Use of pulmonary function tests in the diagnosisof occupational asthmaWasif Anees, MB, BS

Objective: To discuss the different methods of assessing lung function measurements for the diagnosis of occupational asthma,focusing in particular on serial peak expiratory flow rate (PEFR) monitoring, including details on how PEFR records should bekept, plotted, and analyzed and limitations of the method.Data Sources: Published studies on the use of diagnostic methods in occupational asthma, expert opinion, and recently

obtained data from studies performed at a large occupational lung disease clinic.Study Sources: The expert opinion of the author was used to select the relevant data for review.Results: Objective methods are necessary for the diagnosis of occupational asthma, since clinical history alone is not a

satisfactory means of diagnosis. Serial PEFR monitoring has a high diagnostic sensitivity and specificity for occupational asthmaand is more useful than evaluation of cross-shift change in forced expiratory volume in 1 second or change in nonspecificbronchial hyperresponsiveness. Interpretation is best performed by expert visual evaluation of plots of maximum, mean, andminimum daily PEFR readings.Conclusions: Despite some limitations of the method, serial PEFR monitoring is usually the most appropriate first-line

investigation in workers suspected of having occupational asthma.Ann Allergy Asthma Immunol. 2003;90(Suppl):47–51.

INTRODUCTIONAlthough a patient’s medical history may suggest occupa-tional asthma, the history alone is not specific enough for thisdiagnosis to be established reliably. Malo et al1 reported that76% of patients referred to their clinic had an improvement inrespiratory symptoms while away from work but were sub-sequently found to have no objective evidence of occupa-tional asthma. Cote et al2 found that a history suggestiveof western red cedar asthma had a diagnostic specificityof 45%. Considering the implications of a diagnosis ofoccupational asthma, it is clear that this diagnosis must bemade objectively.This review was performed to discuss the different meth-

ods of assessing lung function measurements for the diagno-sis of occupational asthma, focusing in particular on serialpeak expiratory flow rate (PEFR) monitoring, including de-tails on how PEFR records should be kept, plotted, andanalyzed and limitations of the method. Data sources in-cluded published studies on the use of diagnostic methods inoccupational asthma, expert opinion, and recently obtaineddata from studies performed at a large occupational lungdisease clinic. The expert opinion of the author was used toselect the relevant data for review.

DIAGNOSING OCCUPATIONAL ASTHMASeveral objective methods of confirming a diagnosis of oc-cupational asthma are available. Although specific inhalation

challenge testing is considered to be the gold standard diag-nostic test, it is only available in a few centers, can betime-consuming to undertake, and false-positive and false-negative reactions can occur if performed improperly. Othermethods of assessing lung function cannot define the caus-ative agent but can demonstrate the work-related nature ofairflow obstruction. An improvement in nonspecific bron-chial hyperresponsiveness might be expected following re-moval from exposure in a worker with occupational asthma;however, Perrin et al3 found that a 3.2-fold change in hyper-responsiveness had a diagnostic sensitivity of only 43% anda specificity of 65%. Cote et al2 found that a twofold changein hyperresponsiveness had a sensitivity of 62% and a spec-ificity of 78% in occupational asthma due to western redcedar. Cross-shift evaluation of forced expiratory volume in 1second (FEV1) is difficult to perform in practice and is aninsensitive method; for example, Burge et al4 found that only33% of cases of occupational asthma due to colophonyshowed a greater than 10% cross-shift decrease in FEV1.Immunological evaluation can be useful in confirming sen-sitization to a particular agent. The presence of sensitizationas determined by skin prick tests or radioallergosorbent testsin combination with a suggestive history can be specific foroccupational asthma in certain circumstances (eg, for labora-tory animal workers or isocyanates, although for the latterthese are insensitive [approximately 20%]).5 With otheragents, a combination of history and evidence of sensitizationmay be sensitive for diagnosing occupational asthma but notnecessarily specific; for example, a suggestive history andevidence of sensitization were found to have a specificity ofonly 36% for latex in health care workers.6

Occupational Lung Disease Unit, Birmingham Heartlands Hospital, Bordes-ley Green East, Birmingham, England.Received for publication July 28, 2002.Accepted for publication in revised form October 14, 2002.

