ORIGINAL ARTICLE

Effects of continuous positive airway pressure on lung function inpatients with chronic obstructive pulmonary disease and

sleep disordered breathing

D MANSFIELD AND M T NAUGHTON

Alfred Hospital and Monash University Medical School, Prahran, Melbourne, Victoria, Australia

Effects of continuous positive airway pressure on lung function in patients with chronic obstruc-tive pulmonary disease and sleep disordered breathingMANSFIELD D, NAUGHTON MT. Respirology 1999; 4: 365–370Objective: Sleep disordered breathing (SDB), namely hypoventilation and obstructive sleepapnoea, occur in about 50% of patients with severe chronic obstructive pulmonary disease (COPD).Previous studies that have investigated the reversal of SDB in such patients with nasally appliedintermittent positive airway pressure have reported a fall in PaCO2 but little change in airflowobstruction. We reasoned that the lack of improvement in airflow obstruction may be due to insuf-ficient expiratory pressure. Accordingly, we sought to determine the effects of chronic nasal con-tinuous positive airway pressure (CPAP), at highest tolerable levels, upon blood gases and airflowobstruction in patients with severe COPD and SDB. Methodology: Fourteen patients were studied, ten of whom were able to tolerate CPAP (10.2 ±0.9 cmH2O) for at least 3 months. Results: Within the CPAP compliant group, there was a fall in PaCO2 (58.0 ± 3.5 to 48.0 ± 0.9 mmHg,P = 0.015) associated with a rise in PaO2 (54.8 ± 3.8 to 63.2 ± 1.8 mmHg, P = 0.015) and forced expira-tory volume in 1 s (0.95 ± 0.13 to 1.10 ± 0.13 L, P < 0.005). Concurrent with these improvements was asubstantial fall in hospitalization rates (from 3.85 to 0.73 admissions per annum). Conclusion: Improvements in gas exchange, airflow obstruction and hospitalization rates wereobserved in patients with COPD and SDB treated with nasal CPAP during sleep.

Key words: chronic obstructive pulmonary disease, continuous positive airway pressure, sleep dis-ordered breathing.

and hypercapnia, a marker of poor prognosis inpatients with COPD.3

Sleep-related hypoventilation is due to a reductionin respiratory drive, loss of accessory muscle activityand ventilation perfusion mismatch.2 Associated withhypoventilation in such patients is an increased workof breathing,4–7 related to upper and lower airwayobstruction. Respiratory muscles may also fatiguewhich is related to the mechanical disadvantage ofchest wall hyperinflation and possibly due to drugeffects and impaired nutrient supply (increaseddemand for cardiac output8 and reduced nutrientintake due to dyspnoea). There is also a reduction infunctional residual capacity which is related to supineposture and sleep state.1,2,9

This has led to the concept of providing non-invasive ventilatory support during sleep to reversesleep-related hypoventilation particularly in hyper-capnic patients with COPD. In 1988, Carroll and

Respirology (1999) 4, 365–370

INTRODUCTION

The prevalence of sleep disordered breathing (SDB)in patients with severe chronic obstructive pul-monary disease (COPD) is estimated to be approxi-mately 50%.1 Sleep disordered breathing consistsmainly of sleep-related hypoventilation and less com-monly obstructive sleep apnoea (OSA).2 Each maylead to the development of pulmonary hypertension

Correspondence: Dr Matthew T Naughton, Sleep Dis-orders and Ventilatory Failure Service, Department ofRespiratory Medicine, Alfred Hospital, CommercialRoad, Prahran, Melbourne, Victoria 3181, Australia.Email: matthew.naughton@ med.monash.edu.au

Received 25 November 1998; accepted for publication14 May 1999.

Braithwaite reported less nocturnal hypoxaemia andlower awake PaCO2 following 3 months of treatmentwith volume cycled non-invasive ventilatory supportapplied during sleep.10 Thereafter, a further three con-trolled studies addressing the efficacy of assisted ventilation for a medium term drew conflicting con-clusions.11–13 In the study of Stumpf et al., no signifi-cant change was observed in awake arterial bloodgases, lung function tests, or respiratory musclestrength following a 3 month cross-over designedstudy of bilevel positive airway pressure comparedwith air.11 Meecham Jones et al. reported an improve-ment in arterial blood gases, sleep quality and qualityof life with a 3 month cross-over designed study ofbilevel positive airway pressure with supplementaloxygen compared with oxygen alone.12 More recently,Lin failed to demonstrate an improvement in awakeblood gases following a 2 week trial of bilevel positiveairway pressure.13 Three long-term outcome, non-randomized studies have shown improvements inblood gases14–16 but not in lung function.

