Send letter to journal:
Re: Response to David Fishwick

Dear Editor

Experts were given no clinical details except for times of waking and sleeping, and times of starting and leaving work. They were asked to make judgements based on the peak expiratory flow record alone, similar to the judgements made by the Oasys program. Oasys-2 has been shown to have a sensitivity of around 70%, when tested against independent objective diagnoses (mostly specific bronchial provocation testing), and a specificity of 94%. The need is therefore to achieve increased sensitivity.

The experts underscored compared with Oasys-2, and did not appear to be detecting work-related changes missed by Oasys-2. In practice tests are interpreted in the light of clinical information (requiring expertise), however in our practice occupational asthma often occurs in unlikely places, and is frequently diagnosed when the specific exposures are unknown.

We hope we have provided a tool for use by the non-expert inthe initial assessment of occupational asthma. We agree that these records need to be made as soon as the diagnosis is suspected, and before workers are removed from their jobs. Supervising such records does however needs a degree of expertise, with particular emphasis on recording working times, keeping treatment constant, and recording the timings of readings. Help is provided for this on the website occupationalasthma.com, as well as suitable record forms with instructions which can be downloaded.

Ideally, OASYS should be used interactively. The patient returns to clinic with his PEF record stored in an electronic meter. The clinician and patient review the record together. This allows the clinician to ask those question suggested by the record such as "did you have a respiratory infection last week" (if there was an unexpected fall in PEF crossing work/rest interfaces), or "remind me of your work pattern on the 25th of last month" (when a single work-day shows no deterioration when others do). Thus the integration of clinical information and record is even closer enhancing the diagnostic toolkit mentioned by David Fishwick and colleagues.

Send letter to journal:
Re: Man versus machine

Dear Editor

We read with interest the article by Baldwin et al.[1] relating to the level of agreement between expert clinicians and OASYS software when making a diagnosis of occupational asthma. Our clinical unit uses OASYS plotting regularly, and find this of great use as one element of the diagnostic toolkit available for the confirmation of a diagnosis of occupational asthma.

We were interested to note that there was a low level of agreement between experts and OASYS when asked to interpret peak expiratory flow (PEF) records, but agreement within experts was better. We would be interested to know whether the information provided to the experts on the nature of work was used in determining their final outcome, i.e. if an individual was working with a known sensitiser or was in a perceived high risk job, did this influence the outcome more than the graphical and mathematical data.

Practically in clinic, a decision is made to perform regular PEF monitoring in those patients who are thought to have a reasonable chance of having occupational asthma, as judged by the clinical information to date. Perhaps a further study option would be to give experts clinical data first (more like the real life situation), and ask for a likelihood of occupational asthma based on this assessment, followed by a revision of that likelihood after PEF data is supplied. Would then revealing the work effect score lead to further revision of the perceived estimate? Individual experts may be more or less swayed by the clinical data due to variation in their own practice, types of cases seen, geographical location and so on.

Again, experts were deemed to "under report" possible cases of occupational asthma. Whilst this may indeed be the case, an alternative explanation is that the experts were more realistic, taking into account the clinical likelihood as well as the PEF pattern. OASYS systems clearly invoke complex comparisons between known cases of occupational asthma and the record being assessed.

Finally, the authors suggest that PEF interpretation is best left to experts. Whilst we agree that expert centres, consistently diagnosing occupational asthma are needed, as many as one in ten adult asthmatics are likely to have a substantial effect from work.[2] It is therefore important for all such patients in the UK to have access to competent individuals trained to assess these patients. This is where OASYS (or similar) systems are likely to be very important as an initial screen, and could be carried out by primary care or occupational health nurses or other competent non clinical people in the workplace. This would enable patients currently working to undergo PEF assessment, as opposed to the common situation of seeing patients in secondary care already following prolonged period of sickness absence, making diagnosis even more challenging.

At present, the consistency of diagnosis of occupational asthma throughout the UK is likely to be highly variable. We are currently involved in a multicentre UK based study assessing the application of the toolkit to diagnose occupational asthma, and it is evident that practice remains disparate between various expert centres.

We are sure that the future of occupational asthma evaluation will and should rely on OASYS like programmes, but that the diagnosis must be seen also in broader terms, taking into account clinical, immunological and exposure data.

References

(1) DR Baldwin, P Gannon, P Bright, DT Newton, A Robertson, K Venables, B Graneek, RD Barker, A Cartier, J-L Malo, M Wilsher, CFA Pantin, and PS Burge. Interpretation of occupational peak flow records: level of agreement between expert clinicians and OASYS-2. Thorax 2002;57: 860-864.

(2) Blanc P, Toren K. How much adult asthma can be attributed to occupational factors? Am J Med 1999;107:580-587.