VOLUME 90, MAY, 2003 47

Serial PEFR MonitoringAssessment of serial measurements of PEFR at and awayfrom work is a simple and cost-effective way of confirmingoccupational asthma and has been recommended as a first-line investigation in suspected cases.7 The PEFR is defined asthe maximum flow achieved during expiration, deliveredwith maximal force, starting from the level of maximuminspiration. It can be recorded using simple portable meterssuch as the mini-Wright meter. Even in healthy individuals,there is a circadian rhythm in PEFR, with values tending to belowest in the early hours of the morning and maximal in theafternoon. This circadian rhythm is exaggerated in asthma.Numerous studies have found serial measurement of PEFR

to be both sensitive and specific in the diagnosis of occupa-tional asthma, with values ranging from 73 to 100% and 74 to100%, respectively.2,3,8–10 Assessment of changes in nonspe-cific bronchial hyperresponsiveness in addition to PEFRmonitoring does not appear to improve diagnostic accuracy.Also, unsupervised monitoring of FEV1 is not more accuratethan unsupervised PEFR monitoring. Leroyer et al10 found asensitivity of 55% and specificity of 89% for FEV1 comparedwith a sensitivity of 73% and specificity of 100% for PEFR.Instructing Workers on Keeping PEFR RecordsWorkers should be requested to record PEFR every waking 2hours for 4 weeks, including periods at and away from work.The best of three PEFR readings should be recorded on eachoccasion provided the best two readings were within 20L/min of each other. If possible, treatment should be avoided;however, if bronchodilator treatment is required, this shouldbe given after PEFR measurement. Details of the nature ofwork performed should be indicated on the diary cards. It isimportant to indicate waking, starting and finishing work, andsleep times, particularly if shift work is involved. We havefound that quality of PEFR records is best when recorded ondedicated diary cards (available at www.occupationalasthma.com).Data quantity has an important influence on diagnostic

sensitivity and specificity, which are best when readings aremade every waking 2 hours for 4 weeks.11 Performing recordsfor this duration and intensity requires a considerable degreeof cooperation and motivation on the patient’s part. Theshorter PEFR records with fewer readings may still be usefulfor diagnosing occupational asthma. However, sensitivity andspecificity decrease significantly when records are shorterthan 21⁄2 weeks or if there are fewer than four daily readings.Sensitivity is also reduced if there are fewer than 3 consec-utive days in each work period. In most cases of occupationalasthma, a 2-day rest period between work periods is enoughfor PEFR readings to improve significantly. However, occa-sionally, longer rest periods are required, particularly whenoccupational asthma is severe. At our tertiary referral center,we have found that approximately 50% of patients providerecords of adequate quality after being sent written instruc-tions on how to keep PEFR records before their first clinic

appointment. This improves to approximately 70% after be-ing given additional personal instruction.Plotting and Analysis of PEFR RecordsThe PEFR records are best plotted as daily average, maxi-mum, and minimum values, with days at work being easilydifferentiated from rest days (eg, by shading; Fig 1). Plots ofaverage reading for time of day for work and rest periods arealso useful (Fig 2). Confounding factors must be taken intoconsideration before analysis. These include lower respira-tory tract infections, periods where treatment has changed,and learning and laze effects. These periods should be ex-cluded from analysis. If there are days when the subject is notexposed to the causative agent at work, these should beanalyzed as rest days.Analysis is best performed visually by an expert, although

there is a degree of intraexpert and interexpert variability.12–14Computer packages have recently become available that al-low easy data entry and graph plotting. Some, such as theOasys-2 computer program (Occupational Lung DiseaseUnit, Birmingham Heartlands Hospital, Birmingham, En-gland), can also determine the probability of occupationalasthma, have a high degree of diagnostic sensitivity andspecificity (75% and 94%, respectively), and are entirelyrepeatable.15Various rule-based or statistic-based techniques of assess-

