quality measurement of the - thoraxmeasurement ofhealth related quality oflife...

Thorax 1991;46:676-682

Quality of life measurement for patients withdiseases of the airways

P W Jones

Treatment for diseases of the airways islargely palliative, being directed in the maintowards the reduction of acute exacerbationsand limitation of the impact of disease ondaily life. For the purposes of measurement,the latter may be regarded as a series ofdifferent activities, some of which are listed inthe right hand box of figure 1. Disturbance ofthese activities may result from a range ofpathophysiological disorders, which mediatetheir effects through a limited number ofsymptoms. Despite the palliative nature oftreatment for airways disease, relatively littleregard has been paid to the activities listed inthe right hand box of figure 1 as possiblemeasures of therapeutic outcome. It is gen-erally assumed that a measured improvementin the disease processes shown on the left offigure 1 will be accompanied by improvedhealth and wellbeing. This assumption isbased on the belief that the arrows in thisfigure imply a direction of causality that holdsboth during the development of a disease andwith the response to treatment. In fact, thereis surprisingly little scientific evidence thattreatment improves the overall state ofpatients with a chronic disease. Before weconsider treatment, the processes by whichdisease activity in the lungs leads to the dis-ruption of daily living are worth examining.The disturbances in physiological functionlisted in the left hand box of figure 1 are welldocumented. There is also a growing body ofliterature describing measured disturbances ofthe activities listed in the right hand box. Buthow are these disturbances linked and whatare the quantitative relationships betweenthem?To illustrate this problem, it is worth look-

ing more closely at the effects of breathless-ness. Figure 1 suggest a simple linearsequence, but the scheme in figure 2 is morerealistic. This model is not complete, and Imake no claim that it provides profoundinsights; but there is experimental evidencefor most of the links between the boxes, andother links are reasonable hypotheses based oncurrent knowledge. The pathway is not tidy;some arrows linking two boxes are bidirec-tional and some paths form loops. The loopsare important because they allow positivefeedback, which then enables the system tobecome autonomous. Even if the precipitatingevent were entirely corrected, the resultingdisturbances would persist and only decay at arate fixed by the time constants of the system'sindividual components. An obvious exampleis the muscle wasting that results from

reduced physical activity. In a study of 152patients with airways obstruction who per-formed a 10 minute paced step test, 45 stop-ped because of breathlessness, whereas 73stopped because of factors such as fatigue andweak legs (C M Baveystock, P W Jones,unpublished observations). Some pathwaysillustrated in figure 2 may even be irrevers-ible-loss of employment, for example. Thescheme in figure 2 is presented hierarchically.At each level additional factors are fed in andinteract with signals coming up from below.These factors may be unrelated to the basicdisease, yet may modulate the signal producedby the underlying pathological process quiteprofoundly. Without knowledge of the operat-ing characteristics of the individual processesand information on the size of the outsideinfluences, we could not predict the magnitudeof the disturbances at the top of this systemfrom a knowledge of a disturbance at thebottom, no matter how accurately it wasmeasured.

Breathlessness: the link between lungdisease and disabilityQuality of life measurement is concerned withthe events at the top of figure 2, but someintermediate steps must be considered.Breathlessness is the critical link between lungdisease and ensuing disability, yet the factorsthat determine it are poorly understood.Studies during standardised ergometer exer-cise tests in normal subjects have shown widevariations between individuals in the intensityof perceived dyspnoea in relation to ventila-tion.'13 The reasons for this diversity are stillunknown. There is also evidence that the levelof distress that normal subjects associate withtheir breathlessness during exercise isunrelated to their perception of its intensity.4Comparable studies in patients have not beenperformed and we do not know which aspectsof breathlessness limit daily activity. The gen-eration of breathlessness is a critical process inthe production of impaired exercise tolerance,but currently we can neither predict anindividual patient's level of breathlessness norquantify the effect of changes in lung functionon it.

