Thorax 1994;49:1109-1115 1109 Influence of treatment on peak expiratory flow and its relation to airway hyperresponsiveness Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from and symptoms
Huib A M Kerstjens, Paul L P Brand, Petra M de Jong, Gerard H Koeter, Dirkje S Postma, and the Dutch CNSLD Study Group
Abstract dition to the other markers but not instead Background - Despite effective treat- of them. ments, the morbidity and mortality of ob- structive airways disease (asthma and (Thorax 1994;49:1109-1115) COPD) remains high. Home monitoring of peak expiratory flow (PEF) is in- creasingly being advocated as an aid to Peak expiratory flow (PEF) measurements are better management ofobstructive airways increasingly recommended to aid in the dia- disease. The few available studies de- gnosis and management of asthma."1 The scribing effects of treatment on the level National Asthma Education Program, for in- and variation of PEF have involved rel- stance, has recently stated that PEF meas- atively small numbers of subjects and did urements are a valuable clinical tool for not use control groups. assessing the degree of airflow obstruction and Methods - Patients aged 18-60 years were monitoring the response to therapy.7 To date, selected with PC20 < 8 mglml and FEV, however, the effect of different treatments on <95% confidence interval ofpredicted nor- level of PEF - and especially on PEF variation mal. They were randomised to receive, - has not been extensively documented. An in addition to a P2 agonist, either an improvement in PEF with inhaled cortico- inhaled corticosteroid (BA + CS), an anti- steroids, for instance, has been shown only cholinergic (BA + AC), or a placebo in two relatively small controlled studies.89 In (BA + PL). One hundred and forty one of a recent long term study ofmild, newly detected
these subjects with moderately severe ob- adult asthmatic patients, Haahtela and col- http://thorax.bmj.com/ structive airways disease completed seven leagues compared the results of treatment with periods of two weeks of morning and budesonide and terbutaline and showed im- afternoon PEF measurements at home provements in both morning and evening PEF during 18 months of blind follow up. with budesonide. The effects of treatment on Results - Improvements in PEF occurred diurnal variation in PEF were not presented.' within the first three months of treatment To our knowledge, a decrease in diurnal vari- with BA + CS and was subsequently main- ation in PEF with inhaled corticosteroids has
tained: the mean (SE) increase in morning been documented only in children.'01' on September 28, 2021 by guest. Protected copyright. PEF was 51 (8) /min in the BA + CS group The variation in PEF is thought by many compared with no change in the other two to reflect the degree of airway hyper- groups. Similarly, afternoon PEF in- responsiveness.7 It is important to note, how- creased by 22 (7) 1min. Diurnal variation ever, that most of the epidemiological'2-" and in PEF (amplitude %/omean) decreased clinical'5"'" studies are cross sectional, whereas Departnent of from 18*0% to 10-2% in the first three longitudinal, within-subject relations between Puhmonology, University Hospital months of treatment with BA + CS. PC20 and PEF variation (and symptom scores) Groningen, Within-subject relations between changes are more relevant. This is especially pertinent Oostersingel 59, 9713 in diurnal variation in PEF and changes for self-management plans which rely on home EZ Groningen H A M Kerstjens in PC20 were found to be predominandy measurements of PEF. Only one report has P L P Brand negative (median p - 0 40) but with a large documented this within-subject relation be- G H Koeter scatter. Relations between diurnal vari- tween diurnal PEF variation and PC20: a weak D S Postma ation in PEF and changes in symptom relationship was found in a small cohort study Department of scores, FEV1, and bronchodilator response ofasthmatic patients not receiving standardised Pulmonology, were even weaker. therapy. 8 University Hospital - Leiden, Leiden, Conclusions In patients with moderately We have recently completed a double blind, The Netherlands severe obstructive airways disease, PEF controlled, multicentre clinical trial in patients P M de Jong rates and variation are greatly improved with obstructive airways disease and a broad Reprint requests to: by inhaled corticosteroids. Since the re- range ofpatient characteristics. "9 In the current Dr H A M Kerstjens. lation of diurnal PEF variation with PC20, report the influence of treatment with inhaled Received 5 April 1993 symptoms, FEV1, and bronchodilator re- corticosteroids and/or bronchodilators on Retumed to authors 11 October 1993 sponse were all weak, these markers of morning and afternoon PEF levels, as well as Revised version received disease severity may all provide different on diurnal variation in PEF, is tested. Sub- 26 May 1994 Accepted for publication information on the actual disease state. sequently, the interrelationships between di- 2 September 1994 PEF measurements should be used in ad- urnal PEF variation, PC20, and symptoms are 1110 Kerstiens, Brand, de _'ong, Koeter, Postma described, both cross sectionally between sub- the outpatient clinic, patients used a Wright jects, and longitudinally within subjects. mini peak flow meter (Clement Clarke Inter-
