Thorax 1995;50:604-609

Outcome of domiciliary nasal intermittentpositive pressure ventilation in restrictive andobstructive disorders

A K Simonds, M W Elliott*

RespiratorySupport Service,Department ofRespiratory Medicine,Royal BromptonHospital, LondonSW3 6NP, UKA K SimondsM W Elliott*

Reprint requests to:Dr A K Simonds.

*Current address:St James'sUniversity HospitalLeeds, UK.

Received 25 July 1994Returned to authors12 October 1994Revised version received29 December 1994Accepted for publication2 February 1995

AbstractBackground - Nasal intermittent positivepressure ventilation (NIPPV) is a newtechnique which has rapidly supplantedother non-invasive methods of ventilationover the last 5-10 years. Data on its effect-iveness are limited.Methods - The outcome oflong term dom-iciliary NIPPV has been analysed in 180patients with hypercapnic respiratory fail-ure predominantly due to chest wallrestriction, neuromuscular disorders, orchronic obstructive lung disease. One hun-dred and thirty eight patients were startedon NIPPV electively, and 42 following anacute hypercapnic exacerbation. Outcomemeasures were survival (five year prob-ability of continuing NIPPV), pulmonaryfunction, and health status. A crossoverstudy from negative pressure ventilationto NIPPV was carried out in a subgroupof patients.Results - Five year actuarial probabilityof continuing NIPPV for individuals withearly onset scoliosis (n = 47), previouspoliomyelitis (n = 30), following tuber-culous lung disease (n = 20), generalneuromuscular disorders (n = 29), andchronic obstructive pulmonary disease(n = 33) was 79% (95% CI 66 to 92), 100%,94% (95% CI 83 to 100), 81% (95% CI 61to 100), 43% (95% CI 6 to 80), respectively.Most of the patients with bronchiectasisdied within two years. One year after start-ing NIPPV electively the mean (SD) Pao2compared with the pretreatment value was+1-8 (1.9) kPa, mean PacO2 -1P4 (1.3)kPa in patients with extrapulmonary re-strictive disorders, and Pao2 +0-8 (1.0)kPa, PacO2 - 0 9 (0 8) kPa in patients withobstructive lung disease. Arterial bloodgas tensions improved in patients trans-ferred from negative pressure ventilationto NIPPV. Health status was ranked high-est in patients with early onset scoliosis,previous poliomyelitis, and followingtuberculous lung disease. In the group asa whole health perception was comparableto outpatients with other chronic dis-orders.Conclusions - The long term outcome ofdomiciliary NIPPV in patients withchronic respiratory failure due to scoliosis,previous poliomyelitis, and chest wall andpulmonary disease secondary to tuber-culosis is encouraging. The results ofNIPPV in patients with COPD and pro-

gressive neuromuscular disorders showbenefit in some subgroups. The outcomein end stage bronchiectasis is poor.(Thorax 1995;50:604-609)

Keywords: NIPPV, domiciliary ventilation, chronicrespiratory failure.

Home ventilation is one of the earliest formsof technological support to be applied in thecommunity. The methods of providing ventil-atory assistance for individuals with respiratoryinsufficiency have changed over the last 50years.' Negative pressure systems (cuirass,pneumojacket, iron lung) were introduced atthe time of the poliomyelitis epidemics in the1940s and the delivery of ventilation via atracheostomy has been used since the 1950s.Nasal intermittent positive pressure ventilation(NIPPV) is the most recent advance in the fieldof non-invasive ventilatory support, enteringclinical practice in the mid 1980s. As a newform of therapy, NIPPV has been assessedin respect of its short term effects in acutehypercapnic exacerbations of chronic lung dis-ease2-5 and in patients unable to wean fromconventional ventilation.6 However, there islittle information on the impact of NIPPV onsurvival,7 the quality of life,8 and functionallevel in individuals with chronic ventilatoryinsufficiency.