ing PEFR records for occupational effect have been sug-gested, although expert visual analysis is probably better,since the expert can visually evaluate several measures ofairflow variability and is not restricted to using any oneindex. Cote et al16 suggested that an index based on thedifference between maximum PEFR at weekends and mini-mum PEFR on working days was the best indicator ofoccupational asthma. However, this index is unhelpful indifferentiating patients with occupational asthma from non-occupational asthmatic patients who have a high diurnalvariation in PEFR, because the index is in essence a measureof diurnal variation. The difference between mean PEFR onrest days and mean PEFR on workdays is likely to be the bestindex for differentiating workers with occupational asthmafrom those with nonoccupational asthma. Assessment ofchanges in diurnal variation between work and rest periods isa relatively insensitive approach (W. Anees, unpublisheddata, 2001). The PEFR diurnal variation within the normalrange does not exclude occupational asthma; even in nonoc-cupational asthma, increased PEFR diurnal variability has alow diagnostic sensitivity.17,18 In occupational asthma, theeffects of work exposure on PEFR are superimposed on thenormal circadian increase in PEFR during the day; hence,PEFR variability on workdays may actually be lower than onrest days (Fig 2). The measure of airflow variability thatmakes visual evaluation of serial PEFR records useful inoccupational asthma relates to differences in mean PEFRbetween rest days and workdays rather than evaluation ofPEFR diurnal variability or changes in diurnal variability.There is sometimes difficulty in interpreting PEFR records

48 ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY

from workers with symptoms suggestive of occupationalasthma that show relatively small but consistent deteriora-tions in PEFR during work exposure. Some argue that smallchanges in mean PEFR are likely to be due to irritants innature. However, neither PEFR diurnal variability nor themagnitude of the PEFR response to work exposure predictsthe nature of the bronchial inflammatory marker response towork exposure.19 Furthermore, when occupational asthma issevere, mean rest-work PEFR differences may be small unlessPEFRs are performed for a longer rest period.The PEFR records appear to have limited use in the diagnosis

and monitoring of nonoccupational asthma, probably becausethere is a large overlap in the distribution of PEFR diurnal

variability between asthmatic and healthy individuals, resultingin a poor diagnostic sensitivity. The PEFR records appear to bemuch more useful in differentiating workers with occupationalasthma from healthy individuals or nonoccupational asthmaticpatients. This is because, provided there is an adequate quantityof PEFR data, mean rest-work PEFR differences are likely to besmall in nonoccupational asthmatic patients or even healthypatients occupationally exposed to irritant agents (W. Anees,unpublished data, 2001). It is the presence of consistent differ-ences in mean PEFR between rest and work periods that are bestevaluated visually that make PEFR monitoring useful in occu-pational asthma. The magnitude of this rest-work PEFR differ-ence appears to relate poorly to disease severity; the magnitude

Figure 1. Peak expiratory flow rate (PEFR) plot from aworker with occupational asthma caused by isocyanates.Days are marked along the x-axis; lines represent dailymaximum, mean, and minimum values. Shaded areas rep-resent days at work; unshaded areas are rest days. There isa decrease in PEFR during days at work.

Figure 2. Hourly peak expiratory flow rate (PEFR) plotfrom a worker with occupational asthma caused by trigly-cidyl isocyanurate in powder paints. Mean PEFR valuesfor time of day are plotted separately for work and restdays; the shaded area represents the usual work shift. Onrest days, PEFR increases during the day to a peak in theearly evening. However, during workdays, this circadianpattern is blunted by exposure to the causative agent,leading to a relatively low degree of PEFR diurnal varia-tion. Classic early and late reactions may also be seen onhourly PEFR plots.