national Ltd, London, UK) to record PEF Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from values at home. The best of three blows was Methods recorded in the morning (directly after rising, Patients aged 18-60 years with respiratory before bronchodilator therapy) and in the after- symptoms and no other major illnesses20 were noon (before the evening meal, before broncho- selected according to the following criteria: (1) dilator therapy). PEF values were recorded FEV, ranging between 4-5 and 1 64 residual on a separate sheet in the diary. Symptom standard deviations (RSD) below the predicted scores were noted on a four-point scale (0= value - that is, between 2-30 and 0-84 litres no symptoms, 3=severe symptoms) daily for below predicted normal FEV, for men or wheeze, dyspnoea, cough, and phlegm, sep- between 1-71 and 0-62 litres below predicted arately. for women - or FEV1/IVC (inspiratory vital Using data from a standardised history, capacity) more than 1-64 RSD (men: 11-76%, different syndromes were identified which ad- women: 10-68%) below the predicted value, hered to the criteria of the American Thoracic provided that total lung capacity was normal Society.22 Patients reporting attacks of breath- (higher than 1 64 RSD (men: 1.15, women: lessness and wheeze (asthmatic attacks) with- 0-98 litres) below the predicted level2"). FEV, out chronic (that is, for more than three months had to be larger than 1-2 litres; and (2) con- per year) cough and sputum production were centration of histamine causing a 20% fall in identified as having asthma. Current or former FEV, (PC20) s,8mg/ml. smokers without a history of asthmatic attacks, Inhaled corticosteroids were tapered off and reporting either chronic cough with or without discontinued completely four weeks before a sputum production or dyspnoea when walking prerandomisation visit. Other maintenance quietly on level ground, or both, were included medication was withheld for at least six weeks in the chronic obstructive pulmonary disease (ketotifen, antihistamines), four weeks (cro- group (COPD). Patients with both asthmatic molyn sodium), or 48 hours (theophyllines) attacks or recurrent wheeze and chronic cough before the start ofthe study. Maintenance med- and sputum production were labelled asthmatic ication with oral corticosteroids was not al- bronchitics. Subjects with insufficient data to lowed. Atopy was defined on the basis of skin establish a diagnosis from history taking were tests.20 included in an undefined diagnosis group.20 The study was designed as a randomised double blind clinical trial with three parallel treatment arms. Patients were allocated to one DATA ANALYSIS of three double blind regimens from identical PEF records and symptom scores on the diary
metered dose inhalers: all patients received cards were accepted only if at least nine of 14 http://thorax.bmj.com/ an inhaled P2 agonist (terbutaline 250 jig, two days were filled in completely. Of the 274 puffs four times daily) combined with either patients who entered the study, 222 provided an inhaled corticosteroid (beclomethasone acceptable PEF records at the baseline visit and 100 jg, two puffs four times daily (BA+ CS)), 235 acceptable diary cards. As the objectives of or an inhaled anticholinergic (ipratropium the current analyses were focused on time and bromide 20 gg, two puffs four times daily treatment effects on PEF, the analyses are lim- (BA+AC)), or an inhaled placebo, two puffs ited to those patients with acceptable baseline