In each of these studies, the expiratory positiveairway pressure (EPAP) was relatively low. The levels of EPAP used by Stumpf et al.11 and Lin13 weresignificantly lower (0–2 cmH2O) than that used byMeecham Jones et al. (2–4 cmH2O).12 SubtherapeuticEPAP may lead to CO2 rebreathing,17 an inability toovercome the increased work of breathing related to intrinsic positive end-expiratory pressure (PEEP), and an inability to prevent supine and sleep-relatedreductions in functional residual capacity. Finally, theEPAP levels may not have been sufficiently highenough to prevent upper airway closure. On the otherhand, excessively high levels of EPAP may lead toimpaired venous return and activation of expiratorymuscles.

Acutely, continuous positive airway pressure(CPAP) has also been shown to reduce the work ofinspiration in patients with COPD4–6 and asthma.18

Because CPAP provides an expiratory resistance, aswell as inspiratory support, as does bilevel positiveairway pressure, we hypothesized that CPAP that isused long term at highest pressures tolerable mayimprove respiratory function during wakefulness. Inorder to determine whether CPAP treatment results inimproved lung function, we analysed a group 14 patients with severe COPD and sleep-relatedhypoventilation.

METHODS

Patient selection

Consecutive stable patients, aged 35–80 years, withsymptomatic COPD and sleep disordered breathingwho required at least one admission in the previousyear for acute exacerbation of COPD were selected.Clinical stability was determined by no hospitaladmission or change in medication in the previous 4weeks. All patients suffered COPD as evidenced by a smoking history, expiratory wheeze and hyper-inflation. Lung function criteria for COPD were inaccordance with American Thoracic Society (ATS)

guidelines19 with forced expiratory volume in 1 s(FEV1) < 80% predicted and/or reduced maximal mid-expiratory flow (25–75%) rates. Exclusion criteriawere greater than a 15% bronchodilator responsesuggesting asthma or the presence of significant neu-rological, renal or cardiac disease or clinical instabil-ity as defined above. Sleep disordered breathing was defined as oxygen saturation < 90% for > 10% oftotal sleep time and/or the presence of OSA (i.e.apnoea–hypopnoea index (AHI) > 5 events per hour).All patients were on optimal medical therapy forCOPD and had other causes of dyspnoea (cardiacdisease, anaemia, thyroid disease etc.) excluded. Allwere receiving oral or inhaled corticosteriods whichwere continued throughout the period of follow up.No patient had escalation of their long-term mainte-nance therapy during the study period.

Measurements

Spirometry was measured according to ATS criteria19 while patients were awake and in the seatedposition (Jaeger Master Lab, Wuerburg, Germany).Spirometry values reported in this study were post-bronchodilator. An arterial blood gas sample wastaken from the radial artery using a 25 g needle andblood gases and pH determined (ABL 500 Radiome-ter, Copenhagen, Denmark).

Standard polysomnography sleep studies wereconducted and recorded onto a computerized recording system (Somnostar; SensorMedics Corp,Anaheim, CA, USA). Surface electrodes were used torecord electroencephalogram, electro-oculogram,submental electromyogram, cardiograph and ante-rior tibial electromyogram. Standard criteria wereused to manually stage sleep.20 Oxygen saturation(SpO2) was measured by an ear oximeter (OhmedaBiox 3700 Pulse Oximeter; Colorado, USA) and transcutaneous PCO3 (PtcCO2) was measured by acapnograph placed on the anterior chest wall(FasTrac; SensorMedics). Chest and abdominal move-ments were monitored using respiratory effort bands (Resp-ez; EPM Systems, Midlothian, VA, USA).Oronasal flow was monitored by thermocouples.

Hospital admission rates were obtained by way ofpatient and general practitioner interview and analy-sis of hospital records prior to, and following, thecommencement of CPAP. Patients were followed for avariable length of time (minimum 8 months).

Protocol

All patients received an initial polysomnograph, lungfunction and arterial blood gas analysis. All patientswere questioned as to the baseline frequency of hos-pital admissions and hospital records were reviewed.