MethodsSUBJECTSPatients were classified by primary diagnosis.Of 180 patients who were started on NIPPVat the Royal Brompton Hospital over the period1987-92, 47 had early onset idiopathic sco-liosis, 30 had previous poliomyelitis, 20 oldtuberculous lung disease with associated chestwall lesions (thoracoplasty, artificial pneumo-thorax, phrenic nerve crush), 29 had neuro-muscular disorders (eight muscular dystrophy,eight -central hypoventilation, five myopathies,four spina bifida, two polyneuropathies, onespinal muscular atrophy, one motor neuronedisease), 33 had chronic obstructive pulmonarydisease (COPD), and 13 bronchiectasis. A re-maining group comprised eight patients withmiscellaneous disorders (three obesity/hypo-ventilation, two scleroderma, one Proteus syn-drome, one osteogenesis imperfecta, one chylo-thorax). Patient characteristics are given intable 1.The indication for domiciliary NIPPV was

symptomatic, chronic hypercapnic respiratory

604

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from

Domiciliary NIPPV in restrictive and obstructive disorders

Table 1 Mean (SD) age, baseline arterial blood gas tensions and spirometric values of 180 patients in study

Age (years) Pao2 (kPa) Paco2 (kPa) FVC (7) FEV, (7)

Early onset scoliosis (n=47) 49.2 (12 2) 6-6 (1-6) 8 1 (1-4) 0-83 (0-4) 0 64 (0 2)Previous polio (n=30) 51 (10-6) 7-1 (1-4) 7-9 (1 0) 1-1 (0 4) 0-80 (0 2)Previous tuberculosis (n =20) 61 ( 8 4) 6-0 (1 2) 8-4 (1-5) 1-2 (0-6) 0-80 (0 3)Neuromuscular disorders (n =29) 36 (14-4) 7 5 (1-8) 8-4 (1 7) 1 3 (0 7) 0 98 (0 7)COPD (n=33) 574 ( 84) 61 (08) 83 (18) 16 (07) 058 (03)Bronchiectasis (n= 13) 41 (14 4) 5-6 (1 7) 9 1 (1-7) 1 0 (0-5) 0 48 (0 16)Miscellaneous (n=8) 36 ( 9 9) 7-1 (2-0) 8-6 (1 0) 1 7 (1 5) 1-6 (1 3)

Pao2, Paco2, and spirometric measurements available in 170 patients (arterial blood gas data are not included where measurements were made with the patientbreathing supplemental oxygen).

failure which was unresponsive to standardmeasures, including controlled oxygen therapy.Nocturnal hypoventilation was demonstratedby overnight monitoring of arterial oxygen sat-uration and transcutaneous carbon dioxide ten-sion. One hundred and thirty eight patientswere treated electively with NIPPV after a pro-gressive deterioration in symptoms and arterialblood gas tensions, and 42 received treatmentfollowing an acute hypercapnic exacerbationor were weaned from conventional ventilationusing NIPPV. Among the group treatedelectively 15 patients were transferred from neg-ative pressure ventilation (10 cuirass, threepneumojacket, two portable iron lung) whenNIPPV became available at this institution in1987. NIPPV was used as a "bridge" to trans-plantation9 in six subjects in the acute group.At the start of NIPPV all patients had normalor near normal bulbar function, and those re-ceiving NIPPV electively were able to breathespontaneously for at least 12 hours/day.Patients were recruited over a five year periodfrom 1987-92.

VENTILATORY EQUIPMENT AND MONITORINGOne hundred and seventy patients used volumepreset ventilators (BromptonPAC, PneupacLtd; PLV 100, Lifecare; Monnal D, Deva Med-ical), and 10 used pressure preset equipment(BiPAP, Respironics Inc). All ventilators wereoperated in assist/control mode. Patients used

0 1 2 3 4 5

Years

Figure 1 Probability of continuing nasal intermittent positive pressure ventilation(NIPPV) in different diagnostic groups. KS= kyphoscoliosis; polio =poliomyelitis;COPD= chronic obstructive pulmonary disease; TB = tuberculosis.

commercial silicone nasal masks (RespironicsInc or Rescare Ltd).

Initial ventilator settings were determined byovernight monitoring of arterial oxygen sat-uration and transcutaneous carbon dioxide ten-sion, and subsequently modified as necessaryaccording to clinical course. Supplementaloxygen was entrained into the nasal mask at1-2 1/min in individuals in whom arterial Sao2remained below 90% for most of the nightwhen they were ventilated with air. Each patientwas instructed to use NIPPV during sleep.Humidification using a heat and moistureexchanger (Portex Thermovent) was addedintermittently in 10 patients, predominantlythose with bronchiectasis.