four times daily (BA+PL). Additional bron- PEF and diary cards (n = 204) and to patients on September 28, 2021 by guest. Protected copyright. chodilator medication was supplied as sal- who also had at least 18 months of follow up butamol dry powder inhalations (400 jg) on (n = 141). These patients had a slightly higher demand. No other concomitant pulmonary PEF value in 1/min (as measured from flow- medication was allowed, except during ex- volume curves) at the baseline visit than the acerbations when a 12 day course of oral patients excluded from the current analyses prednisolone was administered.20 (384 v 351 1/min, p<0 05), but no significant Follow up visits were scheduled every three difference in PEF as % predicted2' (69-6% and months during which FEV, and, at alternate 74-2% respectively). There were no significant visits, bronchodilator response and PC20 were differences between the entered and excluded measured. FEV, and PC20 were only measured patients with respect to age, smoking habits, during clinically stable periods and not within FEVI, log2PC20, and bronchodilator response. four weeks of completion of a prednisolone Mean moming and aftemoon PEF values course. Pulmonary medication was dis- were calculated for each patient for every period continued at least eight hours before each test. of 14 days before attending the outpatient clin- For bronchodilator response testing FEV, was ics. PEF variation was assessed as mean diumal measured before and 20 minutes after four variation = absolute value of ((afternoon separate inhalations of 250 gg terbutaline reading - moming reading)/mean of these sulphate from a metered dose inhaler ad- two) x 100%. A mean of 14 days was then ministered through a 750 ml spacer device (Ne- calculated. At every visit to the outpatient clinic buhaler). Histamine provocation tests were the PEF meters were checked visually for mal- performed using a two minute tidal breathing function due, for example, to sputum retention. test.20 In order to check long term accuracy the mom- Patients were asked to keep PEF records and ing PEF values at home were compared with a diary card for 14 consecutive days before PEF values from a pneumotachograph at the each visit. After standardised instruction in baseline visit and after 2-5 years. These were Influence of treatment on peak expiratory flow and its relation to airway hyperresponsiveness and symptoms lilll
Table 1 Baseline characteristics by treatment group 525 nA Treatment group
C Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from BA+PL BA+AC BA+CS 475 No. of patients 41 38 62 :LL. Sex: no. (%) male 27 (66%) 27 (71%) 45 (73%) w 0- 425 Mean (SD) age 8.4 (10-8) 39-9 (13 2) 42.0 (12 4) CD FEV, prebronchodilatation: C Mean (SD) litres 2 41 (0-74) 2-32 (0 80) 2 43 (0-79) 0F Mean (SD) % predicted 66-4 (17 7) 62 2 (15 6) 64 9 (15 1) 375 -T------T------T-l----I------T...... Mean (SD) number RSD from ------18 predicted FEV, -2 66 (1-5) -2 92 (1 2) -2-72 (1-2) L -l-- I I--L X 4 Mean (SD) BDR 13-9 (9-1) 12 3 (7-9) 11-2 (7-7) 325 I1 1- 1 I 0 3 6 9 12 15 18 Mean (SD) log2PC20 histamine (mg/ml) -2 04 (2-37) -1-81 (2-27) -1-58 (2 16) Geometric mean PC20 (mg/ml) 0 24 0-29 0 33 Mean (SD) morning PEF (1/min) 373 (93) 364 (103) 393 (130) 525 r Mean (SD) afternoon PEF (1/min) 431 (90) 415 (100) 456 (122) C PEF variation (amplitude %mean) 16-9 (11) 16-4 (11) 18-0 (13) U- 475 Smoking: no. (%) never 14 (34%) 12 (32%) 18 (29%) no. (%) ex 13 (32%) 11 (29%) 19 (31%) w T I no. (%) current 14 (34%) 15 (39%) 25 (40%) 425 -.------1I I .1.I... C --- I---- I----- Atopy: no. (%) atopic 25 (61%) 21 (55%) 35 (56%) 0 Symptom-based diagnosis groups: no. (%) C Asthma 17 (41%) 17 (45%) 20 (32%) 0 375 Asthmatic bronchitis 13 (32%) 13 (34%) 23 (37%) COPD 11 (27%) 8 (21%) 16 (26%) Undefined diagnosis 0 0 3 (5%) 325 _ 0 3 6 9 12 15 18 BA + PL= P2 agonist plus placebo; BA +AC = P2 agonist plus anticholinergic; BA + CS = P2 agonist plus corticosteroid; BDR=bronchodilator response to 1000 sg terbutaline (AFEV,%predicted); PEF = peak expiratory flow (for calculation of diumal variation, see methods); RSD = residual 22 standard deviation (from predicted FEV,, see ref 21). -c C-.-
m 0 19 not significantly different neither at the start . _E nor at the end of the study. Symptom scores 10 over 14 days were averaged for each of the 16 symptoms in the diary and added to obtain a U.0 mean symptom score. 0- -- Calculations with PC20 were performed using 13 the base 2 logarithm as this reflects doubling doses and normalised the distribution. For the 0
purposes of analysis, patients already re- http://thorax.bmj.com/ sponding to saline or to the lowest con- Months centration of histamine (0 03 mg/ml) were Figure 1(A) Mean (SE) morning peakflow (71min). assigned a PC20 vaiue of 0-015, being half the (B) Mean (SE) afternoon peak flow (7/min). (C) Mean lowest concentration applied.20 The data are (SE) variation in diurnal peak flow (amplitude % presented as mean (SE) values unless otherwise mean). BA + CS-2= agonist + corticosteroid (n =62); ---- BA +AC=f12agonist + anticholinergic stated. (n = 38); .... BA +PL = 12 agonist + placebo Skewedness of distributions was assessed (n = 41).