Spirometry, exercise tolerance, anddisabilityIn patients with obstructive airways disease thecorrelation between the results of spirometryand walking distance are poor.56 Even with

Division ofPhysiologicalMedicine, St George'sHospital MedicalSchool, LondonSW17 OREP W JonesReprint requests to:Dr Jones

676

on May 29, 2020 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.46.9.676 on 1 S

eptember 1991. D

ownloaded from

http://thorax.bmj.com/

Quality of life measurement for patients with diseases of the airways

--PATHOPHYSIOLOGYMucus hypersecreationAirway obstructionLung overinflationCompliance changesV/Q mismatchPulmonary hypertensionRight ventricular failurePolycythaemia

Figure 2 Pathways linking disturbances arising in the lungs to factors that may affectpatients' quality of life.

symptom limited ergometer stress tests, themaximum achievable exercise ventilation can-

not be reliably predicted from FEV,.78 Formalexercise tests provide an estimate of patients'physiological impairment. Disability is a more

subjective measure, concerned with the man-

ner in which daily activity is restricted byphysiological impairment. Measures of dis-ability are usually relatively short and simplequestionnaires or scales. The best knownexample in chest medicine is the MedicalResearch Council (MRC) dyspnoea scale,9 butothers used in obstructive airways diseaseinclude the oxygen cost diagraml'; the Borgratio of perceived exertion"; and the Mahlerdyspnoea index.'2 The results of spirometryhave been shown to correlate relatively poorlywith such measures,"1-14 whereas walkingdistance correlates quite well."""2'5 Theseobservations are interesting and important.The most plausible explanation is thatvariability between individuals in the percep-

tion of breathlessness is a more importantfactor than airflow limitation in determining theimpact of lung disease on patients' exercisetolerance and disability. Support for this con-clusion comes from a study of asthmaticpatients. Those patients who showed thelowest level ofbreathlessness in relation to theirventilation during cycle ergometer exercise alsohad the best walking distance performance.This association could not be explained by theseverity or variability of the patients' asthma.'6In summary, there is clear evidence that walk-ing distance and subjective disability are wellcorrelated and probably quite closely linked. Incontrast, links between the pathophysiologicalprocesses measured by spirometry and thepatients' breathlessness, physiological impair-ment, and disability appear to be tenuous. Wemay predict therefore that subjectively per-ceived impairment of health is likely tocorrelate poorly with spirometric measure-

ments of airway function.

General health questionnairesDisturbance to health may be viewed as a seriesof impacts on daily life and wellbeing.Measurement of health related quality of lifecan therefore be defined as "quantification oftheimpact of disease on a patient's life and per-

ceived wellbeing in a formal and standardisedmanner" (the italics emphasise my own

prejudice). The most comprehensivemeasurements of quality of life currentlyavailable are provided by general health ques-tionnaires such as the Quality of WellbeingScale'7 and the Sickness Impact Profile.'8Detailed reviews are available elsewhere.'920These questionnaires were developed toexpress, in numerical terms, disturbances tohealth as seen from the patient's viewpoint.They were the result of painstaking researchdirected towards the development of methodsby which a large proportion of the spectrum ofhuman dys-ease could be quantified in the mosteconomical way (that is, by compressing themaximum amount of information into thesmallest number of terms). There was, ofcourse, no "gold standard" against which thequestionnaires could be judged, so during theirdevelopment extensive tests of validity were

carried out. These concentrated on content,structure, internal reliability, and repeatability.The methods have been described in detail

Figure I List of diseaseprocesses, symptoms, andactivities that may bedisturbed by chronicdiseases of the airways.Note that spiritualactivity is bracketed,because unlike the otherlisted types of activity, itdoes not have reliablemethods of measurement.

--MYMPTOMS ACTIVITIE

ugh PhysicalSputum SocialWheeze EmotionalBreathlessness Intellectual

Economic, (Spiritual) I

I

677



j.com/


eptember 1991. D

ownloaded from


Jones

elsewhere."'2' Although these questionnairescould not be compared with a reference "goldstandard", three important measures ofgeneralhealth, each developed from a differenttheoretical base and with a different structure,have been compared in a comprehensive studyand moderately good correlations were foundbetween them.22