VOLUME 90, MAY, 2003 49

of PEFR responses to bronchoconstrictor stimuli tend to be morevariable and generally less marked than changes in supervisedFEV1 measurements.20,21

Falsification of PEFR RecordsSome authorities have suggested that PEFR records are oflimited value in the evaluation of occupational asthma be-cause readings are unsupervised and because of the possibil-ity that records may be falsified. Quirce et al,22 for example,found that 23% of readings were inaccurate and 23% ofreadings invented. In the United Kingdom, falsification ofPEFR records is not considered to be a major problem. Usingelectronic PEFR meters in patients unaware that readingswere being logged, we found similar rates of falsification tothose described by Quirce et al; however, this did not alter theinterpretation of the PEFR record for occupational effect inmost patients, and PEFR records were still believed to bevalid.23 This is probably because when record falsification didtake place it was usually on occasions when the patients hadnot measured their PEFR and had invented the reading;however, these invented readings were usually not exagger-ated. The consistently high values for diagnostic sensitivityand specificity for PEFR records in occupational asthmareported in numerous studies suggest that the method isrelatively robust in coping with these inaccuracies. Recordfalsification is more likely to occur in patients specificallybeing evaluated for compensation.24 We believe that falsifiedrecords can often be spotted by an expert, but if there isconcern about record falsification, electronic logging metersshould be used or the diagnosis confirmed by specific inha-lation challenge testing.Limitations of Serial PEFR MonitoringThere are several limitations of using PEFR monitoring todiagnose occupational asthma. It does not indicate the caus-ative agent. This should be confirmed by evidence of specificsensitization to a relevant agent (if available) or specificbronchial challenge testing, although where a well recognizedsensitizer is present this may not be necessary (eg, isocya-nates in a spray painter). Serial PEFR monitoring is lessuseful when exposure is intermittent, although to some de-gree this can be overcome by the worker carefully document-ing on the PEFR chart the nature of work and agents to whichhe or she may be exposed. The PEFR records are obviouslyunsuitable if the worker has already been removed fromexposure or if the worker is having severe reactions at work,because it is safer to remove the worker and arrange specificbronchial challenge testing. One needs to be aware of thephenomenon of pseudo-occupational asthma, which can oc-cur if workers wake later on rest days, thereby missing the“morning dip” in PEFR.

CONCLUSIONSerial PEFR monitoring is an appropriate tool for the initialevaluation of most workers suspected of having occupationalasthma. Provided records are of adequate data quantity, the

method has high diagnostic sensitivity and specificity. Inter-pretation for occupational effect is best performed by expertvisual evaluation of daily maximum, minimum, and meanplots, with workdays being differentiated from rest days.

REFERENCES1. Malo JL, Ghezzo H, Archeveque J, Lagier F, Perrin B, CartierA. Is the clinical history a satisfactory means of diagnosingoccupational asthma? Am Rev Respir Dis. 1991;143:528–532.

2. Cote J, Kennedy SM, Chan-Yeung M. Sensitivity and specific-ity of PC20 and peak expiratory flow rate in cedar asthma. JAllergy Clin Immunol. 1990;85:592–598.

3. Perrin B, Lagier F, L’Archeveque J, Cartier A, Boulet LP, et al.Occupational asthma: validity of monitoring of peak expiratoryflow rates and non-allergic bronchial responsiveness as com-pared to specific inhalation challenge. Eur Respir J. 1992;5:40–48.

4. Burge PS, Perks WH, O’Brien IM, Burge A, Hawkins R, et al.Occupational asthma in an electronics factory: a case controlstudy to evaluate aetiological factors. Thorax. 1979;34:300–307.

5. Tee RD, Cullinan P, Welch J, Burge PS, Newman Taylor AJ.Specific IgE to isocyanates: a useful diagnostic role in occupa-tional asthma. J Allergy Clin Immunol. 1998;101:709–715.

6. Vandenplas O, Binard-Van Cangh F, Brumagne A, Caroyer J,Thimpont J, et al. Occupational asthma in symptomatic workersexposed to natural rubber latex: evaluation of diagnostic proce-dures. J Allergy Clin Immunol. 2001;107:542–547.

7. Moscata G, Godnic-Cvar J, Maestrelli P, Malo JL, Burge PS,Coifman R. Statement on self-monitoring of peak expiratoryflows in the investigation of occupational asthma. Eur Respir J.1995;8:1605–1610.