with Kolmogorov-Smirnov tests. If a p value on September 28, 2021 by guest. Protected copyright. <0 05 was obtained, non-parametric tech- niques were applied (Spearman's p for cor- relation, Wilcoxon signed rank sum test, and EFFECT OF TREATMENT ON PEF LEVELS AND Kruskal-Wallis non-parametric analysis ofvari- DIURNAL VARIATION IN PEF ance to compare group means); otherwise para- Changes with treatment in morning PEF are metric analyses were used (Pearson's r for shown in fig 1A. After three months the mean correlation, Student's t test, or one way analysis increase in morning PEF was 51 (8) I/min in ofvariance to compare group means). Although the BA + CS group compared with -4 (7) in patient selection was based on the number of the BA + PL group. The mean difference of 55 residual standard deviations below predicted 1/min (p BETWEEN-SUBJECT RELATION OF DIURNAL WITHIN-SUBJECT RELATION OF DIURNAL VARIATION IN PEF TO OTHER MARKERS OF VARIATION IN PEF WITH OTHER MARKERS OF DISEASE SEVERITY DISEASE SEVERITY on September 28, 2021 by guest. Protected copyright. The cross sectional correlation between diurnal To determine the strength of the association variation in PEF and log2PC20 at the baseline between diurnal variation in PEF and PC20 visit (p= -0 39, p<0 001) became somewhat within patients, individual rank correlation co- stronger at later visits (p-0.50, -0-60, and efficients were calculated (table 3). Although a 25 r predominantly negative association was found (median p-0 40), the scatter was rather large (90% range -1-00 to 0-80, fig 4A). Within 20K patients there was also a significant, though c weak, relation between changes in diurnal vari- 0 0 L- ation in PEF and changes in symptom scores C F- 0 (fig 4B). The relations between diurnal vari- ._ 15 E 0 Cu co Eo ation in PEF and the other markers of disease U-c .W 0.50 0 the different 0- 0. severity were similar between 10K 0 E treatment and diagnostic groups (table 3). '5 C-) >0 5 Discussion To our knowledge this is the first report of the long term effects of double blind treatment 0.13 0 with inhaled corticosteroids and agonists 0 12 15 18 P2 3 6 9 on PEF rates and variation in adults. Mean Months morning and afternoon PEF values improved Figure 3 Temporal relation of changes in diurnal variation in PEF, PC20, and synnptom with the combination of inhaled corticosteroid scores in the group treated with 2 agonist + corticosteroid (BA + CS, n = 62). * = diurnal PEF variation (+ SE); * =PC20 (histamine) mg/ml (+ SE); cross-hatcheId bars and 2 agonist (BA + CS) therapy in patients represent median symptom scores; error bars represent interquartile ranges. with moderately severe obstructive airways dis- Influence of treatment on peak expiratory flow and its relation to airway hyperresponsiveness and symptoms 1113 occurred within the first six months, the im- 25 - A provement continued until at least one year (fig 3). This is an important finding as airway Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from 20 _ hyperresponsiveness continued to improve 01) even when changes in PEF and FEV, could c)0. no longer be demonstrated. The longitudinal a) 15 _ relation between variation in PEF and PC,0 Q .0 cannot therefore be as close as is suggested from 10 - some cross sectional analyses,'625 although not E from others.'4'826 More important than these z 5 between-subject analyses, however, is the weak- ness of the within-subject correlations between changes in PEF variation and other markers of o I I I1 -1 -0.8 -0.6 -0-4 u-.2 0-2 0-4 0-6 0.8 1 disease severity such as symptoms, FEVI, and Within-subject correlation coefficient PC20. Our results in a large group of patients between diurnal PEF variation and PC20 on standardised treatment confirm and extend the observations ofJosephs et al in an open and uncontrolled follow up study of eight children 20r B and 12 adults with mild asthma.'8 Although in our study the relation between diurnal variation