After starting NIPPV as inpatients, patientswere initially reviewed as outpatients every 2-3months and subsequently at 6-12 month in-tervals. Arterial blood gas tensions and spiro-metric parameters were assessed at eachoutpatient attendance. All measurements ofarterial blood gas tensions were made with thepatient breathing air spontaneously during theday.

QUALITY OF LIFEHealth status was assessed on one occasion ina cross section of all patients currently usingNIPPV in 1993. The SF-36 postal ques-tionnaire was sent to 116 patients. Each re-ceived one follow up telephone call. The SF-36 measure of health status assesses physicalfunction, role limitation caused by physicalor mental problems, social function, mentalhealth, energy and vitality, pain and healthperception. 0 Additional questions were in-cluded on sleep quality and the side effectsof NIPPV. The questionnaire was completedanonymously and coded by diagnostic group.Data from the SF-36 questionnaire were com-pared with UK population norms of similarage derived from general practice," and with aUS study of outpatients with chronic medicalconditions (diabetes, hypertension, recentmyocardial infarction, cardiac failure).12

COMPLIANCECompliance with NIPPV was monitored in 43randomly selected patients using volume presetventilators by means of a hidden clock. Thetiming mechanism was activated only whenthe ventilator compressor was operating, andtherefore measured hours ofventilator functionrather than electrical run time.

100

80

0-

Z 600,c

.

C

: 400

20,

O

605

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from

Simonds, Elliott

Table 2 Duration of nasal intermittent positive pressure ventilation (NIPPV) and reasons for withdrawal

Median (range) duration Reasons for withdrawalofNIPPV (years)

Early onset scoliosis 3-2 (0-1-6-9) Seven died (5 respiratory failure, 1 carcinoma breast1 carcinoma ovary). One discontinued- improved

Previous polio 4-3 (066-64) No withdrawalsPrevious tuberculosis 2-1 (0-1-7) One died (leukaemia)Neuromuscular disorders 2-8 (0-3-5 9) Four died (respiratory failure due to progression of underlying

disorder)COPD 1-6 (0-1-6-3) Six died (respiratory failure). Five withdrawn (poor tolerance

NIPPV)Bronchiectasis 0 7 (0-1-2-3) Seven died (respiratory failure). Two withdrawn (heart/lung

transplantation)Miscellaneous 1-6 (0 2-48) One died (respiratory failure)

STATISTICAL ANALYSISThe Kaplan-Meir technique was used tactuarial curves for the probability oftinuing NIPPV. Arterial blood gas tensioispirometric values before and after initialelective NIPPV in the groups with restdisorders and COPD were comparedStudent's t test. Patients beginning NIPan acute measure, during weaning, orbridging to transplantation were excludecthis analysis as blood gas tensions were unat the start of treatment. Responses to t136 questionnaire were examined by diaggroup using ANOVA (Minitab Inc).

ResultsCONTINUATION OF NIPPV (fig 1)The five year actuarial probability ofcontiNIPPV (mean, 95% confidence intervEearly onset scoliosis was 79% (66 to 92)poliomyelitis was 100%, for previousculosis 94% (83 to 100), neuromusculEease 81% (61 to 100), COPD 43% (6 toIn the bronchiectasis group seven patientafter less than two years of treatment. C26 of 180 patients died (23 respiratory fthree unrelated disease), NIPPV wasdrawn in six (five poor tolerance, onproved), and two bronchiectatic pireceived heart-lung transplants. Mediaration ofNIPPV and the reasons for with(from therapy in each diagnostic group are

n =34

n =2

n = 99

-o plotf con-ns andtion ofrictiveusing'PV aswheni fromistablele SF-nostic

in table 2. All patients transferred from negativepressure ventilation to NIPPV expressed a pref-erence to continue with NIPPV long term.Mean (SD) NIPPV use per 24 hours was

7-88 (2 1) hours with 16 of 43 patients usingit for short periods in the day as well as duringsleep at night. Supplemental oxygen therapywas used in 8-5% of patients with idiopathicscoliosis, 3-3% of post poliomyelitis patients,30% of the previous tuberculosis group, 48%of patients with COPD, and in all patients withbronchiectasis.

