Accidents in Obstructive Sleep Apnea Patients Treated With Nasal Continuous Positive Airway Pressure

Accidents in Obstructive Sleep Apnea Patients Treated With Nasal Continuous Positive Airway Pressure* A Prospective Study Jean Krieger, MD; Nicole Meslier, MD; Therese Lebrun, MD; Patrick Levy, MD; Fran9ois Phillip-Joet, MD; Jean-Claude Sailly, MD; Jean-Louis Racineux, MD; the Working Group ANTADIR, Paris; and CRESGE, Lille, France'

Many studies have shown a relationship between obstructive sleep apnea (OSA) and accidents, but to our knowledge, none have investigated prospectively the effects of treatment with nasal continuous positive airway pressure (CPAP). CPAP was proposed to 973 patients, of whom 893 patients actually underwent CPAP. These patients were consecutively invited to enter a prospective follow-up study including a questionnaire before treatment and after 6 and 12 months of treatment; 547 patients completed the study (153 left the study, and only partial data were available for 193). The baseline questionnaire included questions concerning accidents in the previous 12 months, asking whether patients had had an accident and, if so, whether they felt that the accident(s) were related to sleepiness, and whether the patients felt that they had had near-miss accidents due to sleepiness. The questionnaires at 6 and 12 months included the same questions referring to the previous 6 months; the accidents reported on each follow-up questionnaire were cumulated and compared with the accidents during the 1-year period before treatment. The number of patients having an accident decreased with treatment for real accidents (from 60 to 36; p
D uring recent years , emphasis was put on the role

of sleepiness in various types of accidents, including traffic, work, and dom estic accidents.I-6 Among other causes of sleepiness, sleep disorders appear to play a major role, L.2·5 •7 •8 especially sleeprelated breathing disorders,8-13 due to their high prevalence. It has been shown that treatment of obstructive *From the vVorking Group ANTADIR, Paris and CRESGE, Lille, France. Study suppmied by th e Ministere de Ia Reche rche e t d e Ia Technologie, by the Caisse Nationale d'Assurance Maladie des Travailleurs Salalies, and by the TAEMA, Pierre Medical, and SEF AM co mpanies. 1 A complete list of investigators is located in the Appendix. Manuscript received January 7, 1997; revision accepted July 30.

sleep apnea (OSA) with nasal continuous positive ainvay pressure (nCPAP) improved daytime sleepiness, 14 -18 as well as cognitive or psychometric performance including performance on a driving simulator,19-22 but surprisingly, the effects of continuous positive ai1way pressure (CPAP) on performance on driving si mulators were little investigated.20 ·21 In addition, there are only scarce data concerning the effects of treatment with nCPAP on the re al-life occurrence of accidents in OSA patients involving retrospective studies in relatively small groups of patients. 23 ·24 Therefore, we included an evaluation of the occurrence of accidents in a large prospective study of the subjective effects of treatment \v:ith nCPAP; this study has the drawback of lacking a CHEST I 112 I 6 I DECEMBER, 1997

1561

control group, which is unavoidable, as an untreated control group would imply leaving a potentially hazardous condition untreated for a prolonged p eriod of time. MATERIALS AND METHODS

The objective of the study was to evaluate over a 1-year pe1iod the effects of treatment with CPAP in terms of clinical symptoms, quality of life, activity, and use of medical care as part of a large r study designed to evaluate the benefit obtained with CPAP treatment, which was conducted in 59 centers working with the ANTADIR (French respiratory homecare network). The present report relates to data concerning accidents; the other results of the study will be reported elsewhere. Patients