CO in PEF and PC20 was stronger (median p = 15 F- (U0) -0 40) than between PEF variation and either 0. FEV,, bronchodilator response, or symptom Q scores (median p-0-23, 0 20, and 0-25 re- 0 10 0 spectively), considerable scatter existed for all .0 four measurements (table 3, fig 4A,B). This E implies that PEF variation and the other three z 5 markers of disease severity (PC2O, FEVI, symp- tom scores) are not interchangeable and per- haps provide different information on the actual -1 -0.8 -0.6 4-04 -0.2 0 0.2 0-4 0-6 0-8 1 state and are informative in their own right. Within-subject correlation coefficient between The within-subject relation between diurnal diurnal PEF variation and symptom score variation in PEF and PC20 was not significantly different between patients with a symptom Figure 4 Frequency distribution of Spearman p for based diagnosis ofasthma, asthmaticbronchitis, within-subject correlations of diurnal peak flow variation or chronic to (A) log2PC20 and (B) symptom scores. Arows indicate obstructive pulmonary disease median p. (COPD) (table 3). Although a high variation in http://thorax.bmj.com/ PEF is considered characteristic of asthma,45 ease. Simultaneously, diurnal variation in PEF considerable variation in PEF has been found decreased. These improvements occurred to be present in many current or former within the first three months of treatment and smokers without asthmatic attacks.'526 Our were subsequently maintained. Because diurnal study group was selected on the basis of ob- variation in PEF is reduced with inhaled corti- jective criteria of age, FEV, and PC20. In- costeroids, physicians classifying patients as creased responsiveness to methacholine has (non)asthmatic on the basis of a certain degree recently been shown to be present in more than on September 28, 2021 by guest. Protected copyright. of PEF variation'7 should take into account two thirds of patients with early COPD.27 In prior treatment with corticosteroids. our population a large overlap was found, not The improvement in PEF levels and variation only in PC20, but also in PEF variation between with inhaled corticosteroids was paralleled by patients with a symptom based diagnosis of an improvement in FEVy19 and by an im- asthma and COPD.26 Both PEF levels and PC20 provement in symptom scores. By contrast, are also unimodally distributed in the general although the greatest improvement in PC20 population.'228 Reliable separation of patients Table 3 Within-subject Spearman s rank correlation coefficients for longitudinal relations between diurnal variation in PEF and other markers of disease sevenity by treatment and diagnosis group (median values) 10g2PC20 Symptom score FEV,(0opred) BDR Total group median -0 40 0-25 -0-23 0.20 90% range -1 0 to 080 -0-66 to 0-76 -1l0 to 10 -l10 to 10 p* <0001 <0001 004 005 BA+PL -0-26 0 20 -0 20 0 20 BA+AC -0-32 0 32 -0 40 0 40 BA+CS -0 40 0 20 -0 20 0 20 Pt 0005 048 0-78 0 56 Asthma -0 40 0-32 -0 40 0 40 Asthmatic bronchitis -0 40 0.20 -0 40 0-26 COPD -0 40 0-13 0-20 -0-20 Pt 0-75 0-21 0-03 0-02 * Wilcoxon one sample signed rank sum test: non-parametric testing for symmetric distribution about zero. t Kruskal-Wallis one way analysis of variance; non-parametric testing for differences in means between diagnoses or treatment groups. Three patients with an undefined diagnosis are left out of the comparison of diagnoses. For abbreviations see table 1. 