The results were similar to those obtained when using OSA as a continuous outcome variable and provide little evidence of an unseen relationship among the variables studied. Discussion In our study, we found that even modestly increased travel times were significantly associated with the severity of OSA at presentation. After controlling for a variety of cofounders in the multivariate model, each additional 10 minutes of travel time was associated with an increase in AHI of 1.4 events per hour. Because of the increase in the prevalence of OSA and the inability of current health care system infrastructure to keep up with the increasing demand for diagnostic in-laboratory PSG (7), it is well documented that the majority of individuals with OSA have not been given a diagnosis (7, 9). Access to sleep clinics, and in particular diagnostic sleep laboratories, is an understudied but important metric, given these recent trends (7). Our results are consistent with the findings of other studies in the literature investigating the impact of travel times or distance on access to health care services or treatments. A study by Meden and colleagues demonstrated that those with early-stage invasive breast cancer residing at greater distances from radiation therapy treatment facilities were more likely to undergo a mastectomy than were those who lived closer to such facilities, because longer travel distances prevented patients from seeking repeat radiation treatments (12). Numerous other examples have shown that measures of spatial access, such as distance to the nearest health care facility or the density of facilities in a given area, play an important role as facilitators or barriers to health service access for the general public (13 17). However, to the best of our knowledge, this is the first study to date to examine the issue in the context of adult OSA. Our study suggests that patients who live farther from a sleep clinic tend to access the clinic when their OSA is more severe. If these results can be confirmed in future studies, they may provide evidence for the potential advantages of additional diagnostic sleep clinics. All patients with OSA should have access to diagnostic sleep clinics irrespective of the geographic location of their homes. Given the already long wait times for sleep specialist consultations (4 6 mo in western Canada [7]) and the limited number of diagnostic sleep beds, it is important that travel times to the clinics do not act as an additional barrier to access. Other measures to improve access, such as telemedicine or portable monitoring technologies, may be useful as well. Subjective sleepiness as measured using the ESS was not significantly associated with travel time to the clinic. Previous literature has shown that neither the ESS nor other measures of subjective sleepiness are Allen, Amram, Tavakoli, et al.: OSA Severity and Travel Time to the Clinic 721

4 Figure 1. Travel time vs. severity of obstructive sleep apnea (as measured using AHI). AHI = apnea hypopnea index. accurate proxies for objective sleepiness and as a result may not always act as reliable measures of OSA symptoms (18, 19). Thismayaccountinpartforthelackof association between travel times to the clinic and the ESS. Limitations Strengths of our study include the novel use of GIS technologies in the context of OSA to Table 2. Multivariate model with OSA severity (AHI) as outcome Variable Parameter Estimate show the relationship between travel times and disease severity and the use of PSG to objectively analyze sleep in a large cohort of patients. Our study, however, had a number of potential limitations. First, it was conducted only in one clinic in one geographic location, which may limit its generalizability to other health regions. Second, travel times are inherently difficult to estimate given the SE P Value Intercept Age ,0.01 Sex (0 = female, 1 = male) ,0.01 Education* BMI (. and,25 kg/m 2 ) ,0.01 Travel time (per 10 min) Definition of abbreviations: AHI = apnea hypopnea index; BMI = body mass index; OSA = obstructive sleep apnea. *Education variable is categorical; 1 = less than high school diploma, 2 = high school diploma, 3 = college or university degree, 4 = master s or above. Age, sex, education, BMI, and travel time were significant predictors of AHI. The parameter estimate (b) and the SE of the parameter are shown; for example being overweight (BMI. 25) was