Pilot study of risks and long-term effects of carbonmonoxide poisoning

We report on a pilot study intended to explore methods of investigating health effectsin survivors of accidental carbon monoxide poisoning requiring hospitalisation.These effects were to be studied via a questionnaire. We established that casescould be identified in centralised Scottish health record systems, and drew a smallsample, plus a matching sample of appendicitis cases as controls.Contact with the selected subjects was through records of their registrations with GPpractices. Questionnaires for 134 subjects were sent to GPs, with a request that theybe forwarded to the subjects. There were only a small number of refusals by theGPs, but response from the subjects was disappointing, leading to only 24 returnedquestionnaires (10 cases, 14 controls); a response rate of only 18%.Examination of data for the selected cases showed that a number had entriessuggesting that their poisoning may have been intentional rather than accidental. Itis therefore possible that estimates of accidental poisonings for study areoverestimates.Because of the low response rate to the questionnaire survey, we believe that theresults from a full study using these methods could produce results that are notrepresentative of the whole affected population, and we do not recommend that sucha study be commissioned. We discuss briefly some other possible routes ofinvestigation.