All patients were referred to a sleep center for symptoms suggestive of OSA; the final diagnosis was based either on full polysomnography (including recorcling of flow, respiratory effort, arterial oxygen saturation, and EEG+electro-oculogram + electromyogram of chin muscles) or on respirat01y polygraphy (including recording of flow, respirat01y effort, and mterial oxygen saturation without identification of sleep stages). In th e form er group, apnea and apnea + hypopnea indexes were expressed as a function of total sleep time; in the latter group, they were expressed as a function of total recording time. Full polysomnography was used for CPAP titration in all cases. All patients also underwent respiratory function tests and resting daytime blood gas analysis. The study included 1,273 patients between January 1, 1994 and September 30, 1994. Their apnea and apnea+ hypopnea indexes were 52.2:+::28.1 for the patients investigated with full polysomnography, and 48.6:+::23.6 for the patients investigated with respiratory polygraphy. On the basis of each center's usual strategy of choice of treatment for OSA, CPAP was proposed to 973 patients, of whom 893 (637 patients with full, 252 patients with respiratory, 4 with unspecified polysomnography) accepted both to continue home treatment with CP AP and to take part in the study, \vhich was approved by the Ethics Committee of the University Hospital of Angers, France. Methods

The patients were given a self-questionnaire before treatment (baseline) and after 3, 6, and 12 months of treatment with nCPAP. On each occasion, th e patients also had a medical interview and clinical investigation, and the time counters of their CP AP machines were noted. The baseline questionnaire included two groups o f questions concerning accidents in the previous 12 months. The first questions asked whether patients had bad an accident; if the answer was positive, th ey were asked to specify what type of accident (traffic, home, work, or other) and whether th ey felt that the accidents(s) were related to lack of vigilance, fatigu e, or sleepiness; they were also asked to specify whether the accident had required a hospitalization , and if so, for how many days. The second group of questions asked whether the patients felt th ey had had near-miss accidents due to lack of vigilance, fatigue, or sleepiness and what type of accidents. The 6- and 12-months follow-up questionnaires included the same ques tions refening to the last 6 months; the accidents reported on each follow-up questionnaire were cumulated and compared with the accidents reported for the 1-year p eriod before treatment. 1562

Full questionnaire results were available for 547 patients; 153 left the study and only pmtial data (1 or more questionnaire missing) were obtained in 193 patients, who were discarded from the present analysis. In each questionnaire, the patients were also asked to report how frequently ( neve r=1 , sometimes=2, often=3, almost always=4, does not know or not concerned \vith the given activity= O) they dozed off during seven activities of daily life (after a meal, at work, at social meetings, watching TV, while reading, when a passenger in a car or a means of transportation, or while d1iving); a sle epiness score was computed as the average score of th e answers, provided the patients had answered at least four items ; otherwise, the s le epiness score was considered missing; this sleepiness score is close in its principle to tb e Epworth Sleepiness score; it was shown to separate OSA subjects from normal subjects, and to decrease \vith CPAP treatmen t. 2·5 Sleepiness scores could be calculated in 513 of the patients who completed the questionnaire; the 34 patients in whom the sleepiness score could not be calculated were not different from the re maining 513 patients in terms of age, sex ratio, apnea+ hypopnea index, daytime pulmonary function and blood gases, or compliance with CPAP. Statistical Analysis

Comparisons between patients receiving CPAP (n=893 ) and patients with other or no treatment (n=380) and comparisons between patients completing the study or not were made using the Mann-\rVhitney test; the effects of treatment were investigated using the vVilcoxon test. Propmiions were compared using the test of equality of propmtions. Differences were considered significant for p< 0.05. Results are reported as means::'::SD.

RESULTS

Table 1 shows the main characte1istics of the 893 patients who were included in the study. There were no differences in age and sex ratio between patients treated with CPAP (n=893) and those receiving other treatments or untreated (n=380); patients treated with CPAP had higher body mass index (BMI), lower daytime Pa0 2 , higher daytime PaC0 2 , and lower FEV 1 and vital capacity than those receiving other treatments or untreated; as a group, they comprised higher percentages of patients with BMI >29 kglm 2 , and of those exhibiting symptoms of sleepiness, fatigue, witnessed apneas, or morning headaches. They also had higher apnea indexes and apnea+ hypopnea indexes (Table 1). Patients who left the study (n = 153) did so because of one of the following reasons: treatment interruption (n=l08), unavailable for follow-up (n=3l), death (n=3), unwillingness to complete the questionnaires while continuing CPAP (n=7), and unspecified (n=4). As a group, they tended to have a lower BMI and higher daytime Pa0 2 than patients who completed the study. Patients excluded for missing data (n=l93) were not different from patients with complete questionnaire data. Clinical Investigations

