Multicentre drug trials, ethics approval and death of patients.

摘要

The paper by Christie et al. is alarmist, emotive and without foundation. It is based on an invalid calculation of the annual reduction in cancer mortality and makes unwarranted assumptions about how this incorrect figure relates to purported delays in Ethics Committee assessment. By proposing the insulting acronym DIABOLECAL, the authors deliberately slight the good work of Australian Ethics Committees.The claim of 360 fewer deaths from cancer each year assumes that the reduction in cancer mortality rate observed between 1995 and 2001 has continued since (invalidated) and that the overall population is constant. In fact, the Australian population increases at about 1.15% per year, from a 2001 base of 18.77m. Table 1 shows that, at the claimed 'conservative' annual 1% reduction in the cancer mortality rate from the baseline 2001 datum (191 /100 000), cancer deaths tend to increase because the rate of change of population exceeds the presumed reduction in mortality rate. If the authors had used the reduction in cancer mortality rate measured at2001 (1.6%), an estimated annual reduction of approximately 160 is obtained. Thus, under the most reasonable assumptions, the claimed annual cancer death reduction of 360 in the study is a fiction and any calculation that relies on it is invalid.