Authors' Reply

E B Wu 鄔揚正

Dear Editor

With reference to the letter from Dr Tsoi on 20th August 2002, we have the following comments.

This group of 250 patients drawn from the 829 consecutive referrals for coronary angiography was the second in a series of studies on the chest pain score. The small numbers drawn from a large group is not due to overselectiveness, but rather because we were at a large tertiary center that preformed a lot of coronary angiograms on patients with known myocardial disease. In our first study, we compared chest pain characteristics between patients with coronary artery disease and those with normal coronary anatomy. The three items of the chest pain score emerged from this study. Many researches would be satisfied with this and claim that these three factors could "predict" coronary artery disease. We were more careful as we realised the enormous implications of this data. We conducted a second study with 250 patients from our coronary angiography lists. This study confirmed our findings in the case control study. Worried by the same questions that Dr Tsoi has asked - whether these findings can be applied to a primary care setting, we conducted a third study of 363 patients who presented with chest pain either to the open access chest pain clinic directly or through their GPs. The three items of the chest pain score still had predictive validity for coronary artery disease on univariate analysis. The reproducibility score had non-significant predictive value when data from exercise testing was added to the multivariate analysis (as would be expected). Therefore, we can say with certain confidence that this chest pain score works well in the primary care setting and should project well onto the population seen by the GP. Having the same concerns as Dr Tsoi, a fourth study is in the process of being conducted to apply this chest pain score in GP practices across the region. This study will validate the chest pain score and answer the questions posed more directly.

Unfortunately, consensus among cardiologists rarely exists. Like many academic studies the reader has to make up his or her own mind as to the validity of the data presented and alter their management of their patients accordingly. After all this is data published in a peer reviewed journal and not a key point in a textbook. As to the medico-legal implications, we have already emphasised that the GP should not be overly concerned with this aspect, and to do so is ethically suspect. Reassurance is essential and may benefit the patient more than any other test or therapy. A good physician places the benefit of the patient in high regard and may occasionally choose to silently carry the worries of the patient while reassuring them strongly for their good. This is the art of medicine, in which I am sure Dr Tsoi is a far greater expert than I.

E B Wu BSc, MBBS, MRCP
Medical Officer
Cardiology Division,
Department of Medical and Therapeutics,
Prince of Wales Hospital.