Table !-Comparisons Between Patients Treated With CPAP and Patients Receiving Other or No Treatment*

No. of patients Age, yr % male BMI , kglm 2 % BMI > 29 kglm 2 Daytime Pa0 2 , mm Hg Daytime PaC0 2 , mm Hg FEV 1, mL VC, mL Complaint of Snoring,% Sleepiness, % Fatigue,% Witnessed apneas, % Morning headaches, % Polysomnography Total sleep time , min Apnea index, No./h Apnea+ hypopnea index, No.lh Respiratory polygraphy Total recording time, min Apnea index, No.lh Apnea+hypopnea index, No.lh

CPAP

Other

p Value

893 .56.6::': 10.7 86.5 33.7::':6.8 69.2 75.6::':11.6 41.8::':5.0 2820::':858 3701 ::': 999

380 56.1 ::':11.0 86.3 31.0::':7.2 48.3 78.8::':12.3 40.4::':4.5 3017::':919 3933::': 1060

NS NS <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

97.6 89.6 82.5 80.2 51.8 n =725 335.1::':86.2 39.1::':24.9 59.8::':25.8

95.0 78.2 68.4 74.7 43.9 n=321 351.3::':87.6 18::':20.8 35.0::':25.3

n=168 446.6::':69.8 34.9::':21.1 53.6::':21.6

n = 59 463.9::':60.6 19.9::':18.0 34.6::':23.4

NS <0.01 <0.01 <0.05 <0.0 1 <0.01 < 0.01 <0.01

NS <0.0 1 <0.01

*VC=vital capacity; NS=not significant.

Table 2 shows that both the number of patients and the average number of accidents per patient decreased with treatment, for real accidents (p<0.01 ) as well as for near-miss accidents (p< 0.01 ). It also shows that the proportion of accidents related by the patients to lapses in vigilance was lower dming treatment than before treatment. All together, the patients had 94 accidents during the year before treatment, of which they attributed 49 to lapses in vigilance; they had 40 accidents during the first year of receiving treatment (with 13 attributed to lapses in vigilance). Compliance rates (measured by th e average daily use of the device, calculated from the timing records on the equipment) were not

different between patients reporting accidents and patients reporting no accidents while receiving CPAP (5 h 57 min±1 h 53 min vs 6 h 7 min:±:1 h 58 min). Of the 94 accidents during the untreated year, 16 patients required a hospitalization, accounting for a total number of 885 days in hospital, whereas eight accidents during the treatment year required 84 days in hospital. Table 3 shows the comparisons between the 60 patients who had real accidents and the 487 patients who had no accident before treatment; only the sleepiness score was higher in the accident group. DISCUSSION

This study indicates a dramatic decrease in the number of accidents, including traffic, home, and work accidents, occurring over equal periods of time in OSA patients treated with nCPAP. It also shows that in addition to traffic accidents, which have been investigated extensively, OSA patients are also exposed to other types of accidents, whose decrease with CPAP treatment suggest that they are related to the disease. There are several limitations to the study that should be discussed before attributing the observed effects to the elimination of sleep apneas by CP AP. First, the data are based on self-questionnaires, and are therefore dependent on the patients' recollection of their own accidents. It should be noted that the way the data were collected, over one single 1-year period b efore treatment, and in 2 six-month periods of receiving treatment, would rather favor a better recollection of accidents during the follow-up period, especially of minor ones. Second, the absence of a placebo group, which is difficult to constitute for a disease known to increase morbidity and mortality, does not make it possible to separate from among the observed effects those that are specifically related to the elimination of CP AP and the nonspecific effects related to the emphasis