General health questionnaires in airwaysdiseaseThe Quality of Wellbeing Scale'7 and theSickness Impact Profile2"28 have been used inpatients with airways disease. Their scoreshave been shown to correlate with severalrelevant indices of disease activity, disability,and distress" "' and their ability to distin-guish between different levels ofdisease activityhas been clearly shown. For example, inpatients whose breathing never returned tonormal between attacks of breathlessness andwheeze quality of life measured in terms of theSickness Impact Profile score was considerablyworse than in patients who did have symptomfree periods.25 Spirometric measures in generalhave been correlated relatively poorly withgeneral health indices such as the Quality ofWellbeing Scale'7 and the Sickness ImpactProfile.2"27 In contrast, the six minute walkingdistance has been-shown to account for 40% ofthe variance in the Sickness Impact Profilescore." This score has also been shown tocorrelate quite well with disability as assessedby the MRC dyspnoea scale25 26 and the oxygencost diagram.26 Nevertheless, although thesecorrelations were statistically significant, theywere too low for general health to be accuratelypredicted from simple measurements of dis-ability. This overall picture is rather similar tothe pattern of correlations observed betweenspirometric measurements, walking distance,and disability.The Sickness Impact Profile was designed for

application to a very wide range of diseases, andinevitably its content provides restrictedcoverage of any particular clinical condition.This may limit its precision and result in lowsensitivity. For example, in patients withchronic obstructive airways disease, SicknessImpact Profile scores did not differ in patientswith the two mildest grades of disability on theMRC dyspnoea scale, but did show progressivedifferences between the three higher grades.25There also appears to be a non-linear relationbetween disease severity, as measured by pre-bronchodilator FEV, and Sickness ImpactProfile score2"28 (summarised in fig 3). In studiesin which the mean FEV, was less than 50%predicted there was a progressive worsening inthe score by comparison with studies in whichthe mean FEV, was above this level. Thisrelative insensitivity for mild to moderatedisease could mean that the questionnaire ismore responsive to deterioration than toimprovement. This may be critically importantfor attempts to measure the response to treat-ment.

Advantages of standardised healthquestionnairesGeneral health measures such as the SicknessImpact Profile and Quality of Wellbeing Scaleare standardised. Each patient completesexactly the same questionnaire, which is alwaysscored in the same way. Both ofthese question-naires produce a single index or summaryscore, and with the Sickness Impact Profile aprofile ofcategory scores may also be calculatedto provide a more detailed description of thedisturbances of daily life. Such profiles forpatientswith differentlevelsofairflowlimitationhave been published.252729 The advantage ofstandardised health scores is that they allowcomparisons not only between subjects withina given study but across study populations, asshown in figure 3. Furthermore, if a standar-dised questionnaire were sufficiently sensitive,its score would allow direct comparisons ofefficacy between therapeutic trials. Some of thesolutions to the problem of designing a sen-sitive questionnaire have overlooked thisproperty. Examples include some otherwisegood and successful measures. One of these isthe Mahler transition score for breathless-ness,12 which measures one symptom only,though an important and disabling one.Therapeutic benefit from theophylline3' andtargeted muscle training3" has been demon-strated with this measure. As its name suggests,this measure grades changes in symptoms withrespect to each patient's baseline state, so itallows only semiquantitative comparisons be-tween responses from different patients orstudies. The first quality of life measuredeveloped specifically for chronic airflow limita-tion, the Chronic Respiratory Questionnaire,32is also not completely standardised. It has theworthy property of allowing patients partiallyto tailor the questionnaire to suit their state, butunfortunately this "individualisation" does notallow a standardised score to be calculated. Useofthis measure has shown quality oflife benefitsof bronchodilators,33 but direct comparisons ofefficacy between this and other studies in whichthe same questionnaire may be used will not bepossible. This problem has been exemplifiedrecently in a study using the chronic res-piratory questionnaire to compare patientswith chronic obstructive airways disease andcystic fibrosis. In that study the authors had toadjust the scores from the patients with cysticfibrosis to allow comparisons with the patientswith chronic obstructive airways diseasebecause the former identified fewer areas ofdaily life causing dyspnoea.34

Disease specificmeasures ofquality oflifeThere is a need for a standardised and sensitivemeasure of the impact of chronic airwaysdiseases on the daily life and perceived well-being of patients with these conditions. Asdiscussed previously, comprehensive coverageof many different diseases may render themeasure too insensitive for specific diseasestates. Sensitivity may be increased by limiting

678



j.com/


eptember 1991. D

ownloaded from


Quality of life measurementfor patients with diseases of the airways

Figure 3 Mean valuesfor pre-bronchodilatorFEV, and SicknessImpact Profile (SIP)scoresfrom publishedstudies. The number ineach box is the number ofthe study given in thereference list, with theexception of * which iscurrently unpublished(Quirk FH, Jones PW).The patients in references23-25 had chronicobstructive airwaysdisease. In reference 26chronic obstructiveairways disease andasthma were combined.The patients in reference28 had "non-specific"airways obstruction. Theremaining two studies wereon patients with asthma.The total number ofpatients in these studiesexceeds 1800.Note that the psychosocialand physical scores aresubscores of the SicknessImpact Profile, with self-explanatory titles.

