Article Text


Evidence based guidelines for the prevention, identification, and management of occupational asthma
  1. P J Nicholson1,
  2. P Cullinan2,
  3. A J Newman Taylor2,
  4. P S Burge3,
  5. C Boyle4
  1. 1Procter & Gamble, Whitehall Lane, Egham, Surrey, UK
  2. 2National Heart & Lung Institute, Manresa Road, London, UK
  3. 3Heartlands Hospital, Bordesley Green East, Birmingham, UK
  4. 4Health & Safety Executive, Magdalen Court, Bootle, UK
  1. Correspondence to:
 Dr P J Nicholson
 British Occupational Health Research Foundation, 6 St Andrews Place, Regents Park, London NW1 4LB, UK;


Background: Occupational asthma is the most frequently reported work related respiratory disease in many countries. This work was commissioned by the British Occupational Health Research Foundation to assist the Health and Safety Executive in achieving its target of reducing the incidence of occupational asthma in Great Britain by 30% by 2010.

Aim: The guidelines aim to improve the prevention, identification, and management of occupational asthma by providing evidence based recommendations on which future practice can be based.

Methods: The literature was searched systematically using Medline and Embase for articles published in all languages up to the end of June 2004. Evidence based statements and recommendations were graded according to the Royal College of General Practitioner’s star system and the revised Scottish Intercollegiate Guidelines Network grading system.

Results: A total of 474 original studies were selected for appraisal from over 2500 abstracts. The systematic review produced 52 graded evidence statements and 22 recommendations based on 223 studies.

Discussion: Evidence based guidelines have become benchmarks for practice in healthcare and the process used to prepare them is well established. This evidence review and its recommendations focus on interventions and outcomes to provide a robust approach to the prevention, identification, and management of occupational asthma, based on and using the best available medical evidence. The most important action to prevent cases of occupational asthma is to reduce exposure at source. Thereafter surveillance should be performed for the early identification of symptoms, including occupational rhinitis, with additional functional and immunological tests where appropriate. Effective management of workers suspected to have occupational asthma involves the identification and investigation of symptoms suggestive of asthma immediately they occur. Those workers who are confirmed to have occupational asthma should be advised to avoid further exposure completely and early in the course of their disease to offer the best chance of recovery.

Statistics from

Work related asthma, where there is an association between symptoms and work, includes: work aggravated asthma (pre-existing or coincidental new onset asthma worsened by workplace exposure); and occupational asthma (OA; new onset asthma caused by workplace exposure). OA can be either: hypersensitivity induced OA characterised by latency between first exposure to a substance at work and the development of symptoms; or irritant induced OA that occurs shortly after a high exposure to an irritant gas, fume, or vapour at work. Almost 90% of cases of occupational asthma are of the hypersensitivity type1–6 and are the focus for this review.

This review was commissioned and funded by the British Occupational Health Research Foundation (BOHRF) to provide a robust approach to the prevention, identification, and management of OA, based on and using the best available medical evidence.7


We systematically searched Medline and Embase from 1966 and 1974 respectively to the end of June 2004. Pairs of working group members reviewed over 2500 abstracts independently to select 474 papers for further, independent critical review; single case studies and narrative reviews were excluded. Data from 223 papers were entered into evidence tables7 and contributed to the evidence statements.

Criteria for grading evidence and recommendations were developed primarily to guide inferences about treatment; questions about disease frequency, aetiology, diagnosis, and prognosis require other hierarchies.8 There are few meta-analyses and randomised controlled trials (RCTs) related to OA, and so little Scottish Intercollegiate Guidelines Network (SIGN) grading system level 1 evidence. Therefore we graded evidence statements (ES) and recommendations using both the revised SIGN grading system 2000 for levels of evidence (table 1) and recommendations (table 2), and the modified 1995 Royal College of General Practitioners (RCGP) three star system used in the BOHRF Occupational Health Guidelines for the Management of Low Back Pain at Work (table 3).9 In accordance with current opinion, recommendations appear in behaviourally specific terms.10

Table 1

 Revised SIGN grading system for levels of evidence

Table 2

 Revised SIGN grading system for recommendations

Table 3

 Modified RCGP three star system



OA is the most frequently reported occupational respiratory disorder in westernised industrialised populations.5,11,12 Where mining is common, for example, South Africa and the Czech Republic, OA is the second commonest occupational respiratory disorder after pneumoconiosis.4,13

Main messages

  • Reducing airborne exposure reduces the number of workers who develop occupational asthma.