Continuous positive airway pressure was com-menced over 2–3 nights during an elective in-patienthospital admission. Continuous positive airway pres-sure via a nasal mask was initiated during wakeful-ness at a pressure of 5 cmH2O. Patients wereencouraged to use CPAP for 30–60 min periods in the

366 D Mansfield and MT Naughton

morning and again in the afternoon. Thereafter,patients were encouraged to use CPAP overnight.Continuous positive airway pressure was thenincreased by 1–2 cmH2O aliquots depending onpatient tolerability over the next 2 days and results ofeither overnight oximetry or polysomnography. Thismode of CPAP initiation was chosen because it is ourexperience that such patients are often claustropho-bic and feel inhibited with the combination of CPAPand the electrodes required for full polysomnography.This mode of CPAP titration is similar to that used forpatients with congestive heart failure and Cheyne-Stokes respiration21 who also feel claustrophic.Patients were reviewed in the clinic at monthly inter-vals until compliance was determined and thereafterat 3, 6 and 12 month intervals. Continuous positiveairway pressure compliance was determined fromCPAP hour meters and was defined as CPAP use forgreater than 3 h on average per night.

Patients were followed longitudinally for a mini-mum of 8 months. Resting arterial blood gas analysis,pulmonary function testing, polysomnography andadmission rate were re-evaluated.

Statistics

Data are presented as the mean and standard error.Comparisons of data between compliant and non-compliant patients, and within compliant patients,were made using unpaired and paired t-tests asappropriate. Significance was indicated by P < 0.05.

RESULTS

Fourteen patients met the eligibility criteria of whom10 were compliant with nasal CPAP (Table 1). The

group non-compliant of CPAP had less severe OSA(AHI 4.4 ± 3.3 [range 1–7] vs 16.9 ± 6.1 [range 0–60]events per hour, P < 0.05) and less hypoxaemiaovernight (total sleep time spent with SpO2 less than90%: 53 ± 24 vs 72 ± 9%, P < 0.05) compared with theCPAP compliant group. There were no significant dif-ferences in severity of COPD, age, gender or bodymass index between the two groups. Within the 10patients who were compliant with CPAP, the baselinesleep efficiency was 71 ± 4% and the percentage sleeptime spent in slow wave and rapid eye movementsleep were 10 ± 2% and 10 ± 1%, respectively.

Within the compliant group, the average level ofnasal CPAP was 10.2 cmH2O and compliance 4.8 h.The average duration of follow up after the introduc-tion of CPAP was 16.5 (range of 8–40) months. Thenumber of admissions per annum, pre-CPAP, was 3.85(calculated from the previous 4 years) and fell to 0.73admissions per year following initiation of CPAP (P = 0.0003). Similarly, the total in-patient days wasreduced from 25.6 to 5.1 days per annum.

Severity of sleep disordered breathing improvedwith nocturnal CPAP use (Table 2). The AHI fell (16.9± 6.1 to 6.4 ± 3.4 events per hour, P = 0.025) as did thepercentage total sleep time spent with SpO2 less than90% (72 ± 9 to 30 ± 10%, P = 0.002) and less than 80%(20 ± 11 to 4 ± 3%, P = 0.09).

Lung function improved in all but one patient following extended CPAP usage (Table 2). The FEV1

increased from 0.96 ± 0.13 to 1.10 ± 0.13 L (P = 0.005).A similar trend was observed in forced vital capacity(2.24 ± 0.28 to 2.38 ± 0.25 L, P = 0.084). Resting awakePaO2 (54.8 ± 3.8 to 63.2 ± 1.8 mmHg, P = 0.015) andPaCO2 (58.0 ± 3.5 to 48.0 ± 2.9 mmHg, P = 0.015) levelstaken improved while breathing room air. CoexistentOSA was observed as well as hypoventilation in threepatients. To control for the isolated effect of OSA, theeffect of CPAP on lung function and arterial bloodgases was analysed in two subsets within the compli-ant group; those with an AHI greater than 15 eventsper hour and those with an AHI less than 15 eventsper hour. The benefit was seen in both groups inde-pendent of the presence of OSA.