SIPtotalscore

301

20 -

10-

0 30

SIPphysicalscore

30-1

20 -

10

n_A1u- 1 I-.0 30

SIPpsychosocialscore30,

20-

10-

0

0

40 50 60 70 80 90FEV1 (% predicted)

I I I I I

40 50 60 70 80 9(FEV1 (% predicted)

II30 40 50 60 7b 80 90

FEV1 (% predicted)

the questionnaire's content to items selected toreflect abroad range ofeffects ofa specific diseaseor a limited group of related diseases. Thisrestriction of the questionnaire's content hasoccasionally led to the criticism that a diseasespecific measure is something less than a "true"quality of life measure as it addresses a limitedarea of disturbance to health. Such criticismmay be based on a narrow understanding of thevaried and complex nature of any single dis-ease. General health measures are a bestattempt to cover the whole spectrum of dys-ease, but this inevitably reduces the number ofitems referable to a specific clinical condition.For example, the Sickness Impact Profileincludes items concerned with intravenousfeeding, but has a few items that are relevant to

patients with airways disease. In a study of 152patients with moderately severe chronic ob-structive airways disease 20% of the items inthis questionnaire were left completely un-answered by every patient, and a quarter of thequestionnaire items accounted for two thirds ofthe total number of positive responses (P WJones, CM Baveystock, unpublished observa-tions). The overall effect of this is to producevery low scores even in patients who havemoderate levels of disability.25 A correctlydesigned and appropriately applied diseasespecific questionnaire might therefore provide abetter and more precise measurement ofquality of life than a general index.

The St George's respiratoryquestionnaireWe have developed a measure of impairedhealth in patients with diseases causing airwayobstruction, both asthma and chronic obstruc-tive airways disease, known as the St George'srespiratory questionnaire. This questionnaireis divided into three components: symptoms,activity, and impacts. A total score is alsocalculated. The "impacts" section covers socialand emotional disturbances due to the disease.Although some items in this section relate tothe psychological impact of the disease, ques-tions designed specifically to measure anxietyand depression were excluded as several suit-able questionnaires are available for this pur-pose. Each item in the St George's respiratoryquestionnaire is accorded a weight for theamount of distress associated with the symp-tom or state described. These weights wereobtained during the questionnaire's develop-ment from studies in 140 asthmatic patients insix countries. Factors such as age, sex, and theduration, severity, and variability of diseaseeach contributed to less than 2% ofthe variancein weights between patients.35 There was nosignificant difference in the weights obtained inEngland, Finland, Italy, Thailand, or theUnited States'3; but the weights from theNetherlands were on average 19% higher thanthose obtained in the other countries. Noobvious explanation was found for this excep-tion. The weights in asthmatic patients werealmost identical to weights collected in a groupof older patients with chronic obstructive air-ways disease who had more severe disease.37 Asa result ofthese studies, we have concluded thatthe weights are suitable for a wide range ofpatients with airways disease.The final version of the questionnaire has

good repeatability for this type of measure.37To assess its reliability-that is, its ability todistinguish between different levels ofhealth-the questionnaire was compared with othermeasures of disease severity, disability, anddistress, including spirometric measurements,bronchodilator response, results of oximetryduring exercise, six minute walking distance,the MRC respiratory questionnaire, the hos-pital anxiety and depression scale,38 and theSickness Impact Profile, in a study on largenumbers of patients with a wide spectrum ofairflow obstruction.37 The component parts of

679

I



j.com/


eptember 1991. D

ownloaded from


680

the questionnaire correlated with appropriatereference measures and the total score wasshown to sum a range of different disturbancesto health. The pattern of correlations betweenthe St George's respiratory questionnairescores and the reference variables closely fol-lowed the pattern obtained with the SicknessImpact Profile, but the St George's question-naire was over twice as sensitive to differences indisease severity as the Sickness Impact Profile.39Measurements performed one year apartshowed that changes in the St George's res-piratory questionnaire scores correlated withchanges in the reference measures listed aboveand showed that the total score aggregatedchanges in several different areas of diseaseactivity.37