  • Occupational rhinitis may indicate an increased risk of a worker developing occupational asthma, particularly when IgE associated and especially in the year after the onset of rhinitis.

  • Health surveillance can detect occupational asthma at an earlier stage of disease and contribute to improved clinical outcome.

  • Asking workers about symptoms that improve regularly when away from work is more sensitive than enquiring about symptoms that deteriorate when at work.

  • Outcome is better in workers who have shorter duration of symptoms prior to diagnosis, relatively normal lung function at diagnosis, and no further exposure to the causative agent after diagnosis.

There are no complete registries of OA so the true frequency of the disease is unknown. Published frequencies come from surveillance schemes, compensation registries, or epidemiological studies of the relation between asthma and occupation. The incidence, predominant causative agents, and jobs most commonly reported to incur risk depend on the methodology of data collection, case definition, and predominant work sectors and occupations. Studies generally do not distinguish between hypersensitivity and irritant causes. One systematic review showed an attributable risk of 9%,14 while the highest scoring studies showed an attributable risk of 15%, a figure supported by another review.15

ES1*** SIGN 2++

Occupational factors are estimated to account for 9–15% of cases of asthma in adults of working age, including new onset or recurrent disease.14,15

ES2*** SIGN 2++

The annual population incidence of occupationally related asthma ranges from an estimated 12–170 cases per million workers with an estimated mean of 47 cases per million workers.6,14,16,17


The population incidence of OA may be underestimated by as much as 50%.18


The reported incidence of OA has not decreased in recent years.19–21

Policy implications

  • Employers should implement programmes to prevent occupational asthma by removing or reducing exposure to its causes.

  • Employers and their health and safety personnel should provide regular health surveillance (respiratory questionnaire with functional and immunological tests, where appropriate) to workers where a risk of occupational asthma is identified.

  • Employers and their health and safety personnel should ensure that measures are taken to ensure that workers diagnosed as having of occupational asthma avoid further exposure to its cause in the workplace.

ES5*** SIGN 2++

The most frequently reported agents include isocyanates, flour and grain dust, colophony and fluxes, latex, animals, aldehydes, and wood dust.3,4,6,12,13,17,19,22–24

ES6*** SIGN 2++

The workers most commonly reported to surveillance schemes of OA include paint sprayers, bakers and pastry makers, nurses, chemical workers, animal handlers, welders, food processing workers, and timber workers.3,6,12,13,17,21–25

ES7** SIGN 2+

The workers reported from population studies to be at increased risk of developing asthma include bakers, food processors, forestry workers, chemical workers, plastics and rubber workers, metal workers, welders, textile workers, electrical and electronic production workers, storage workers, farm workers, waiters, cleaners, painters, plastic workers, dental workers, and laboratory technicians.26–29

Four main risk factors have been identified for OA: the causative factor of exposure to an agent at work; the predisposing factors of atopy and genetic predisposition; and the contributing factor of cigarette smoking.

A direct relation between OA and exposure to an agent at work occurs with acid anhydrides,30,31 cimetidine,32 colophony,33 enzymes,34–38 green coffee and castor bean,39 flour allergens,40–45 crab,46 isocyanates,47–49 laboratory animal allergens,50–52 piperazine,53 platinum salts,54 prawns,55 and western red cedar.56 Most studies also showed a positive exposure-response relation for sensitisation. Studies limited to sensitisation showed a relation with exposure to acid anhydrides,57 enzymes,37,58,59 laboratory animals,60 and platinum salts.61

ES8*** SIGN 2++

The risk of sensitisation and OA is increased by higher exposures to many workplace agents.30–32,34–56