1114 Kers6ens, Brand, de _ong, Koeter, Postma with asthma from both normal subjects and in PEF levels and variation in PEF was found those with COPD on the basis ofany one single with bronchodilator treatment alone. We are, parameter is not therefore possible. however, cautious to suggest the absence of a Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from From a clinical point ofview, it would be very deleterious effect ofregular 12 agonist treatment valuable if diurnal variation in PEF, measured alone, both because of our large withdrawal before institution of treatment, was a good rate (mainly due to increases in pulmonary predictor of improvement with inhaled corti- complaints) and because all our patients re- costeroids as not all patients respond to ceived regular treatment with a 12 agonist corticosteroids.2930 In our study, however, base- (2000,g terbutaline per day). line diurnal variation in PEF was too weakly In conclusion, this randomised controlled related to changes in FEVy and PC20 (p = 0 29 longitudinal study of 141 patients shows that and 0 07, respectively) to help in deciding to inhaled corticosteroids improve morning and whom to administer inhaled corticosteroids. afternoon PEF levels, thereby reducing diurnal Baseline PC20, FEVI, bronchodilator response, variation in PEF. This should be kept in mind allergy, and smoking habits may be more useful when interpreting PEF variation for diagnostic in this respect.'931 classification in research and clinical practice. To date it is unclear how the variation in The relationship between diurnal variation in PEF should best be measured and expressed. PEF and both symptoms and objective para- Several points are currently unresolved: firstly, meters such as PC20 and FEV1 is relatively how many PEF readings should be taken each weak. We therefore regard home PEF mon- day, and when.32 It is clear that more readings itoring as an informative and valuable aid to will not only increase diurnal (circadian) PEF physicians but always in conjunction with other variation, but will also lead to somewhat more markers of disease severity. stable figures for PEF variation. However, in this 2 5 year clinical trial, as perhaps in routine The authors gratefully acknowledge the constructive criticism of S T Weiss as well as the support from three pharmaceutical patient care, it was thought that too frequent companies who supplied the study medication in identical measurements of PEF, especially at night, metered dose inhalers: Astra Pharmaceuticals, Boehringer In- gelheim, and Glaxo. would seriously hamper patient compliance. This study was supported by a government grant from the Secondly, Ryan and coworkers have advocated Netherlands' Health Research Promotion Program (SGO). The Dutch Chronic Non-Specific Lung Disease (CNSLD) the computation of the PEF amplitude as the group consists of a steering committee (K F Kerrebijn, PhH highest postbronchodilator value minus the Quanjer and H J Sluitert), of members from the Departments of Pulmonology of the University Hospital of Amsterdam lowest prebronchodilator value.'6 Expressing (E M Pouw, D F M E Schoonbrood, C M Roos, H M Jansen), amplitude in this way they found a considerably Groningen (P L P Brand, A de Gooyer, H A M Kerstjens, D S Postma, Th W van der Mark, H J Sluitert, G H higher correlation coefficient with PC20 (r= Koeter), Leiden (P M de Jong, P J Sterk, A M J Wever, - 8 1) than when calculating amplitude from J H Dijkman), Nijmegen (P N R Dekhuijzen, HT M Fol- prebronchodilator values only (r= -0O41). gering, C L A van Herwaarden), Rotterdam (S E Overbeek, J M Bogaard, C Hilvering) and Utrecht (S J Gans, H J J http://thorax.bmj.com/ This might effectively make home PEF meas- Mengelers, B A H A v d Bruggen-Bogaarts, J Kreukniet), urements more informative, probably by add- from the Departments of Pediatric Pulmonology at the Sophia Children's Hospital, Rotterdam (E E M van Essen-Zandvliet, ing another component of disease severity - K F Kerrebijn), the Juliana Children's Hospital, the Hague namely, the degree ofreversibility to a agonist (EJ Duiverman, J M Kouwenberg, J E Prinsen), University Hos- 02 pital ofGroningen (HJWaalkens,J Gerritsen, KKnol), from the - to the equation. When performing analyses Department of Allergology, University Hospital of Groningen with this measure of variation in PEF one (J G R de Monchy), from the Department of General Practice, University of Leiden (F W Dekker), from the