Post ventolin readings / non reproducible readings in Oasys

Here are a few more questions that you may be able to answer right now: in your methods, you said that you excluded a few records because they were not adequate: what were the reasons for excluding them? It was surely not based on reproducibility of the readings as in theory the workers enter only the best of 3 readings. In OASYS, there does not seem to be any place for entering post-Ventolin data per se. There is no special mark on the graph nor any place in the database to signify that a particular PEF reading was done after Ventolin. Do you have any suggestions for that? If PEF readings are frequently done after Ventolin, it will obviously affect the score and how do you correct for that? If a record contains only PEF in the morning and nothing in the afternoon and evening, would you disregard it? at least this is what your methods suggest. Amitiés,

Oasys, Specialist, 2/20/2006, 2/20/2006,

We have never excluded poorly reproducible readings, as we have never measured reproducability formally. Instructions are to make as many readings as necessary to achieve the best 2 within 20 litres/min. How often this is achieved is unknown to me. We instruct workers to measure PEF before bronchodilator, and to use the bronchodilator as often on days off as on work days. I think either the pef goes down with exposure and can be measured, or the treatment goes up. It is very difficult to incorporate both in an analysis. Abreviated instructions are printed on the downloadable pef record form from this website. We have changed our criteria for acceptability over time, and have looked at reducing the number of readings a day, without a huge loss of sensitivity. We remove obviously prefabricated records, sections containing major changes in treatment, respiratory infections etc. At the moment minimal criteria are 5 complexes (about 3 weeks), minimum 3 day consecutive work periods, usual exposure during the record, and 4 evenly spaced readings per day. One reading at 5pm is not bad however. I will ask wasif anees to send you his data on this. Sherwood

2/20/2006