Mood state in patients with diseases ofthe airwaysA high incidence of clinically importantpsychological morbidity, as measured by thegeneral health questionnaire, has been reportedin patients with chronic obstructive airwaysdisease.26" ' In one study of patients with amean post-bronchodilator FEV, of 53%predicted 47% of the patients had an anxietyscore at or above the borderline of clinicalsignificance on the hospital anxiety and depres-sion scale and 29% had depression scores inthis range.25 A further study recorded moderatelevels of anxiety on the basis of the Speilbergerindex, and borderline levels of depressionaccording to the Beck depression inventory.42In younger asthmatic patients anxiety levelshave been reported to be higher than in normalsubjects.4" A correlation between psychologicalmorbidity and poor spirometric results hasbeen found in patients with chronic obstructiveairways disease"2 and psychological state mayinfluence patients' reporting of respiratorysymptoms.26 The latter has important implica-tions for quality of life measurement. Anxietyand depression scores were found to correlatewith the symptoms score from the St George'srespiratory questionnaire. To analyse this fur-ther, responses to the MRC respiratory ques-tionnaire items about the frequency of coughand wheeze were included with the anxiety ordepression score in a multivariate statisticalmodel.'7 In this model, in which the StGeorge's symptoms score was the dependentvariable, respiratory symptoms were thedominant correlates and correlations withmood scores were much weaker. A similarpicture was seen with the St George's activityscore. Correlations with anxiety and depressionwere found with this score, but these correla-tions became weaker when the MRC dyspnoeascore was included in a multivariate model andthe dyspnoea score then became the dominantcorrelate. In contrast, correlations betweenmood score and the impacts score of the StGeorge's questionnaire remained strong, evenin multivariate models that included otherpowerful correlates of the impact score, such asresponses to MRC questionnaire items con-cerned with dyspnoea and frequency ofwheeze. Mood state scores appear therefore to

correlate with the symptoms and activity sec-tions of the St George's questionnaire, largelythrough a shared association with measures ofdisease activity, including symptom severity,physical impairment, and perceived disability.The impacts section contains items thatcorrelate with anxiety independently of otherdisease related factors. Clearly relations be-tween anxiety, depression, and the perceptionof health are complex and still poorly delin-eated.

Quality of life questionnaires in clinicaltrials"Quality of life" measures are now beginningto appear in clinical trials, but are they anythingmore than a marketing ploy? I believe that theyare. Most people would agree, that with a life-long disease, a major requirement of anytherapy should be a clear demonstration of itsbeneficial effects on daily life and wellbeing. Asdiscussed earlier, it is not possible to makereliable predictions of a patient's disability orimpaired wellbeing from measurements of air-ways function. There is as yet, limited dataconcerning the relationship between changes inspirometry following therapy and changes inhealth, but the available evidence suggests agenerally poor correlation. In patients withasthma studied over an eighteen monthinterval, changes in FEV, correlated withchanges in quality of life measured with thequality of wellbeing scale (r2 = o040).'7 In astudy of bronchodilators in patients withchronic obstructive airways disease there was asignificant correlation (r2 = 0 30) betweenFEV, and quality of life as determined by thechronic respiratory questionnaire." In bothof these studies the correlations, thoughmoderately good, were too low to allowaccurate prediction of the quality of life scorefrom the FEVI. In another study in patientswith chronic obstructive airways disease over aone year interval changes in quality of life asassessed by the St George's respiratory ques-tionnaire score correlated poorly with changein FEV, but rather better with the patient'swalking distance or theMRC dyspnoea score.'7A study on the effect of salbutamol in chronicobstructive airways disease found a very lowand non-significant correlation betweenchanges in spirometric values and perceivedbreathlessness."3 Finally, in a study using theo-phylline, breathlessness improved but with nosignificant change in spirometric results,arterial blood gas tensions, or walking dis-tance.'0 Clearly we cannot assume thatimprovement in subjective health or disabilitywill accompany a measured improvement inairways function.