Table 2-0ccurrence of Real Accidents and Near-Miss Accidents During the Year Before Treatment and the First Year Receiving Treatment With nCPAP in 547 OSA Patients Near-M iss Accidents

Real Accidents

No. of patients having had an accident Average No. of accidents per patient Car accidents, No. (%) Domestic accidents, No. (%) Work accidents, No. (%) Others+unspecified, No. (%) Accidents related to lack of vigilance, %

Before CPAP

With CPAP

Before CPAP

With CPAP

60 1.57::': 1.33 46 (48.7) 25 (263 ) 12 (1 32) 11 (11.8) 52.6

36 1.11::':0.32 14 (34.2) 13 (31.6) 7 (18.4) 6 (15.8) 33 .3

151 4.47::': 6.5 505 (74.8) 65 (9.6) 63 (9.3) 42 (6. 3)

32 1.8::': 1.4 39 (66.2) 8 (13.2) 2 (4.4) 9 (16.2)

All

All

CHEST I 112 I 6 I DECEMBER, 1997

1563

Table 3-Comparisons Between the Patients Who Had Real Accidents Before Treatment and Those Who Had No Accident*

Age, yr BMI, kglm 2 AHI, No./h Sleepiness score

Accident Group (n = 60)

No Accident Group (n=487)

p Value

56.5:+:ll.2 34.3:+:7.4 65.0:+:28.1 2.4:+:0.6

57.1:+: 10.8 33.9:+:6.8 60.8:+:25.1 2.2:+:0.7

NS NS NS < 0.05

*AHI=apnea hypopnea index; NS=not significant.

put on the patients' sleepiness once they were diagnosed as having sleep apnea, or the care taken of them, or the mere recognition that they were being observed. However, given the known effects of OSA on daytime vigilance, the recognized relationship between sleepiness and accidents, 1 -6 and the improvement in daytime sleepiness with CPAP treatment, 14- 18 it appears reasonable to accept that the observed decrease in accidents is at least in part related to the specific effects of CP AP on the symptoms of OSA. In addition , even if part of the reduction in accidents was to be ascribed to nonspecific effects of treatment, the benefit for the patients would still be obvious. Because of the lack of an adequate control group, we considered data on accidents in the general population. Table 4 shows the number of accidents in the general French population and the corresponding rates of accidents in our patient sample before and on treatment with CPAP. However, it is clear that no comparison is possible for many reasons. The age range of the patients is different from the general population but our patient sample was not large enough to constitute subgroups by age (and sex) for adequate comparisons; still, it is well known that younger people are more at risk for traffic accidents, and older people are more at risk for domestic accidents. The type of accidents reported is also different; especially for car accidents, patients

reported all types of accidents, whereas national statistics include only accidents reported to police authorities or insurance companies. This probably explains the much higher rate of car accidents in our patients. Bearing in mind the above caveats and excluding car accidents, it could be suggested that patients had slightly more accidents than the general population before treatment, and less accidents than the general population when receiving treatment. Third, exposure to risk was not evaluated in this study. If the time spent driving or the time spent at work had been lower during the year of receiving treatment than during the year before treatment, it would have caused a decrease in the number of accidents unrelated to the effects of CPAP. No specific information concerning number of kilometers driven or number of hours worked was collected. However, the number of patients who were actually working during the two pe1iods (293) did not change, as the number of patients who stopped working due to retirement (n=5) was balanced out by the number of patients who resumed working after sta1ting CPAP (n=8). In addition, one may expect that sleepy patients might have reduced their exposure to risk (some patients avoiding driving) before treatment. Therefore, it can reasonably be accepted that the reduced number of accidents with treatment was not the result of a reduced exposure to risk of accidents. To our knowledge, only two other studies investigating the effects of treatment with CPAP on the occurrence of accidents in OSA patients have been published so far. Both concerned only traffic accidents. Minemura et al 23 found that 42% of 14 patients with OSA reported at least one accident during the 3 years before treatment and 0% of these subjects had an accident during a year of treatment with CPAP. The small number of patients and the complete absence of accidents during the treatment period suggests that a selection bias must have