ARTERIAL BLOOD GAS TENSIONS ANDSPIROMETRIC VALUESArterial blood gas tensions improved in allpatient groups. One hundred and eight patients

inuing with extrapulmonary restrictive disorders andal) for 30 patients with COPD received NIPPV elect->, after ively. Arterial blood gas data at the start oftuber- NIPPV were available in 99 of the 108 patientsar dis- with restrictive disorders and in all 30 patients80%). with COPD. After one year of elective NIPPVts died mean (SD) Pao2 compared with initial values)verall increased by 1-8 (1.9) kPa (p<0Q0001) andailure, Paco2 fell by 1-4 (1-3) kPa (p<0Q0001) inwith- patients with extrapulmonary restrictive dis-

ie im- orders (idiopathic scoliosis, neuromuscularatients disease including poliomyelitis, and postn dur- tuberculous disease). Mean (SD) Pao2 rose bydrawal 0-8 (1.0) kPa (p<0-01) and Paco2 decreasedgiven by 0 9 (0 8) kPa (p<0'001) in the group with

COPD. The trends in arterial blood gas ten-sions over five years of NIPPV in the groupwith extrapulmonary restriction are shown infig 2. In the subgroup of patients transferredfrom negative pressure ventilation to NIPPVthe best levels of arterial blood gas tensionsachieved during the period of negative pressureventilation were Pao2 7-5 (1-2) kPa, Paco27-1 (1.2) kPa. Values for this group improved

20 significantly after using NIPPV for an averageof 38 (17-7) months to Pao2 9-6 (1-4) kPa

[Pao2l (p<0-001), Paco2 6-3 (0 7) kPa (p<0.005).FVC and FEV, did not change significantly

over time in any patient group.

[Paco2]

1 2 3 4 5

Years

Figure 2 Mean (SE) arterial blood gas tensions over five years in patients withrestrictive disorders who started nasal intermittent positive pressure ventilation for electivereasons.

HEALTH STATUSThe response rate to the SF-36 questionnairewas 90 5% (105 of 116). The mean (SD)age of the respondents was 54-2 (13-5) years;duration ofNIPPV use in these patients rangedfrom six months to seven years. Results aregiven in fig 3 (where a score of 100% representsnormal function or no pain, and 0% indicates

(Ua-

8

0,e I*f,

606

10 F

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from

Domiciliary NIPPV in restrictive and obstructive disorders

L NIPPVW= Chronic disease (USA)

| Normals (HLS, UK)I... ...._..._..... i

100

80

60

40IL

20K

HP PF RLP RLM SF MH EV PAIN

SF-36 indices

Figure 3 Health status measured by SF-36 questionnaire in patients receiving nasal intermittent positive pressureventilation (NIPPV) in a USA group with chronic diseases and in UK normal subjects (Healthy Life Survey (HLS),Oxford). HP= general health perceptions; PF=physical functioning; RLP= role limitations due to physical problems;RLM= role limitations due to emotional problems; SF= social functioning; MH= general mental health; EV= energylvitality.

lowest possible value or maximal pain). Phys-ical function was significantly higher in groupswith idiopathic scoliosis, after poliomyelitis,and previous tuberculosis compared with theCOPD group. Patients with neuromusculardisorders and COPD had a similar level ofphysical function, but role limitation due tophysical factors was greater in the COPD group(65% versus 18-7%, p<0-01). Health statusoutcome in the patients receiving NIPPV wascomparable to the group with chronic disordersreported from the USA. Although physicalfunction was reduced compared with the UKpopulation norms, mental health and energyand vitality were similar.

Sleep quality was rated poor in 5% ofpatients, average in 67%, and very good in21 %. There was no correlation between hoursof ventilator use and sleep quality. The com-monest reported disadvantages ofNIPPV wereinconvenience (20 patients), nasal/mask prob-lems (10), gastric distension (three), and noise(three).

0 1 2 3 4 5Years

Figure 4 Percentage survival in patients with COPD receiving NIPPV compared withdata from NOTT trial (oxygen 19 hours/day), MRC trial (oxygen 15 hourslday) andcontrol group from MRC trial.