Department of should then be aware of a higher inter- Psychiatry, University Hospital Leiden (A A Kaptein), and from relationship between the different markers of the Department of Physiology, University of Leiden (P J F M H counsel: S M D Merkus, Ph Quanjer). Scientific J Pocock, on September 28, 2021 by guest. Protected copyright. disease severity. Hughes, (London, UK), E R Bleecker, D A Meyers (Baltimore, A limitation of the current analysis is the USA). reduction in number of patients from the ori- 1 Haahtela T, Jnrvinen M, Kava T, Kiviranta K, Koskinen S, ginal group, even though the analysed group Lehtonen K, et al. Comparison ofa P,-agonist, terbutaline, was not significantly different from the group with an inhaled corticosteroid, budesonide, in newly de- scores tected asthma. N EnglJ Med 199 1;325:388-92. not selected. PEF records and symptom 2 British Thoracic Society RURCPL. Guidelines for man- on the diary cards were accepted only ifreliable agement ofasthma in adults: I. Chronic persistent asthma. - - data of the BMJ 1990;301:651-3. that is, sufficiently completed 3 British Thoracic Society RURCPL. Guidelines for man- prerandomisation visit were available (n = 204). agement of asthma in adults: II. Acute severe asthma. BMJ 1990;301:797-800. This is a prerequisite to study changes with 4 International Asthma Management Project. International treatment. As a result of withdrawal before the consensus report on the diagnosis and treatment of was asthma. Bethesda, Md. Publication no. 92-3091. NHLBI, 18-month visit,'9 the number of patients NIH, 1992. further reduced to 141. This second criterion 5 Cross D, Nelson HS. The role of the peak flow meter in was in order to term the diagnosis and management of asthma. Jf Allergy Clin employed study long Immunol 1991;87:120-8. variability and within-subject correlation co- 6 Clark NM, Evans D, Mellins RB. Patient use of peak flow efficients in a set of data. We re- monitoring. Am Rev Respir Dis 1992;145:722-5. complete 7 National Asthma Education Program, Expert Panel Report. analysed the data including all patients who Guidelines for the diagnosis and management of asthma. achieved a follow up of at least three months Bethesda, MD Publ. No. 91-3042: US Dept. of Health and Human Services, 1991. (n = 187 instead of 141) and found the changes 8 Griffin E, Hakansson L, Formgren H, J6rgensen K, Peterson in PEF levels and variation with treatment to C, Venge P. Blood eosinophil number and activity in relation to lung function in patients with asthma and with be highly comparable and all significance levels eosinophilia. AUley Clin Immunol 1991;87:548-57. unchanged. We therefore feel confident that 9 Molema J, van Herwaarden CLA, Folgering HThM. Effects of long-term treatment with inhaled cromoglycate and the data presented are representative of the budesonide on bronchial hyperresponsiveness in patients complete group, even though a relatively large with allergic asthma. Eur Respir Jf 1989;2:308-16. 10 Waalkens HJ, Gerritsen J, Koeter GH, Krouwels FH, van reduction of patients occurred. Aalderen WMC, Knol K. Budesonide and terbutaline or No change (improvement or deterioration) terbutaline alone in children with mild asthma: effects on Influence of treatment on peak expiratory flow and its relation to airway hyperresponsiveness and symptoms 1115 bronchial hyperresponsiveness and diurnal variation in 21 Quanjer PhH. Standardized lung function testing. Bull Eur peak flow. Thorax 1991;46:499-503. Physiopathol Respir 1983;19(Suppl 5):1-95. 11 van Essen-Zandvliet EEM, Hughes MD, Waalkens HJ, 22 American Thoracic Society. Standards for the diagnosis and Duiverman EJ, Pocock SJ, Kerrebijn KF, et al. Effects of care ofpatients with chronic obstructive pulmonary disease Thorax: first published as 10.1136/thx.49.11.1109 on 1 November 1994. Downloaded from 22 months treatment with inhaled corticosteroids and/or (COPD) and asthma. Am Rev RespirDis 1987;136:225-44. beta-2-agonists on lung function, airway responsiveness 23 Miller MR, Pincock AC. Predicted values: how should we and symptoms in children with asthma. Am Rev Respir use them? Thorax 1988;43:265-7. Dis 1992;146:547-54. 