Table 4-Number of Accidents in the General French Population Compared to the Number of Accidents in the Study Patients*

Source/Yearffype of Accts CNAMTS/1992 Nontraffic nonwork accts Domestic accts INSEE/1990 traffic accts CNAMTS/1994 Work accts (excluding car accts)

No. of Accts for l 00 Persons

Our Study Before CPAP

With CPAP

3.7 0.4

6.6 4.6 8.4

3.8 2.4 2.5

2.1

2.2

1.3

5.1

*CNAMTS=Caisse Nationale d'Assurance Maladie des Travailleurs Salaries; INSEE=Institut National des Statistiques et Etudes Economiques; accts =accidents.

1564

Clinical Investigations

occurred. Engleman et al 24 divided traffic incidents into near-misses, casualty-free (minor) collisions, and major collisions; the investigation was based on 204 replies to a questionnaire sent to CPAP users, asking for the occurrence of road traffic incidents during the 5 years before starting CPAP and the time since starting CPAP, which ranged from 16 to 2,921 days. There was a decrease in all types of incidents, which was significant only for near-miss incidents (probably because of the small number of minor and major collisions). Because of differences in method, it is difficult to compare our study with this previous study. It is clear that data on accidents can be collected only retrospectively, after the accident has occurred. However, we would like to stress that the above studies were fully retrospective, in that the information on accidents for the two study periods (baseline and treated) was collected at the same time, after patients had been treated. Since in our study, questionnaires were completed before CPAP was started for the pretreatment accidents, the data collection for accidents during the treatment period was prospective. Our results thus strengthen the previously reported decrease in accidents with CPAP treatment. Indeed, the risk that patients before treatment may be tempted to deny accidents for fear of being banned from driving, and the retrospective collection of data for the pretreatment period, would rather lead to an underestimation, if anything, of accidents before treatment period, whereas the prospective collection of the data during the treatment period would rather have minimized the risk of forgetting minor accidents. It should also be emphasized that the reduced cost for the community related to the reduction of the number of accidents 26 ·27 should be taken into account when evaluating the cost/benefit ratio of treating OSA patients. This reduction in cost is best demonstrated in our study by the reduction in the number of accident-related days of hospitalization from 885 to 84; this dramatic reduction was due both to the decreased number of accidents and to the shorter duration of hospitalizations after residual accidents in treated patients. Given the known severity of sleepiness-related accidents, 10 the shorter duration of stays in hospital after the residual accidents during the treated period is in agreement with the lower proportion of accidents attributed to sleepiness by the patients. In conclusion, even though the nature of our study does not make it possible to ensure that the observed effect is a specific effect of the elimination of apneas with nCPAP treatment, this study shows a significant decrease in the number of self-reported accidents in OSA patients when treated with nCPAP.

ACKNOWLEDGMENTS: The authors are grateful to V. GrilletLanoir and M. Coisy-Vialettes for preparing the questionnaires and for coordinating the data collection, to N. Rolland for the statistical analysis, and to B. Fleury and M. Tiberge for their thoughtful comments on the manuscript.