DiscussionAs nocturnal hypoventilation is the commondenominator in patients with chest wall dis-orders, neuromuscular conditions, and chronicobstructive lung disease who progress to corpulmonale and premature death, it is logical toexamine the effects of long term correctionof nocturnal arterial blood gas tensions withNIPPV on outcome. There have been no longterm controlled studies of domiciliary NIPPVin chronic ventilatory failure. In patients withhypercapnic respiratory failure secondary torestrictive chest wall disorders such as idio-pathic scoliosis, previous poliomyelitis andtuberculosis, premature death'3 14 is an in-evitable consequence if chronic hypoxaemiaand hypercapnia are not reversed. The five yearprobability of continuing NIPPV is a closeapproximation to five year survival in thesegroups, as death was the principal cause ofwithdrawal from the study. Furthermore,deaths in these subgroups were largely due toconcurrent disease (malignancy) or progressionof pulmonary hypertension in patients whopresented late for NIPPV. With a five yearsurvival figure of around 80% or higher, andevidence that withdrawal of NIPPV results ina recurrence of nocturnal hypoventilation,'5 arandomised controlled study would appear un-ethical in patients with idiopathic scoliosis, pre-vious poliomyelitis, and old tuberculous lungdisease.'6 Leger et al' have reported data froman uncontrolled French multicentre study ofNIPPV in groups with scoliosis, previous tuber-culosis, Duchenne muscular dystrophy,COPD, and bronchiectasis. Although similarventilatory methods were used, the Frenchpatients differ from the population in the cur-rent study in that they tended to have a higherPo2 and lower Pco2 on initiation of NIPPV, atleast 60% of the French patients with scoliosisreceived supplemental oxygen therapy, and the

0'F

CP

0)co40~

607

00X

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from

Simonds, Elliott

follow up period was shorter (36 months). Des-pite these differences the probability of con-tinuing NIPPV at 36 months in the Frenchstudy is similar to the results presented here.

Control of diurnal arterial blood gas tensionswas maintained over four years following elect-ive initiation of NIPPV, although there is asuggestion ofsome deterioration after this time.As greater numbers of patients are followed,these trends should become clearer. Arterialblood gas values were improved by NIPPV inthe patients who were transferred from negativepressure ventilation, although it should beborne in mind that a direct comparison of theefficacy of these techniques was not made.Negative pressure systems often provoke upperairways obstruction during sleep'7 and cancause chest wall discomfort. As a result, NIPPVappears to be better tolerated by patients andis usually easier to institute.

Patients with end stage COPD differ fromthose with extrapulmonary restrictive disordersas, in addition to alveolar hypoventilation, gasexchange is also severely impaired. Long termoxygen therapy (LTOT) is the mainstay oftreatment of respiratory failure in COPD, andhas been shown to produce improved survivalin those using oxygen for 15 or more hours aday.'8 However, there are problems associatedwith LTOT, including the fact that no con-sistent increase in quality of life has been dem-onstrated in recipients, compliance withtherapy may be poor, and a small proportionof patients with COPD are unable to toleratecontrolled oxygen therapy due to progressiveand/or symptomatic hypercapnia. A controlledtrial ofdomiciliary negative pressure ventilationused with the aim of "resting" the respiratorymuscles,'9 and a crossover study of pressuresupport ventilation20 in end stage COPD,both produced no overall benefit in terms ofsymptom control or physiological measures.By contrast, Elliott et al2' demonstrated animprovement in nocturnal and diurnal arterialblood gas tensions in a pilot study ofdomiciliaryNIPPV over one year in hypercapnic COPDpatients. This beneficial effect was attributedto an increase in ventilatory drive and reductionin ventilatory load. No change in respiratorymuscle function was seen.22 Severely hyper-capnic patients and those with symptomaticsleep disturbance showed the most im-provement in nocturnal and diurnal arterialblood gas tensions. Individuals with end stageemphysema were unlikely to derive benefit.The patients with COPD included in the pres-ent study represent the most severe end of thespectrum of respiratory failure in COPD (meanPco2 on entry 8-3 kPa) and most were de-teriorating despite long term oxygen therapy.Hypercapnic patients have a poor prognosis.'8Although withdrawal from NIPPV due to in-tolerance was higher in the patients with COPDthan in the other groups, the probability ofcontinuing NIPPV at five years of43% is com-parable to the survival results of long termoxygen therapy in previous studies (fig 4).182324Such retrospective comparisons with un-matched populations are unsatisfactory, but theresults of this study suggest that long term