Health questionnaires were developed as ascientific response to the problem of quantify-ing aspects of disease severity that could not beassessed by existing measures. Properlydeveloped health related quality of life ques-tionnaires are validated measures of the impactof a disease on the patient's daily life andperceived wellbeing. In tests of drug efficacy itis clearly appropriate to measure the drug's

Jones



j.com/


eptember 1991. D

ownloaded from


Quality of life measurementfor patients with diseases of the airways

effect on the patient's health and not justspirometric values. The use of standardisedmeasurements will allow direct comparisons ofefficacy, measured in terms of overall effects ondaily life and wellbeing, between different trials,drugs, and therapeutic modalities. Only theaccumulation of a large body of studies willallow a judgment about whether spirometrycan provide an adequate surrogate measure ofimprovement in health (as opposed to airwaysfunction).

Quality of life measures in audit androutine practiceQuality of life measures are clearly suitable forauditing the outcome of medical care in termsof patients' perceived health, but they shouldnot be confused with measures of "satisfac-tion".45 Measurement of satisfaction is aprocess used to assess the success with whichthe supplier or a process meets or exceeds therequirements of a consumer or client. It has tobe scored largely in relative terms. A discussionof the intriguing and complex relations be-tween satisfaction and perceived health andwellbeing is beyond the remit of this review. Itshould be realised, however, that in the contextof medical audit there may be significant dif-ferences between the conclusions obtainedfrom an audit using a quality of life measureand one using a satisfaction index. Considerthese two statements: "In this audit 80% ofpatients were satisfied with their improvementwith treatment" and "In this audit the meanbaseline health score of the patients was 40units; 60% of patients achieved or surpassedthe minimum criterion for improvement." Thefirst statement says nothing about the patients'health, gives no indication of which aspect oftheir treatment satisfied them, and provides nomeasure of the size of any improvement. Thesecond gives a clear statement about thebaseline state of health of the patients and theproportion who improved by a predeterminedamount, though it does not state what pro-portion felt satisfied with this level ofimprovement.The role of quality oflife measures in routine

clinical practice has yet to be established. TheSt George's respiratory questionnaire, forexample, takes about 10 minutes to complete-a little less than the time taken to testbronchodilator responsiveness. We are inves-tigating its application to routine practice andare examining the possibility of developingshortened versions. One problem may arisebecause quality of life measures may not bequite as repeatable as spirometry. This maylimit their routine use for testing therapeuticresponses in individuals as opposed to groupsof patients. Despite this qualification I believethat, either in its current form or in shortenedforms, quality of life questionnaires willprovide valuable information for routine clin-ical management and that they will com-plement but not replace physiologicalmeasurement.

Limitations of quality of life measures inairways diseaseQuestionnaires for quality of life measurementshould be applied only to diseases in whichtheir use has first been validated. Even whenappropriately selected the presence of inter-current disease could be a confounding factor.Unreliable results may be produced if qualityof life measures are used incorrectly. In gen-eral, these questionnaires are easy to use andfairly robust, but they should be handled withthe same care and attention to detail as anyother instrument. One perceived limitationmay be the rate of change with treatment. Bycomparison with measurement of disturbancesclose to the underlying pathological process (asillustrated in figs 1 and 2) quality of life scoresmay respond to therapeutic intervention quiteslowly. This is not due to an inherent weaknessin the questionnaires but is a reflection of theinfluence of factors that can modulate theimpact of disease on the patient's life. Many ofthese factors may be unrelated to the under-lying disease process and therefore would notbe directly influenced by treatment. Disuseatrophy of leg muscles due to exercise limita-tion resulting from breathlessness is an obviousexample. Instant cure would not immediatelyreturn the patient's life to its premorbid state.Finally, expectations of quality of lifemeasurement should be realistic. If a treatmenthas little measurable influence on the underly-ing disease processes or on processes close to it,it is unlikely to improve subjectively perceivedwellbeing unless it also possesses directpyschoactive properties, as may be the casewith oral corticosteroids.42 Good quality of lifemeasures are well validated instruments thatprovide hard data on health. They are notmagic wands to be used when othermeasurements have shown little benefit fromthe treatment under study.

1 Stark RD, Gambles SA, Lewis JA. Methods to assessbreathlessness in healthy subjects: a critical evaluation andapplication to analyse the acute effects of diazepam andpromethazine on breathlessness induced by exercise or byexposure to raised levels of carbon dioxide. Clin Sci1981;61:429-39.