DISCUSSION

This study examined the effects of nasal CPAP appliedduring sleep upon awake lung function and hospitali-zation rates in patients with severe COPD and sleepdisordered breathing. The main findings were of a sig-nificant reduction in PaCO2, a rise in both PaO2 andFEV1 associated with a fall in hospitalization rateswith CPAP treatment. Although previous publishedreports suggest similar improvements in blood gaseswith bilevel positive airway pressure or volume cycledventilators,11–13 we believe this is the first descriptionof a similar improvement with CPAP treatment. More-over, there was an improvement in lung function, forwhich several mechanisms are possible, namelybronchodilatation, reduced work of breathing andalterations in ventilation perfusion matching.

Initial reports by Carr and Essex were of a 33–71%increase in the diameter of large conducting airways

CPAP on lung function 367

Table 1 Patient characteristics

Non-compliant Compliant P

Number 4 10 NSAge (years) 68 ± 3 58 ± 3 NSGender (M : F) 3 : 1 6 : 4Body mass index 30.7 ± 5.3 36.5 ± 2.3 NS

(kg/m2)FEV1 (L) 0.94 ± 0.13 0.95 ± 0.13 NSFVC (L) 2.15 ± 0.22 2.38 ± 0.90 NSFEV1/FVC (%) 44 ± 3 40 ± 3 NSpH 7.43 ± 0.02 7.39 ± 0.03 NSPaO2 (mmHg) 62.0 ± 3.5 54.8 ± 3.8 NSPaCO2 (mmHg) 48.3 ± 4.8 58.0 ± 3.5 0.18Apnoea–hypopnoea 4.4 ± 3.3 16.9 ± 6.1 0.027

index (events/h)Sleep efficiency (%) 70.0 ± 5.4 66.5 ± 5.7 NSMean SpO2 (%) 87.5 ± 4.7 86.8 ± 2.1 NSMinimum SpO2 (%) 73.3 ± 6.0 63.3 ± 6.7 NSSpO2 < 90% (% TST) 53 ± 24 72 ± 9 0.013

Sleep efficiency, total sleep time/time in bed; TST, totalsleep time.

of anaesthetized spontaneously breathing dogs given20 cmH2O CPAP suggesting a significant bron-chodilatory effect.22 Barach and Swenson observedsimilar findings in humans with asthma given posi-tive airway pressure.23 Further support for the CPAP-induced bronchodilatory effect was provided byreports of improvements in objective markers ofasthma, namely peak expiratory flow rates, in stablepatients with OSA following treatment with nocturnalCPAP.24 In patients with acute asthma, CPAP has alsobeen found to reduce both respiratory rate and dysp-noea scores.25 Continuous positive airway pressurehas been shown to acutely improve overnight oxy-genation in patients with combined apnoea andCOPD,3 however, longer term effects on blood gasesand spirometry were not studied.

Previously it has been shown that CPAP can sig-nificantly reduce inspiratory effort and work ofbreathing in patients with acute respiratory failureweaning from invasive mechanical ventilation.6 It hasbeen reported that CPAP can reduce inspiratory effortto overcome intrinsic PEEP during sleep in patientswith COPD.4,5 These observations may explain theimprovement in respiratory muscle strength associ-ated with an increase in exercise endurance observedin patients with COPD and mild sleep disorderedbreathing treated with 5–8 cmH2O CPAP duringsleep.26 Moreover, low levels of CPAP (5 cmH2O) givenwhile awake to patients with severe COPD duringexercise results in an improvement in exerciseendurance time and sensation of dyspnoea,27 aneffect thought to be due to unloading of inspiratorymuscles and attenuation of dynamic compression ofsmall airways on expiration.28

Similarly, positive end-expiratory pressure hasbeen shown to improve ventilation perfusion match-ing without adverse effects upon respiratory mechan-ics nor on haemodynamics in patients with severeCOPD intubated and mechanically ventilated29 andreduce inspiratory work of breathing.4–7

Previous groups have observed an improvement inarterial blood gases in patients with OSA. Sforza et al.reported an increase in PaO2 (69.9 to 72.8 mmHg) anda fall in PaCO2 (48.5 to 44.5 mmHg) but no change inlung function tests.30 The improvements we report aregreater, and may be explained by the greater severityof underlying lung disease in our population group.Reversal of nocturnal hypoxaemia and hypercapniaand ‘resetting’ of the chemoreceptor sensitivity arethought to be the likely mechanisms responsible forthe improvements in arterial blood gases.