Differences in ascertaining atopy, from personal or family history of allergy to tests for specific immunoglobulin E (IgE), can cause inconsistencies between observations. Atopy is reported to increase the risk of OA in workers exposed to enzymes,38,62,63 isocyanates,47,64 laboratory and other animals,50–52,64–74 bakery allergens,44,75–79 and some reactive dyes.80 Other studies show no association between atopy and OA due to exposure to cork,81 isocyanates,48,82,83 enzymes,35 glutaraldehyde,84 salmon,85 crab,86 hexahydrophthalic anhydride,87 platinum salts,61,88 and plicatic acid.89 Atopy is associated with increased risk of sensitisation in workers exposed to enzymes,36,37,42,58,63,90–93 green coffee and castor bean,39,94 bakery allergens,42–44,75,76,95,96 laboratory animals,50 crab,86 prawn,55 and acid anhydrides.57,97

ES9*** SIGN 2++

Atopy increases the risk of developing OA caused by exposure to many high molecular weight (HMW) agents that induce the production of specific IgE antibodies.38,50–52,62,63,65,66,68–79

That only some workers develop OA despite similar exposures, suggests underlying genetic susceptibility. Studies have examined genetic predisposition to OA attributed to isocyanates,98–105 complex platinum salts,106 western red cedar,107 acid anhydrides,108 and laboratory animal proteins.70,109 Studies had small sample sizes and their findings were inconsistent or unreplicated.

ES10** SIGN 2−

Genetic polymorphisms that code for human leucocyte antigen (HLA) class II genes or respiratory antioxidant mechanisms may predispose to OA for a number of agents.70,98,101–103,107

Smoking increases the risk of OA in workers exposed to: isocyanates,47,64,83 platinum salts,54,88,111 salmon,85 and snow crab.86 Smoking increases the risk of sensitisation in studies with exposure to green coffee and castor bean,39,94 platinum salts,61,110,111 prawn,55 and flour.95 The role of smoking is unclear for OA due to other agents. Some studies show increased risk of laboratory animal asthma in smokers,50,74,112 whereas others show no effect.51,65,68,69,113 For acid anhydrides, studies show both negative31,87 and positive97 correlation with specific IgE. Conflicting evidence also exists for detergent enzymes.38,114 In bakery workers, one study showed an increased risk of sensitisation in smokers,95 but the risk of OA does not appear to be increased.42,44,75,95

ES11** SIGN 2+

Cigarette smoking can increase the risk of developing OA with some sensitising agents.47,54,64,83,85,86,88,111

Co-morbid rhinitis or rhino-conjunctivitis is reported in 45–90% of subjects suffering from IgE associated OA attributable to acid anhydrides,115,116 laboratory animals,50,68,69 snow crab,86 and wheat flour.44

ES12** SIGN 2+

Occupational rhinitis and OA frequently occur as co-morbid conditions in IgE associated OA.19,44,50,68,69,86,115–117

ES13** SIGN 2+

Rhino-conjunctivitis is more likely to appear before the onset of IgE associated OA.19,50,68,69,115,117,118

ES14* SIGN 2−

The risk of developing OA is highest in the year after the onset of occupational rhinitis.19,118

While the latent interval between first exposure and the onset of OA can extend to many years,19,119–121 risk is highest soon after first exposure to laboratory animal allergens,50,52,65,68,69,112 isocyanates,122 platinum salts,54,88,111 and azodicarbonamide.123

ES15** SIGN 2+

IgE sensitisation and OA are most likely to develop in the first years of exposure for workers exposed to azodicarbonamide, enzymes, complex platinum salts, isocyanates, and laboratory animal allergens.50,52,54,65,68,69,88,111,112,122,123


With any study of preventive measures, it is difficult to distinguish the relative effect of one measure against another, since they are usually implemented as a composite programme. Studies show that reducing exposure leads to fewer cases of OA from acid anhydrides,31,124 detergent enzymes,34,36 isocyanates,49 laboratory animals,66,125 and latex.126–128

