Search

veterans with one of the defined syndromes and 20 unaffected veterans. Compared with the controls, the cases had a variety of subtle, but detectable neurological abnormalities. The researchers concluded that the three syndromes are variants of a single pathological process but that they present differently because of differences in the extent of damage or the part of the nervous system involved. Finally, the team analysed veterans’ self-reported exposure to potential causative agents (eg, insect repellents, flea collars, and the anti-nerve gas medicine pyridostigmine), and their “perceived chemical weapons exposure”. This analysis found that the risk of “impaired-cognition” syndrome was highest among younger veterans, those who wore flea collars, and those in security details; the risk of “confusion-ataxia” syndrome was highest among those reporting chemical weapons exposure or side-effects from pyridostigmine tablets; and the risk of “arthro-myo-neuropathy” syndrome was highest among older veterans who had used government-issued insect repellent and who had had side-effects after taking pyridostigmine. The researchers suggest that different combinations of chemicals, perhaps working synergistically, and the ages of exposed veterans may explain the three syndromes. But, in an accompanying editorial, Philip Landrigan, of the Presidential Advisory Committee on Gulf War Veterans’ Illnesses (see p 188), points to limitations in the study. First, since only one battalion was studied, the results may not be generalisable; second, only 41% of the battalion participated in the surveys, so there may be selection bias; third, virtually all information on illnesses was self-reported—detailed clinical tests were done on only 23 veterans; and, finally, all exposure data was selfreported. These limitations, Landrigan said, “substantially weaken the authors’ strong conclusions”.

The poor correlation between the marks awarded in the clinical tests in medicine and surgery, and between the medical clinical test and the written objective paper, suggested the need for more detailed investigations into the reliability of clinical examinations. Two examiners independently marked 29 candidates in a clinical examination in medicine, and the marks awarded were compared. In another examination 28 different students were examined on one short case by two examiners and the examination was recorded on videotape. This was subsequently marked by a further twelve examiners and again on two further occasions by the original examiners. Marks awarded by one examiner often varied considerably from those awarded by another examiner for the same performance. Inconsistency in the same examiner was also evident. This observed variation must be more widely recognised in determining the fate of candidates in clinical examinations.

The relative efficacy of clinical and laboratory tests used to monitor disease activity in rheumatoid arthritis was determined by consensus analysis in a study of 21 patients treated for 6 months. Erythrocyte sedimentation rate (ESR), which is influenced by the anaemia of chronic disease and by variation in the blood concentration of acute-phase proteins, was the most effective single test. ESR was a better guide to disease severity than measurement of plasma viscosity, serum C-reactive protein, and serum orosomucoid—tests that reflect the blood concentration of acute-phase proteins only. Clinical tests performed poorly unless combined with laboratory data in a numerical index. Consensus analysis can be used to assess the efficacy of clinical and laboratory tests and to identify redundant tests.