8. Burge PS, O’Brien I, Harries M. Peak flow rate records in thediagnosis of occupational asthma due to colophony. Thorax.1979;34:308–316.

9. Burge PS, O’Brien I, Harries M. Peak flow rate records in thediagnosis of occupational asthma due to isocyanates. Thorax.1979;34:317–323.

10. Leroyer C, Perfetti L, Trudeau C, L’Archeveque J, Chan-YeungM, Malo J. Comparison of serial monitoring of peak expiratoryflow and FEV1 in the diagnosis of occupational asthma. Am JRespir Crit Care Med. 1998;158:827–832.

11. Anees W, Huggins V, Burge PS. Peak expiratory flow recordsfor the diagnosis of occupational asthma: the effect of dataquantity on diagnostic sensitivity and specificity. Eur Respir J.2001;18(Suppl):281s.

12. Malo J, Cote J, Cartier A, Boulet L, L’Archeveque J, Chan-Yeung M. How many times per day should peak expiratory flowrates be assessed when investigating occupational asthma? Tho-rax. 1993;48:1211–1217.

13. Liss GM, Tarlo SM. Peak expiratory flow rates in possibleoccupational asthma. Chest. 1991;100:63–69.

14. Zock JP, Brederode D, Heederik D. Between- and within-observer agreement for expert judgement of peak flow fromgraphs from a working population. Occup Environ Med. 1998;40:969–972.

15. Burge PS, Pantin CF, Newton DT, Gannon PF, Bright P, et al.Development of an expert system for the interpretation of serialpeak expiratory flow measurements in the diagnosis of occupa-tional asthma. Occup Environ Med. 1999;56:758–764.

50 ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY

16. Cote J, Kennedy SM, Chan-Yeung M. Quantitative versus qual-itative analysis of peak expiratory flow in occupational asthma.Thorax. 1993;48:48–51.

17. Lewis S, Weiss ST, Britton JR. Airway responsiveness and peakflow variability in the diagnosis of asthma for epidemiologicalstudies. Eur Respir J. 2001;18:921–927.

18. Goldstein MF, Veza BA, Dunsky EH, Dvorin DJ, BelecanechGA, Haralabatos IC. Comparisons of peak diurnal expiratoryflow variation, postbronchodilator FEV1 responses, and metha-choline inhalation challenges in the evaluation of suspectedasthma. Chest. 2001;119:1001–1010.

19. Anees W, Huggins V, Pavord ID, Robertson AS, Burge PS.Occupational asthma due to low molecular weight agents: eo-sinophilic and non-eosinophilic variants. Thorax. 2002;57:231–236.

20. Leone FT, Mauger EA, Peters SP, Chinchilli VM, Fish JE, et al.The utility of peak flow, symptom scores, and beta-agonist useas outcome measures in asthma clinical research. Chest. 2001;119:1027–1033.

21. Giannini D, Paggiaro PL, Moscato G, Gherson G, Bacci E, et al.Comparison between peak expiratory flow and forced expira-

tory volume in one second (FEV1) during bronchoconstrictioninduced by different stimuli. J Asthma. 1997;34:105–111.

22. Quirce S, Contreras G, Dybuncio A, Chan-Yeung M. Peakexpiratory flow monitoring is not a reliable method for estab-lishing the diagnosis of occupational asthma. Am J Respir CritCare Med. 1995;152:1100–1102.

23. Anees W, Huggins V, Burge PS. Peak expiratory flow recordsin occupational asthma are still valid in most workers despitethe presence of some fabricated readings. Am J Respir Crit CareMed. 2002;165:A524.

24. Malo J, Trudeau C, Ghezzo H, L’Archeveque J, Cartier A. Dosubjects investigated for occupational asthma through serialpeak expiratory flow measurements falsify their results? J Al-lergy Clin Immunol. 1995;96:601–607.

Requests for reprints should be addressed to:Wasif Anees, MB, BSOccupational Lung Disease UnitBirmingham Heartlands HospitalBordesley Green EastBirmingham B9 5SS, England

VOLUME 90, MAY, 2003 51