ES16** SIGN 2+

Reducing airborne exposure reduces the number of workers who become sensitised and who develop OA.31,34,36,49,66,125–128

Respiratory protective equipment (RPE) only offers protection when worn properly, removed safely, and either replaced or maintained regularly. Studies are few and small; one observed a significant association between symptoms and even brief removal of RPE.48

ES17* SIGN 3

The use of RPE reduces the incidence of, but does not completely prevent, OA.48,83,129

Pre-placement examinations should be used to establish a baseline for periodic health surveillance rather than to detect and exclude susceptible individuals from workplaces. Little is known about host susceptibility factors, except for atopy in those exposed predominantly to high molecular weight agents. The efficiency of screening out susceptible job applicants depends partly on the frequency of the trait in the general population.

ES18* SIGN 3

The positive predictive values of screening criteria are too poorly discriminating for screening out potentially susceptible individuals, particularly in the case of atopy where the trait is highly prevalent.67,77,111,123,130–132

ES19* SIGN 3

A previous history of asthma is not significantly associated with OA.67,68

Periodic health surveillance aims to identify sensitised workers or cases of OA at early and reversible stages of disease. There are few, and no concurrent comparison studies of the efficacy of health surveillance for OA. The only study from which valid conclusions can be drawn is of isocyanate workers, in whom health surveillance was linked to a mandatory programme of control of exposure. Workers with isocyanate induced asthma were diagnosed sooner after symptoms began, had better lung function and better outcome than workers who had asthma attributable to other workplace agents133 although the results might partly be attributable to concomitant reduction in exposure. Few studies have evaluated the separate components of health surveillance, questionnaires, spirometry, and identification of specific IgE. In one, all true cases of OA were identified by questionnaire, spirometry producing many false positives due to poor inspiratory effort and no additional cases of asthma.134 In another, spirometry detected one case of OA in addition to two cases identified by questionnaire.135 Skin prick and serological tests can detect specific IgE in workers who have become sensitised to HMW agents and a few low molecular weight (LMW) agents (complex platinum salts, acid anhydrides, and some reactive dyes).

ES20* SIGN 3

Health surveillance can detect OA at an earlier stage of disease and outcome is improved in workers who are included in a health surveillance programme.133

ES21* SIGN 3

Screening questionnaires may lead to an underestimate of the prevalence of asthmatic symptoms.136,137

ES22* SIGN 3

Spirometry detects few cases of OA that would not otherwise be detected by respiratory questionnaire.134,135

ES23** SIGN 2+

Skin prick testing and blood sampling of exposed workers to conduct immunological tests is feasible in the workplace.138–140

ES24** SIGN 2+

Prospective surveillance for the development of specific IgE antibodies can be used as part of a broader risk management programme to reduce the incidence of OA.36,90

Identification and evaluation of OA

Most evidence relating to the diagnosis of OA emanates from specialist settings where the prior probability of disease and positive predictive values of tests may be high. History alone produces false positive diagnoses, requiring further validation. Serial measurement of peak expiratory flow (PEF) is the simplest initial investigation. When performed and interpreted to validated standards there are few false positive results, but about 20% false negatives. Skin prick or blood tests for specific IgE are available for most HMW agents and a few LMW agents but few standardised allergens are commercially available. OA can usually be diagnosed without specific bronchial provocation testing (SBPT).

The symptoms of any type of asthma have high sensitivity but lower specificity, whereas the question “have you been told by a doctor that you have asthma” has a higher specificity but lower sensitivity.141 Asking about deterioration at work is an insensitive method of diagnosing OA. Questioning about symptoms that are better on days away from work has a sensitivity of 58–100% for validated OA and is enhanced by asking about symptoms improving on holiday. Work related symptoms are common in those with negative SBPT.