This is the first report to our knowledge thatdescribes the use of long-term CPAP in patients withCOPD and sleep disordered breathing. Previous workhas concentrated on the effect of long-term use ofpressure set or volume cycled ventilators upon bloodgases and lung function in such patients. Given thatmost of the reports suggest a reduction in PaCO2

levels, there has not been a consistent increase in lungfunction. Meecham Jones et al. saw no change in lungfunction after 3 months of regular BiPAP® (Respiron-ics Inc., Murrysville, PA, USA) use, where EPAP levelsof 2–4 cmH2O were used.12 Although Mezzanotte etal.26 reported increases in respiratory muscle strength

368 D Mansfield and MT Naughton

Tab

le2

Co

nti

nu

ou

s p

osi

tive

air

way

pre

ssu

re (

CPA

P)

com

pli

ant

pat

ien

t d

ata

bef

ore

an

d a

fter

CPA

P t

reat

men

t

Pati

ent

CPA

PA

HI

SpO

2<

90%

SpO

2<

80%

FE

V1

(L)

FV

C (

L)P

aO2

PaC

O2

pH

no.

cmH

2O(e

ven

ts/h

)(%

TST

)(%

TST

)(m

mH

g)(m

mH

g)B

LF

-UP

BL

F-U

PB

LF

-UP

BL

F-U

PB

LF

-UP

BL

F-U

PB

LF

-UP

BL

F-U

P

1.10

01

100

113

01.

01.

293.

123.

0157

6563

437.

357.

392.

72

099

44

00.

480.

521.

321.

7059

6750

467.

437.

423.

106

236

12

01.

962.

003.

633.

3867

6953

487.

397.

424.

87

337

140

00.

861.

161.

591.

9272

7247

507.

397.

385.

1710

294

6176

30.

640.

881.

201.

5238

5466

537.

337.

436.

1010

275

501

00.

880.

962.

522.

4357

5650

457.

417.

407.

1010

371

4518

70.

880.

962.

723.

1242

6168

507.

487.

428.

1222

088

105

01.

001.

143.

103.

2759

6153

497.

437.

409.

942

3224

04

01.

201.

122.

282.

3861

6649

477.

397.

3910

.9

6019

9999

8931

0.72

0.92

0.87

1.07

3661

8149

7.30

7.40

Mea

n10

.216

.96.

4*72

3020

40.

951.

102.

242.

3854

.863

.258

.048

.07.

397.

41SE

M0.

96.

13.

49

1011

30.

130.

130.

280.

253.

81.

83.

50.

90.

030.

00P

0.02

50.

002

0.09

0.00

50.

075

0.01

50.

015

0.39

2

AH

I, a

pn

oea

–hyp

op

no

ea in

dex

; TST

, to

tal s

leep

tim

e; P

, lev

el o

f si

gnifi

can

ce b

etw

een

bas

elin

e (B

L) a

nd

fo

llow

up

(F

-UP

).

and functional capacity, they saw no change inmarkers of airflow obstruction after 4 weeks of regularnocturnal CPAP usage in eight stable COPD patientsgiven 5–8 cmH2O.26 It is possible that the differencesin lung function represent differences, in pressureand/or duration of treatment.

An interesting observation from this study was thereduction in hospital admission rates. Hospitalizationrates for COPD have never correlated well with sever-ity of airflow limitation nor is there evidence to indicate that any therapy alters the admission rate.31

The explanation for the finding in our study is likelyto be complex. One of the likely reasons is the greaterawareness the patient develops for his or her illness through the education process that inevitablyaccompanies the introduction of CPAP. Together with a more intensive follow-up programme thepatients were likely to be more aware of disease exacerbations and responding promptly to the earlysigns with antibiotics and steroids. Second, ourpatients described an improved sense of well-beingand a number reported a significant improvement in exertional tolerance (although not objectively measured in this study). This may have had a positive impact on the psychological health of ourpatients which is a well-known contributor to hospi-tal admission frequency. Our findings support previous observations of reduced hospitalizationwith reversal of sleep disordered breathing in patientswith COPD.32,33