associated with an AHI increase of 6.67 events per hour. fact that different people use different modes of transport. Third, it is possible that over the span of recruitment, a trend toward increasing severity of OSA may have developed in patients. Fourth, although the degree of sleepiness was not associated with travel time, we suspect that the ESS may not have reflected an accurate burden of the patient s symptoms. Fifth, we cannot exclude the possibility that residual confounding may have been a partial explanation for the association that was found. Finally, we were unable to collect comorbidity data to show if patients who lived farther from the clinic were less healthy than the people who lived closer. Conclusions Moderate travel times to the clinic are associated with an increased severity of OSA when compared with shorter travel times and may thus be a barrier to OSA diagnosis. If these results can be verified in other centers, this may help guide the planning process for diagnostic centers within the health care system. Future research should also focus on the relationship between symptoms of OSA and travel times to better understand the driving force behind an individual s decision to seek care. n Author disclosures are available with the text of this article at AnnalsATS Volume 13 Number 5 May 2016

5 References 1 Ip M, Chung KF, Chan KN, Lam SP, Lee K. Previously unrecognized obstructive sleep apnea in Chinese subjects with essential hypertension. Lung 1999;177: Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993;328: Gurubhagavatula I. Consequences of obstructive sleep apnoea. Indian J Med Res 2010;131: Engleman H, Joffe D. Neuropsychological function in obstructive sleep apnoea. Sleep Med Rev 1999;3: AlGhanim N, Comondore VR, Fleetham J, Marra CA, Ayas NT. The economic impact of obstructive sleep apnea. Lung 2008;186: Young T, Evans L, Finn L, Palta M. Estimation of the clinically diagnosed proportion of sleep apnea syndrome in middle-aged men and women. Sleep 1997;20: Evans J, Skomro R, Driver H, Graham B, Mayers I, McRae L, Reisman J, Rusu C, To T, Fleetham J. Sleep laboratory test referrals in Canada: sleep apnea rapid response survey. Can Respir J 2014;21: e4 e10. 8 Flemons WW, Douglas NJ, Kuna ST, Rodenstein DO, Wheatley J. Access to diagnosis and treatment of patients with suspected sleep apnea. Am J Respir Crit Care Med 2004;169: Wang F, McLafferty S, Escamilla V, Luo L. Late-stage breast cancer diagnosis and health care access in Illinois. Prof Geogr 2008; 60: Berry RB, Budhiraja R, Gottlieb DJ, Gozal D, Iber C, Kapur VK, Marcus CL, Mehra R, Parthasarathy S, Quan SF, et al.; American Academy of Sleep Medicine: Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events.Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. J Clin Sleep Med 2012;8: Schuurman N, Crooks VA, Amram O. A protocol for determining differences in consistency and depth of palliative care service provision across community sites. Health Soc Care Community 2010;18: Meden T, St John-Larkin C, Hermes D, Sommerschield S. MSJAMA. Relationship between travel distance and utilization of breast cancer treatment in rural northern Michigan. JAMA 2002;287: Buor D. Distance as a predominant factor in the utilization of health services in the Kumasi metropolis, Ghana. GeoJournal 2002;56: McLafferty S, Grady S. Prenatal care need and access: a GIS analysis. J Med Syst 2004;28: Tanser F, Gijsbertsen B, Herbst K. Modelling and understanding primary health care accessibility and utilization in rural South Africa: an exploration using a geographical information system. Soc Sci Med 2006;63: Arcury TA, Gesler WM, Preisser JS, Sherman J, Spencer J, Perin J. The effects of geography and spatial behavior on health care utilization among the residents of a rural region. Health Serv Res 2005;40: Stock R. Distance and the utilization of health facilities in rural Nigeria. Soc Sci Med 1983;17: Olson LG, Cole MF, Ambrogetti A. Correlations among Epworth Sleepiness Scale scores, multiple sleep latency tests and psychological symptoms. J Sleep Res 1998;7: Sadeghniiat-Haghighi K, Moller HJ, Saraei M, Aminian O, Khajeh- Mehrizi A. The Epworth Sleepiness Scale for screening of the drowsy driving: comparison with the maintenance of wakefulness test in an Iranian sample of commercial drivers. Acta Med Iran 2014;52: Allen, Amram, Tavakoli, et al.: OSA Severity and Travel Time to the Clinic 723