ES25** SIGN 2+

In the clinical setting questionnaires that identify symptoms of wheeze and/or shortness of breath that improve on days away from work or on holiday have a high sensitivity, but relatively low specificity for OA.138,142–147

ES26* SIGN 3

Free histories taken by experts have high sensitivity, but their specificity may be lower.117,142,143,145,148–150

No good studies compare cross-shift changes with SBPT. Such testing is unlikely to be either sensitive or specific since measures of airflow obstruction show diurnal variation in most normal workers that is increased in most asthmatics. In one study a fall in FEV1 of >10% post-shift was found in 5% of asymptomatic workers and 32% of those with work related asthma symptoms.

ES27* SIGN 3

Pre- to post-shift changes in lung function cannot be recommended for the validation or exclusion of OA.151,152

Serial PEF was studied in small case series of patients attending specialist clinics or workplace surveys. Over 70% of subjects in four clinical series and each of the workforce populations returned acceptable records. A single case series of 74 patients attending a specialist clinic reports the highest combination of sensitivity and specificity with PEF measured at least four times daily. Less frequent readings produced higher specificity but lower sensitivity. Six series report high levels of agreement (averaging 80%) between expert assessors. A single series, using non-expert assessors, reports much lower inter-observer agreement. Eight case series report direct and blinded comparisons of serial PEF measurement and either SBPT or an expert diagnosis based on a combination of other evidence. Sensitivities and specificities were consistently high, averaging 80% and 90% respectively. Three case series compare visual inspection of PEF records by experts with a variety of statistical indices. One computed analysis reported sensitivity of 75% and specificity of 94% when analysis was calibrated using the opinion of one expert in cases whose OA was mostly attributable to LMW agents.

ES28** SIGN 3

Acceptable PEF series can be obtained in around two thirds of those in whom a diagnosis of OA is being considered.139,146,147,153–158

ES29* SIGN 3

The diagnostic performance of serial PEF measurements falls when fewer than four readings a day are made.159

ES30** SIGN 3

There is high level of agreement between expert interpretations of serial PEF records.155,159–164

ES31** SIGN 3

The sensitivity and specificity of serial PEF measurements are high in the diagnosis of OA.146,147,155,159,162,164–166

ES32** SIGN 3

Statistical analysis of serial PEF measurements is of limited diagnostic value compared to expert interpretation.147,155,162

ES33** SIGN 2+

Computed analysis of PEF records has good diagnostic performance.160,167

Many studies show that non-specific bronchial hyper-reactivity is normal in 5–40% of SBPT positive workers. Testing with higher concentrations of methacholine or histamine at which some non-asthmatics react reduces the number of, but still leaves some, non-reacting asthmatics. A normal test of non-specific reactivity is not sufficiently specific to exclude OA.

ES34*** SIGN 2++

A large number of concordant studies from different centres using different methodologies showed that increased non-specific reactivity is often found in workers with OA. There are however many reports of normal methacholine or histamine reactivity within 24 hours of exposure in workers with confirmed OA.142–145,149,150,162,166,168–177

ES35** SIGN 2−

Changes in non-specific reactivity at and away from work alone have only moderate sensitivity and specificity for diagnosis.146,162,177

ES36* SIGN 3

Paired measurements of non-specific reactivity may be achieved in the workplace.177,178

The sensitivities and specificities of skin prick or serological tests to detect specific IgE vary between allergens and with positive cut-off levels. Serological tests may not be as sensitive as skin prick testing.179 The presence of specific IgE confirms sensitisation to an agent at work, but alone does not confirm OA, or necessarily its cause. There is a high false positive rate although, with HMW agents, few false negatives. Specific IgE is an insensitive but specific test for isocyanate induced OA,180 although this may depend on the time since last exposure. A small study reported greater sensitivity for MDI (83%) than TDI (27%).181

ES37** SIGN2+

Both skin prick and serological tests are highly sensitive for detecting specific IgE and OA caused by most HMW agents, but are not specific for diagnosing OA.51,182

ES38** SIGN2+

Both skin prick and serological tests are sensitive for detecting specific IgE and OA caused by acid anhydrides and some reactive dyes, but have a lower specificity for diagnosing OA.179,183–186

ES39** SIGN2+

Skin prick tests are highly sensitive but less specific for OA caused by complex platinum salts.138,174,175

SBPT is used as the gold standard for OA diagnosis, making assessments of its diagnostic validity difficult. Standardised methods are lacking for many occupational agents. The threshold exposure increases with time since last exposure, making tests less sensitive after prolonged absence from work. Individuals can have non-specific reactions to SBPT at concentrations likely to be found in the workplace and negative SBPT with otherwise good evidence of OA when challenge concentrations are below occupational exposure standards.151,152,170,173,176

ES40− SIGN 4

Carefully controlled specific challenges come closest to a gold standard test for some agents causing OA.