2 Adams L, Chronos N, Lane R, Guz A. The measurement ofbreathlessness induced in normal individuals: individualdifferences. Clin Sci 1986;70:131-40.

3 Wilson RC, Jones PW. A comparison of the visual analoguescale and modified borg scale for the measurement ofdyspnoea during exercise. Clin Sci 1989;76:277-82.

4 Wilson RC, Jones PW. Differentiation between the intensityof breathlessness and the distress it evokes in normalsubjects during exercise. Clin Sci 1991;80:65-70.

5 McGavin CR, Gupta SP, McHardy GJR. Twelve minutewalking test for assessing disability in chronic bronchitis.BMJ 1976;i:822-3.

6 Swinbum CR, Wakefield JM, Jones PW. Performance,ventilation and oxygen consumption in three differenttypes of exercise test in patients with chronic obstructivelung disease. Thorax 1985;40:581-6.

7 Carter R, Peavler M, Zinkgraf S, Williams MS, Fields S.Predicting maximal exercise ventilation in patients withchronic obstructive pulmonary disease. Chest 1987;92:252-9.

8 Matthews JL, Bush BA, Ewald FW. Exercise responsesduring incremental and high intensity and low intensitysteady state exercise in patients with obstructive lungdisease and normal control subjects. Chest 1989;%:1 1-7.

9 Medical Research Council Committee on the Aetiology ofChronic Bronchitis. Standardised questionnaire on res-piratory symptoms. BMJ 1960;ii:1665.

10 McGavin CR, Artvinli M, Naoe H, McHardy GJR. Dysp-noea, disability and distance walked: comparison of

681



j.com/


eptember 1991. D

ownloaded from


Jones

estimates of exercise performance in respiratory disease.BMJ 1978;ii:241-3.

11 O'Reilly JF, Shaylor JM, Fromings KM, Harrison BDW.The use ofthe 12 minute walking test in assessing the effectof oral steroid therapy in patients with chronic airwaysobstruction. Br J Dis Chest 1982;76:374-82.

12 Mahler DA, Weinberg DH, Wells CK, Feinstein AR.Measurement of dyspnoea. Contents, interobservercorrelates of two new clinical indexes. Chest 1984;85:751-8.

13 Wolkove N, Dajozman E, Colacone A, Kreisman H. Therelationship between pulmonary function and dyspnea inobstructive lung disease. Chest 1989;96:1247-51.

14 Mahler DA, Wells CK. Evaluation of clinical methods forrating dyspnea. Chest 1988;93:580-6.

15 Guyatt GH, Thompson PJ, Berman LB, Sullivan MJ,Townsend M, Jones NL, et al. How should we measurefunction in patients with chronic lung disease? J Chron Dis1985;38:517-24.

16 Wilson RC, Quirk FH, Jones PW. Correlation in asthmaticpatients of magnitude estimates of dyspnoea during cycleexercise with walking distance and subjective disabilitybut not with disturbed lung function [abstract]. Thorax1991;46:277P.

17 Kaplan RM, Atkins CJ, Timms R. Validity of a well-beingscale as an outcome measure in chronic obstructivepulmonary disease. J Chron Dis 1984;37:85-95.

18 Bergner M, Bobbitt RA, Carter WB, Gilson BS. Thesickness impact profile: development and final revision ofahealth status measure. Med Care 1981;19:878-85.

19 McDowell I, Newell C. Measuring health. Oxford: OxfordUniversity Press, 1987.

20 Walker SR, Rosser RM. Quality of life: assessment andapplication. Boston: MTP, 1987.

21 Streiner DL, Norman GR. Health measurement scales: apracticalguide to their development and use. Oxford: OxfordUniversity Press, 1989.

22 Read JL, Quinn RJ, Hoefer MA. Measuring overall health:an evaluation of three important approaches. J Chron Dis1987;40(suppl):7-21 S.

23 Nocturnal Oxygen Therapy Trial (NOTT) Group. Contin-uous or nocturnal oxygen therapy in hypoxemic chronicobstructive lung disease. Ann Intern Med 1980;133:14-20.

24 Intermittent Positive Pressure Trial Group. Intermittentpositive pressure breathing therapy ofchronic obstructivepulmonary disease. Ann Intern Med 1983;99:612-20.