Although no patient deteriorated with CPAP, four ofthe 14 patients were non-compliant with CPAP. Thismay have been due to less severe sleep disorderedbreathing. Alternatively, CPAP levels may have causedrecruitment of expiratory muscles and paradoxicallyincreased the expiratory work of breathing. We doubtthis to be the case because the mean level of CPAPused, 10 cmH2O, is a level generally thought to over-come intrinsic PEEP and to diminish inspiratory workof breathing.4–6 Alternatively, CPAP may have bron-chodilated and resulted in an increase in respiratorydeadspace, and lead to hypercapnia and increasedrespiratory rate.34

In summary, we report a reduction in PaCO2 and anincrease in PaO2 and FEV1 in a group of patients withsevere COPD and sleep disordered breathing follow-ing treatment with nocturnal CPAP treatment. Morerandomized controlled studies will be required toconfirm our findings and to remove the placeboeffects of medical attention.

ACKNOWLEDGEMENT

The authors wish to thank the staff of the sleep andrespiratory function laboratories.

REFERENCES

1 Heidra Y, Dekhuijzen P, van Herwaarden C, Folgering H.Nocturnal saturation and respiratory muscle function inpatients with chronic obstructive pulmonary disease.Thorax 1995; 50: 610–12.

2 McNicholas WT. Impact of sleep in respiratory failure.Eur. Respir. J. 1997; 10: 920–33.

3 Sampol G, Sagales MT, de Roca A, la Calzada MD, BofillJM, Morell F. Nasal continuous positive airway pres-sure with supplemental oxygen in coexistent sleepapnoea–hypopnoea syndrome and severe chronicobstructive pulmonary disease. Eur. Respir. Dis. 1996; 9:111–16.

4 Petrof BJ, Kimoff RJ, Levy RD, Cosio MG, Gottfried SB.Nasal continuous positive airway pressure facilitates res-piratory muscle function during sleep in severe chronicobstructive pulmonary disease. Am. Rev. Respir. Dis.1991; 143: 928–35.

5 Petrof BJ, Legare M, Goldberg P, Milic-Emili J, GottfriedSB. Continuous positive airway pressure reduces work ofbreathing and dyspnoea during weaning from mechan-ical ventilation in severe chronic obstructive pulmonarydisease. Am. Rev. Respir. Dis. 1990; 141: 281–9.

6 Katz JA, Markes JD. Inspiratory work with and withoutcontinuous positive airway pressure in patients withacute respiratory failure. Anaesthesiology 1985; 63:598–607.

7 Renson JP, DiMarco AF, Supinski GS. Respiratory musclerest using nasal BiPAP ventilation in patients with stablesevere COPD. Chest 1994; 105: 1053–60.

8 Robertson CH, Foster GH, Johnson RL. The relationship of respiratory failure to the oxygen con-sumption of, lactate production by, and distribution of blood flow among respiratory muscles during increas-ing inspiratory resistance. J. Clin. Invest. 1977; 59: 31–42.

9 Ballard R, Clover W, Suh B. Influence of sleep on respira-tory function in emphysema. Am. J. Respir. Crit. CareMed. 1995; 151: 945–51.

10 Carroll N, Braithwaite MA. Control of nocturnalhypoventilation by nasal intermittent positive airwaypressure ventilation. Thorax 1988; 43: 349–53.

11 Stumpf DA, Millman RP, Carlisle CC et al. Nocturnal positive-pressure ventilation via nasal mask in patientswith severe chronic obstructive pulmonary disease. Am.Rev. Respir. Dis. 1991; 144: 1234–9.

12 Meecham Jones D, Paul E, Jones P, Wedzicha J. Nasalpressure support ventilation plus oxygen compared withoxygen therapy alone in hypercapnic COPD. Am. J.Respir. Crit. Care Med. 1995; 152: 538–44.

13 Lin CC. Comparison between nocturnal nasal positivepressure ventilation combined with oxygen therapy andoxygen monotherapy in patients with severe COPD. Am.J. Respir. Crit. Care Med. 1996; 154: 353–8.

14 Elliot M, Simonds A, Wedzichs J, Branthwaite M. Domiciliary nocturnal nasal intermittent positive pres-sure ventilation in hypercapnic respiratory failure due tochronic obstructive lung disease: Effects on sleep andquality of life. Thorax 1992; 47: 342–8.

15 Leger P, Bedicam J, Cornette A et al. Nasal intermittentpositive pressure ventilation. Long term follow-up inpatients with severe chronic respiratory insufficiency.Chest 1994; 105: 100–5.

