Editor, Professor Fletcher raises important points in his editorial 'Expensive medical technology -- we can, but should we?' Aust Prescr 1994;17:54-5. His concern is shared by the Australian and New Zealand Association of Physicians in Nuclear Medicine which believes that quality assurance and technology assessment are essential to the rational use of these procedures.1 However, I wish to take issue with several points in his editorial.
He refers to a study2 that is said to use 'the most modern methods of thallium scanning', which shows that 'only age and definite evidence of coronary disease were identified as statistically significant predictors of postoperative cardiac adverse events'. But 'definitive evidence of coronary disease' and the most common 'cardiac adverse event', perioperative ischaemia, were defined using clinical or ECG criteria, and therefore a correlation is not unexpected. Furthermore, the method of analysis of scans was not that usually applied and is likely to lead to false positive and false negative results.
Variation in the use of thallium scanning relates to many factors including access and the understanding of its role by referring practitioners. In general, it is not a self referred procedure.
When examining the efficacy of a diagnostic test, outcome measures such as mortality and infarction rates may not always be appropriate as many factors can intervene between test result and outcome, including what the referring practitioner does with the information, patient preferences and the quality and nature of subsequent procedures. The use of intermediate outcomes is preferred, but they must be specific to each context.
R. McLean
President
Australian and New Zealand Association
of Physicians in Nuclear Medicine
Sydney, NSW.