ES41− SIGN 4

A negative test in a worker with otherwise good evidence of OA is not sufficient to exclude the diagnosis.

Management of the worker with OA

The outcome of interventions following a confirmed diagnosis of OA may depend on several factors, including the worker’s age and the causative agent. Early diagnosis and early avoidance of further exposure, either by relocation of the worker or substitution of the hazard, offer the best chance of complete recovery. If these are impossible, workers should be relocated to low or occasional exposure areas and have increased health surveillance. Studies investigating the effectiveness of RPE in those with OA are limited to small studies in provocation chambers or limited case reports. The risk of unemployment may148 or may not,187,188 be higher than in other adult asthmatics and may fall with increasing time from diagnosis.189 No studies make direct comparisons between different systems of rehabilitation. There is one small RCT of the effect of inhaled corticosteroids on recovery from OA after ceasing exposure.

ES42*** SIGN 2+

The symptoms and functional impairment of OA caused by various agents may persist for many years after avoidance of further exposure to the causative agent.89,122,190–209

ES43*** SIGN 2++

The likelihood of improvement or resolution of symptoms or of preventing deterioration is greater in workers who have no further exposure to the causative agent.89,204,205,210–216

ES44** SIGN 2+

The likelihood of improvement or resolution of symptoms or of preventing deterioration is greater in workers who have relatively normal lung function at the time of diagnosis.89,193,201,205,206,217,218

ES45** SIGN 2+

The likelihood of improvement or resolution of symptoms or of preventing deterioration is greater in workers who have shorter duration of symptoms prior to diagnosis.89,189,204–206,213,218,219

ES46** SIGN 2+

The likelihood of improvement or resolution of symptoms or of preventing deterioration is greater in workers who have shorter duration of symptoms prior to avoidance of exposure.189,193,205,213,218,219

ES47* SIGN 3

Redeployment to a low exposure area may lead to improvement or resolution of symptoms or prevent deterioration in some workers, but is not always effective.85,204,205,210,211,220,221

ES48* SIGN 3

Air fed helmet respirators may improve or prevent symptoms in some but not all workers who continue to be exposed to the causative agent.204,222–226

ES49** SIGN 2−

Approximately one third of workers with OA are unemployed up to six years after diagnosis.148,187–189,207,209,227–229

ES50** SIGN 2−

Workers with OA suffer financially.148,188,207,227,228,230

ES51− SIGN 4

Systems that incorporate retraining may be more effective than those that do not.227,231

ES52* SIGN 1+

Inhaled corticosteroids used after cessation of exposure may provide small clinical benefits to workers with OA.161


Recommendations appear in table 4 and are followed by references to supporting evidence statements in brackets.

Table 4

 Principal recommendations


BOHRF Research Working Group: AJ Newman Taylor (Chair), PJ Nicholson (Vice Chair), PS Burge, C Beach, P Cullinan, C Francis, PFG Gannon, C Boyle, M Levy, R Miguel, MJ Nieuwenhuijsen, RG Rawbone, D Romano-Woodward, AJ Scott, O Tudor, EV Warbrick.

BOHRF External Reviewers: CAC Pickering, SC Stenton. Patient representative: S Ozanne.

View Abstract


  • Financial support: This study was supported by a grant from the British Occupational Health Research Foundation, 6 St Andrew’s Place, London NW1 4LB. Catherine Boyle was funded by the Health & Safety Executive.

  • Competing interests: none

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.