16 Simonds A, Elliot M. Outcome of domiciliary nasal inter-mittent positive pressure ventilation in restrictive andobstructive disorders. Thorax 1995; 50: 604–9.

17 Ferguson GT, Gilmartin M. CO2 rebreathing during BiPAPventilatory assistance. Am. J. Respir. Crit. Care Med. 1995;151: 1126–35.

CPAP on lung function 369

18 Martin JG, Shore S, Engel LA. Effect of continuous posi-tive airway pressure on respiratory mechanics andpattern of breathing in induced asthma. Am. Rev. Respir.Dis. 1982; 126: 812–17.

19 American Thoracic Society Statement. Guidelines forlung function testing: selection of reference values andinterpretative strategies. Am. Rev. Resp. Dis. 1991; 144:1202–18.

20 Rechstschaffen A, Kales A. A Manual for StandardizedTerminology, Techniques and Scoring System for SleepStages of Human Subjects: Publication No. 204. Washington, DC: National Institutes of Health, 1968.

21 Naughton MT, Liu PP, Benard DC, Goldstein RS, BradleyTD. Treatment of congestive heart failure and Cheyne-Stokes respiration during sleep by continuous positiveairway pressure. Am. J. Respir. Crit. Care Med. 1995; 151:92–7.

22 Carr DT, Essex HE. Certain effects of positive pressurerespiration on the circulatory and respiratory systems.Am. Heart J. 1946; 31: 53–73.

23 Barach A, Swenson P. Effect of breathing gases underpositive pressure on lumens of small and medium sizedbronchi. Arch. Int. Med. 1939; 63: 946–8.

24 Chan CS, Woolcock AJ, Sullivan CE. Nocturnal asthma:Role of snoring and obstructive sleep apnea. Am. Rev.Respir. Dis. 1988; 137: 1502–4.

25 Shivaram U, Miro AM, Cash ME, Finch PJP, Heurich AE,Kamholz SL. Cardiopulmonary responses to continuouspositive airway pressure in acute asthma. J. Crit. Care1993; 8: 87–92.

26 Mezzanotte W, Tangel D, Fox A, Ballard R, White D. Nocturnal nasal continuous positive airway pressure inpatients with chronic obstructive pulmonary disease.Chest 1994; 106: 1100–8.

27 O’Donnell D, Sanii R, Younes M. Improvement in exer-cise endurance in patients with chronic airflow limita-tion using continuous positive airway pressure. Am. Rev.Resp. Dis. 1988; 138: 1510–14.

28 O’Donnell D, Sanii R, Giesbrecht G, Younes M. Effect ofcontinuous positive airway pressure on respiratory sen-sation in patients with chronic obstructive pulmonarydisease during submaximal exercise. Am. Rev. Resp. Dis.1988; 138: 1185–91.

29 Rossi A, Santos C, Roca J, Torres A, Felez MA. Effects ofPEEP on VA/Q mismatching in ventilated patients withchronic airflow obstruction. Am. J. Respir. Crit. Care Med.1994; 149: 1077–84.

30 Sforza E, Krieger J, Weitzenblum E, Apprill M, LampertE, Ratamaharo J. Long-term effects of treatment withnasal continuous positive airway pressure on daytimelung function and pulmonary hemodynamics inpatients with obstructive sleep apnea. Am. Rev. Respir.Dis. 1990; 141: 866–70.

31 American Thoracic Society Statement. Standards for thediagnosis and care of patients with chronic obstructivepulmonary disease. Am. J. Respir. Crit. Care Med. 1995;152: S78–S121.

32 Kryger MH, Roos L, Delaive K, Walld R, Horrocks J. Utilization of health care services in patients with severeobstructive sleep apnea. Sleep 1996; 19: S111–S116.

33 Peker Y, Hedner J, Johansson A, Bende M. Reduced hos-pitalisation with cardiovascular and pulmonary diseasein obstructive sleep apnea patients on nasal CPAP treat-ment. Sleep 1997; 20: 645–53.

34 Folkow B, Pappenheimer J. Components of the respira-tory dead space and their variation with pressure breath-ing and with bronchoactive drugs. J. Appl. Phys. 1962; 8:102–10.

370 D Mansfield and MT Naughton