domiciliary NIPPV is a practical propositionin carefully selected patients with end stageCOPD who are deteriorating despite long termoxygen therapy. However, self-reported healthstatus is less good than in those with restrictivedisorders.The heterogenous nature of the neuro-

muscular and miscellaneous groups makes itdifficult to draw conclusions on the outcomein different diseases, including those with pro-gressive neuromuscular disorders, multiplehandicaps (for example, spina bifida) and rareconditions (for example, scleroderma, osteo-genesis imperfecta). However, the outlookin individuals with stable neuromuscular dis-ease appears good. Sawicka et al25 showed thatthe survival in patients with progressive neuro-muscular disease receiving domiciliary negativepressure ventilation was largely related to theprogress of the underlying condition, and thisis likely to remain the case regardless of ventil-atory technique. Quality of life or health statusis the most appropriate outcome measure inthese patients. In this study no patient electedto transfer from NIPPV to ventilation via atracheostomy, although only one individual(with motor neurone disease) developed severebulbar problems.

In the bronchiectasis group the probabilityof continuing NIPPV after two years was lessthan 20%, and most patients became in-creasingly more ventilator-dependent withtime. All were extremely unwell at the start ofNIPPV, having been referred for considerationfor heart-lung transplantation. Only two re-ceived transplants - in the remainder eitherorgans were unavailable or patients were noteligible for the transplant programme becauseof comorbidity. The response to NIPPV in thisgroup is disappointing and worse than hasbeen shown in a previous study.7 However,our patients had more severe hypoxaemia andhypercapnia at the start of NIPPV suggestingthat ventilation was introduced later in thenatural history of the disease.

Several conclusions can be drawn from thisopen study. Domiciliary NIPPV is generallywell tolerated and associated with excellent fiveyear survival in patients with extrapulmonaryrestrictive disorders. The SF-36 was ad-ministered on one occasion after a variableperiod of NIPPV and therefore cannot givean indication of the impact of the treatment,merely a single view of health status at the timeof measurement. Despite this major limitationit can be seen that, although physical functionis reduced compared with normal subjects,levels ofsocial function and self-reported healthstatus contradict the commonly held belief thatdomiciliary ventilator-dependent patients ex-perience a severely limited lifestyle and poorquality of life. The outcome in patients with endstage, hypercapnic COPD who fail to tolerateLTOT is encouraging, and suggests that con-trolled studies are warranted. A European trialcomparing LTOT and an LTOT/NIPPV com-bination is in progress. For groups with pro-gressive disorders NIPPV can producesymptom relief and a satisfactory quality of life.There are important ethical issues involved in

608

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from

Domiciliary NIPPV in restrictive and obstructive disorders

carrying out controlled trials in these patients.In the first instance more data are required onthe outcome of NIPPV in progressive neuro-muscular disease, and in other conditionswhich affect relatively small numbers ofpatients. This should be facilitated by a nationalregister of home ventilator patients which hasrecently been established by the British Thor-acic Society.

We are indebted to Dr M A Branthwaite who established theRespiratory Support Service at the Royal Brompton Hospital,and to Ms Nikki Potter, Senior Respiratory Support technician,for data collection and technical skills.

1 Muir J-F. Home mechanical ventilation. Thorax 1993;48:1264-73.

2 Brochard L, Isabey D, Piquet J, Piedade A, Mancebo J,Messadi A-A, et al. Reversal of acute exacerbations ofchronic obstructive lung disease by inspiratory assistancewith a face mask. N Engl Med 1990;323:1523-30.

3 Elliott MW, Steven MH, Phillips GD, Branthwaite MA.Non-invasive mechanical ventilation for acute respiratoryfailure. BMJ7 1990;300:358-60.

4 Bott J, Carroll MP, Conway JH, Keilty SE, Ward EM,Brown AM, et al. Randomised controlled trial of nasalventilation in acute ventilatory failure due to chronicobstructive airways disease. Lancet 1993;341:1555-7.

5 Meduri GU, Conoscenti CC, Menashe P, Nair S. Non-invasive face mask ventilation in patients with acute re-spiratory failure. Chest 1989;95:865-70.

6 Udwadia ZF, Santis GK, Steven MH, Simonds AK. Nasalventilation to facilitate weaning in patients with chronicrespiratory insufficiency. Thorax 1992;47:715-8.