APPENDIX Investigators: P. Meekel, CH. Troyes; 0. Gerin, Clinique Prevallon Cottage, Marseille; F. Philip-Joet, CHU Nord, Marseille; J.N. Prevost, CHU Cote de Nacre, Caen; J. Vincent, CHG SaintLouis, La Rochelle; 0. Reybet-Degat, CHU de !a Trouhaude, Dijon; S. Arabs, CHG Lannion; S. Lacroix, CH Dujassic de !a Rivie re, Perigueux; V. Gillet, Don du Souffle, Besan~on; L. Vascaut, Hopital La Muse, St Sebastien de Morsent; F. Martin, CH Victor Jousselin, Dreux; Leophonte, Hopital de Rangueil, Toulouse; Robert, Clinique des Cedres, Cornebarsieu; Sapene, Polyclinique de Pins Francs, Bordeaux; J.M. Vernejoux, CHU Bordeaux, Pessac; B. Desrues, CHU de Pontchaillou, Rennes; E. Goarant, CHG Broussais, Saint-Malo; G. Ginies, CHU Bretonneau, Tours; P. Levy, CHU La Tronche, Grenoble; CourtFortune, CHU St-Etienne; S. Chollet, CHU Nantes; N. Meslier, CHU Angers; Maury, CHG Cholet; M. Becu, CHG Chalons-surMarne; A. Cazabat, CHU Reims; M. Desjobert, CHG Laval; A. Cornette, CHU Nancy Brabois, Vandoeuvre; P. Tagu, CHG Earle-Due; H. Joffreau, Hopital Belle-Isle, Metz; P. Cervantes, Hospitalor, Saint Avoid; A. Mallard, CHU Calmette, Lille; F. Steenhouwer, CHG Victor Provo, Roubaix; F. Fortin, Clinique de !a Louviere, Lille; Demarcq-Delerue, Clinique Teissier, Valenciennes; E. Fournier, Clinique Darcy, R enin-Beaumont; J. Amourette, CHG Lens; F. Bart, CHG Bethune; H. Monnot, CHG Saint-Omer; A. Greil, AIRRA Clermont-Ferrand; J. Camborde, SCP Pau; J.M. Chavaillon, CHG La Fontonne, Antibes; J. Krieger, CHU Strasbourg; J.M. Zipper, CHG Mulhouse; G. Carelli, CH Rouffach; J. Mouret, CHS Le Vinatier, Bron; E. Orvoen-Frija, Hotel Dieu, APHP Paris; B. Fleury, Rakotonanahary, Hopital Saint-Antoine APHP Paris; B. Fleury, CTAR Paris; C. Tardif, CHU Charles Nicolle, Rouen; J. Quieffin, CHG Monod, Le Havre; G. Ozenne, Centre medico-chirurgical BoisGuillaume, Rouen; F. Portier, CHU BoisGuillaume, Rouen; Rida, APAIR, Amiens; T. Pigeanne, CHG Les Sables d'Olonne; J.C. Meurice, Patte, CHU La Miletrie, Poitiers; J. Garnier, CHG Belfort; A. Foucher, Hopital Ambroise Pare, APHP Boulogne; D. Baud, Hopital de Chevilly La Rue; J.L. Sarraux, CHG Emile Roux, Eaubonne.

REFERENCES 1 Roth T, Roehrs T. Sleepiness and motor vehicle accidents. Int Clin Psychopharmacol1988; 3:111-16 2 Arbus L, Tiberge M, Serres A, et al. Somnolence et accidents de Ia circulation routiere: importance du diagnostic. Neurophysiol Clin Clin Neurophysiol 1991; 21:39-44 3 Gold DR, Rogacz S, Bock N, et al. Rotating shift work, sleep, and accidents related to sleepiness in hospital nurses. Am J Public Health 1992; 82:1011-14 4 Accidents and sleepiness: A consensus statement from the International Conference on Work Hours, Sleepiness and Accidents, Stockholm, September 8-10, 1994. J Sleep Res 1994; 3:195 5 Dinges DF. An overview of sleepiness and accidents. J Sleep Res 1995; 4:4-14 6 Horne JA, Reyner LA. Sleep related vehicle accidents. BMJ 1995; 310:565-67 7 Aldrich MS. Automobile accidents in patients with sleep CHEST I 112 /6/ DECEMBER, 1997