25 Jones PW, Baveystock CM, Littlejohns P. Relationshipsbetween general health measured with the SicknessImpact Profile and respiratory symptoms, physiologicalmeasures and mood in patients with chronic airflowlimitation. Am Rev Respir Dis 1989;140:1538-43.

26 Williams SJ, Bury MR. Impairment, disability and handicapin chronic respiratory illness. Soc Sci Med 1989;29:609-16.

27 Schrier AC, Dekker FW, Kaptein AA, Dijkman JH. Qualityof life in elderly patients with chronic non-specific lungdisease seen in family practice. Chest 1990;90:894-9.

28 Jones PW, Baveystock CM. Impaired general health andmood in asthmatic patients in six countries [abstract]. EurRespir J 1990;3(suppl 10):239S.

29 McSweeny J, Grant I, Heaton RK, Adams KM, TimmsRM. Life quality of patients with chronic obstructivepulmonary disease. Arch Intern Med 1982;142:473-8.

30 Mahler DA, Matthay RA, Snyder PE, Wells CK, Loke J.Sustained-release theophylline reduces dyspnea in non-reversible obstructive airways disease. Am Rev Respir Dis1985;131:22-5.

31 Harver A, Mahler DA, Daubenspeck JA. Targeted in-spiratory muscle training improves respiratory musclefunction and reduces dyspnea in patients with chronicobstructive pulmonary disease. Ann Intern Med 1989;111:117-24.

32 Guyatt GH, Berman LB, Townsend M, Pugsley SO,Chambers LW. A measure of quality of life for clinicaltrials in chronic lung disease. Thorax 1987;42:773-8.

33 Guyatt GH, Townsend M, Pugsley SO, Keller JA, Short D,Taylor DW, Newhouse MT. Bronchodilators in chronicair-flow limitation. Am Rev Respir Dis 1987;135:1069-74.

34 Weir DC, Freeman W, Roberts KT, Stableforth DE, BurgePS. Comparison of a measure of quality of life in patientswith cystic fibrosis and chronic airflow obstruction[abstract]. Thorax 1991;46:288P.

35 Quirk FH, Jones PW. Patients' perception of distress due tosymptoms and effects of asthma on daily living and aninvestigation of possible influential factors. Clin Sci1990;79- 17-21.

36 Quirk FH, BaveystockCM, Wilson RC, Jones PW. Influenceofdemographic and disease related factors on the degree ofdistress associated with symptoms and restrictions ondaily living due to asthma in six countries. Eur Respir J1991;4:167-71.

37 Jones P, Quirk F, Baveystock C, Littlejohns P. A self-complete measure for chronic airflow limitation-the StGeorge's Respiratory Questionnaire. (Submitted.)

38 Zigmund AS, Snaith RP. The hospital anxiety and depres-sion scale. Acta Psychiatr Scand 1983;67:361-70.

39 Jones PW, Quirk FH, Baveystock CM. The St George'sRespiratory Questionnaire. Respir Med (in press).

40 Morgan AD, Peck DF, Buchanan DR, McHardy GJR.Psychological factors contributing to disproportionatedisability in chronic bronchitis. J Psychosom Res1983;27:259-63.

41 Rosser R, Denford J, Heslop A, Kinston W, Macklin D,Minty K, et al. Breathlessness and psychological mor-bidity in chronic bronchitis and emphysema: a study ofpsychotherapeutic management. Psychol Med 1983;13:93-110.

42 Swinburn CR, Wakefield JM, Newman SP, Jones PW.Evidence of prednisolone induced mood change ("steroideuphoria") in patients with chronic obstructive airwaysdisease. Br J Clin Pharmacol 1988;26:709-13.

43 Yellowlees PM, Ruffin RE. Psychological defenses andcoping styles in patients following a life-threatening attackof asthma. Chest 1989;95:1298-303.

44 Guyatt GH, Townsend M, Keller J, Singer J, Nogradi S.Measuring functional status in chronic lung disease:conclusions from a randomized control trial. Respir Med1989;83:293-7.

45 Fitzpatrick R. Surveys of patient satisfaction: 1-Importantgeneral considerations. BMJ 1991;302:887-9.

682



j.com/


eptember 1991. D

ownloaded from


quality measurement of the - thoraxmeasurement ofhealth related quality oflife...

Documents