7 Leger P, Bedicam JM, Cornette A, Reybet-Degat 0, Lange-vin B, Polu JM, et al. Nasal intermittent positive pressureventilation. Long term follow-up in patients with severechronic respiratory insufficiency. Chest 1994;105:100-5.

8 Pehrsson K, Olofson J, Larsson S, Sullivan M. Quality oflife in patients treated by home mechanical ventilationdue to restrictive ventilatory disorders. Respir Med 1994;88:21-6.

9 Hodson ME, Madden BP, Steven MH, Tsang VT, YacoubMH. Non-invasive mechanical ventilation for cystic fib-rosis patients - a potential bridge to transplantation. EurRespirJ7 1991;4:524-7.

10 Ware JE. Measuring patients' views: the optimum outcomemeasure. SF-36: a valid, reliable assessment of healthfrom a patient's point of view. BMJ3 1993;306:1429-30.

11 Brazier JE, Harper R, Jones NMB, O'Cathain A, ThomasKH, Usherwood T, et al. Validating the SF-36 healthsurvey questionnaire: new outcome measure for primarycare. BMJ 1992;305:160-4.

12 McHorney CA, Ware JE, Raczek AE. The MOS 36-itemshort-form health survey (SF-36): II. psychometric andclinical tests of validity in measuring physical and mentalhealth constructs. Med Care 1993;31:247-63.

13 Freyschuss V, Nilsonne U, Lundgren KD. Idiopathic scoli-osis in old age. 1. Respiratory function. Arch Med Scand1968;184:365.

14 Piper AJ, Sullivan CE. Sleep-disordered breathing in neuro-muscular disease. In: Saunders NA, Sullivan CE, eds.Sleep and breathing. New York: Marcel Dekker, 1994:765-8.

15 Hill NS, Eveloff SE, Carlisle CC, Goff SG. Efficacy ofnocturnal nasal ventilation in patients with restrictive thor-acic disease. Am Rev Respir Dis 1992;145:365-71.

16 Hill N. Noninvasive ventilation. Does it work, for whom,and how? Am Rev Respir Dis 1993;147:1050-5.

17 Levy RD, Bradley TD, Newman SL, Macklem PT, MartinJG. Negative pressure ventilation. Effects on ventilationduring sleep in normal subjects. Chest 1989;95:95-9.

18 Medical Research Council Working Party Report. Longterm domiciliary oxygen therapy in chronic hypoxic corpulmonale complicating chronic bronchitis and em-physema. Lancet 1981;i:681-5.

19 Shapiro SH, Ernst P, Gray-Donald K, Martin JG, Wood-Dauphinee S, Beaupre A, et al. Effect of negative pressureventilation in severe chronic obstructive pulmonary dis-ease. Lancet 1992;340:1425-9.

20 Strumpf DA, Millman RP, Carlisle CC, Gratton LM, RyanSM, Erickson AD, et al. Nocturnal positive-pressure vent-ilation via nasal mask in patients with severe chronicobstructive pulmonary disease. Am Rev Respir Dis 1991;144:1234-9.

21 Elliott MW, Simonds AK, Carroll MP, Wedzicha JA, Bran-thwaite MA. Domiciliary nocturnal nasal intermittentpositive pressure ventilation in hypercapnic respiratoryfailure due to chronic obstructive lung disease: effects onsleep and quality of life. Thorax 1992;47:342-8.

22 Elliott MW, Mulvey DA, Moxham J, Green M, BranthwaiteMA. Domiciliary nocturnal nasal intermittent positivepressure ventilation in COPD: mechanisms underlyingchanges in arterial blood gas tensions. Eur Respir J 1991;4:1044-52.

23 Nocturnal Oxygen Therapy Trial Group. Continuous ornocturnal oxygen therapy in hypoxaemic chronic ob-structive lung disease, a clinical trial. Ann Intern Med 1980;93:391-8.

24 Cooper CB, Waterhouse J, Howard P. Twelve year clinicalstudy of patients with hypoxic cor pulmonale given longterm domiciliary oxygen therapy. Thorax 1987;42: 105-10.

25 Sawicka EH, Loh L, Branthwaite MA. Domiciliary ventil-atory support; an analysis of outcome. Thorax 1988;43:31-5.

609

on Decem

ber 9, 2020 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.50.6.604 on 1 June 1995. Dow

nloaded from