24 Norusis MJ. Statistical Package for the Social Sciences, PC 12 Higgins BG, Britton JR, Chinn S, Jones TD, Jenkinson D, version 4.0. Chicago: SPSS Inc, 1990. Burney PGJ, et al. The distribution of peak expiratory 25 Woolcock AJ, Yan K, Salome CM. Effect of therapy on flow variability in a population sample. Am Rev Respir Dis bronchial hyperresponsiveness in the long-term man- 1989;140:1368-72. agement of asthma. Clin Allergy 1988;18:165-76. 13 Quackenboss JJ, Lebowitz MD, Krzyzanowski M. The nor- 26 Brand PLP, Postma DS, Kerstjens HAM, Koeter GH, mal range of diurnal changes in peak expiratory flow rates. Dutch CNSLD Study Group. Relationship of airway hy- Relationship to symptoms and respiratory disease. Am perresponsiveness to respiratory symptoms and diurnal Rev Respir Dis 1991;143:323-30. peak flow variation in patients with obstructive lung dis- 14 Higgins BG, Britton JR, Chinn S, Cooper S, Burney PGJ, ease. Am Rev Respir Dis 1991;143:916-21. Tattersfield AE. Comparison of bronchial reactivity and 27 Tashkin DP, Altose MD, Bleecker ER, Connett JE, Kanner peak expiratory flow variability measurements for epi- RE, Wong-Lee W, et al. The lung health study: airway demiologic studies. Am Rev Respir Dis 1992;145:588-93. responsiveness to inhaled methacholine in smokers with 15 Ramsdale EH, Morris MM, Hargreave FE. Interpretation mild to moderate airflow limitation. Am Rev Respir Dis of the variability of peak flow rates in chronic bronchitis. 1992;145:301-10. Thorax 1986;41:771-6. 28 Rijcken B, Schouten JP, Weiss ST, Meinesz AF, de Vries 16 Ryan G, Latimer KM, Dolovich J, Hargreave FE. Bronchial K, van der Lende R. The distribution of bronchial re- responsiveness to histamine: relationship to diurnal vari- sponsiveness to histamine in symptomatic and in asympto- ation ofpeak flow rate, improvement after bronchodilator, matic subjects. Am Rev Respir Dis 1989;140:615-23. and airway calibre. Thorax 1982;37:423-9. 29 Corrigan CJ, Brown PH, Barnes NC, Szefler SJ, Tsai JJ, 17 Hetzel MR, Clark TJH. Comparison of normal and asth- Frew AJ, etal. Glucocorticoid resistance in chronic asthma. matic cicardian rhythms in peak expiratory flow rate. Glucocorticoid pharmacokinetics, glucocorticoid receptor Thorax 1980;35:732-8. characteristics, and inhibition of peripheral blood T cell 18 Josephs LK, Gregg I, Mullee MA, Holgate ST. Non-specific proliferation by glucocorticoids in vitro. Am Rev Respir bronchial reactivity and its relationship to the clinical Dis 1991;144:1016-25. expression of asthma. A longitudinal study. Am Rev Respir 30 Lane SJ, Lee TH. Glucocorticoid receptor characteristics Dis 1989;140:350-7. in monocytes of patients with corticosteroid-resistant 19 Kerstjens HAM, Brand PLP, Hughes MD, Robinson NJ, bronchial asthma. Am Rev Respir Dis 1991;143:1020-4. Postma DS, Sluiter HJ, et al. A comparison of bron- 31 Kerstjens HAM, Overbeek SE, Schouten JP, Brand PLP, chodilator therapy with or without inhaled corticosteroid Postna DS, Dutch CNSLD Study Group. Airways hy- therapy in obstructive airways disease. NEnglJMed 1992; perresponsiveness, bronchodilator response, allergy, and 327:1413-9. smoking habit predict improvement in FEV, during long- 20 Brand PLP, Kerstjens HAM, Postma DS, Sterk PJ, Quanjer term inhaled corticosteroid treatment. Eur Respir J 1993; PhH, Sluiter HJ, et al. Long-term multicentre trial in 6:868-76. chronic non-specific lung disease: methodology and base- 32 Gannon PFG, Newton DT, Pantin CFA, Burge PS. Effect line assessment in adult patients. Eur Respir 7 1992;5: of the number of peak expiratory flow readings per day 21-31. on diurnal variation. Eur Respiry 1992;5(Suppl 15):200s. http://thorax.bmj.com/ on September 28, 2021 by guest. Protected copyright.