1565

disorders. Sleep 1989; 12:487-94 8 Findley L, Unverzagt M, Guchu R, et la. Vigilanl:e and automobile accidents in patients with sleep apnea or narcolepsy. Chest 1995; 108:619-24 9 George CF, il:kerson PW, Hanley PJ, et a!. Sleep apnea patients have more automobile accidents [ letter] . Lancet 1987; 2:447 10 Findley LJ , Unverzagt ME , Suratt PM. Automobile accidents in volving patients with obstructive sleep apn ea. Am Rev Respir Dis 1988; 138:337-40 11 Cassel W, Ploch T. Sleep apnea and accidents: health risk for healthy people? In: Peter JH, Penzel T, Podzus T, et al, eds. Sleep and health risk. Berlin: Springer-Verlag, 1991; 279-85 12 Stoohs RA, Guilleminault C, Itoi A, et l.aTraffic accidents in com me rcial long-haul truck drive rs: the influence of sleepdisordered breathing and obesity. Sleep 1994; 17:619-23 13 Wu H, YanGo F . Self-reported automobile accide nts involving patients with obstructive sleep apnea. Neurology 1996; 46:1254-57 14 Rajagopal KR, Bennett LL, Dillard TA , et al. Overnight nasal CPAP improves hypersomnolence in sleep apnea. Chest 1986; 90:172-76 15 Lamphe re J, Roe hrs T, Wittig R, e t al. Recovery of alertness after CPAP in apnea. Chest 1989; 96:1364-67 16 Bedard MA, Montplaisir J, Malo J, et a!. Persistent neuropsychological d eficits and vigilance impairment in sleep apnea syndrom e after treatment vvith continuous positive ai1way pressure. J Clin Exp Neuropsychol 1993; 15:330-41 17 Engleman HM , Cheshire KE, Deary IJ, et al. Daytime sleepiness, cognitive performance an d mood after continuous positive airway pressu re for the sleep apnoea hypopnoea syndrom e. Thorax 1993; 48:911-14

1566

18 Sforza E , Kri eger J. Daytime s leepiness after long-term nasal co ntinuous positive airway pressure (CPAP) treatment in obstructive s el ep apnea syndrome. J Neural Sci 1992; 110: 21-26 19 Borak J , Cieslicki JK, Koziej M , et l.a Effects of CPAP treatment on psychological status in patients with severe obstructive sleep apnoea. J Sleep Res 1996; 5:123-27 20 Findley LJ, Fabrizio MJ, Knight H, et la. Driving simulator performance in patients with sleep apnea. Am Rev Respir Dis 1989; 140:529-30 21 Engleman HM , Martin SE , Dea1y IJ, et la. Effect of continuous positive airway pressure treatm ent on daytime function in sleep apnoealhypopnoea syndrome. Lancet 1994; 343: 572-75 22 Ovesen J, Nielsen PW, vVildschiodtz G. Sh01tened reaction time during nasal CPAP treatment of obstructive sleep apnea. Acta Otolatyngol 1992; 492(suppl):ll9-21 23 Minemura H, Akashiba T, Yamamoto H, et l.a Traffic accidents in obstructive sleep ap nea patients and effects of nasal CPAP treatm ent. Jpn J Thorac Med 1993; 31:1103-08 e od AL, et a!. 24 Engleman HM , Asgari-Jirhandeh N , McL Self-reported use of CPAP and benefits of CPAP therapy: a patient survey. Chest 1996; 109:1470-76 25 Petiau C, Delanoe C, Hecht MT, et al. Somnolence e t syndrome d'apnees du sommeil: analyse de 188 questionnaires (102 malades et 86 temoins) . Neurophysiol Clin 1993; 23:77-85 26 Leger D. The cost of sleep-related accidents-a report for the national commission on sleep disorders research. Sleep 1994; 17:84-93 27 \\'ebb WB. The cost of sleep-related accidents: a reanalysis. Sleep 1995; 18:276-80

Clinical Investigations