Study of white coat effect on the prognosis of women with breast cancer
TO THE EDITOR:
We were fascinated to read the study of white coat effect on the prognosis of women with breast cancer (Costa et al. 2003). However there are several unanswered questions:
Methodology
Most studies quantify white coat effect (WCE) by the difference between office and home ambulatory blood pressure values (Pickering et al. 2002), with home seen as the more tranquil environment. In this study WCE was quantified as the difference between blood pressure measured by the doctor and the nurse, in a clinic environment.
This design was a record-based, retrospective study. It is important to know whether an attempt was made to control for variables that may be associated with blood pressure measurement (Reeves 1995). For example: doctor and nurse interpersonal variations in measurement techniques, training, equipment, timing of measurements during the consultation and the chemotherapeutic regimes the women were having. It is also likely that the nature of the initial doctor consultation and the subsequent nurse encounters differed with respect to the goals and activities of the appointments. Such variations could easily result in significant variations in blood pressure, particularly within the stressful environment of the oncology clinic. We also found no mention of patients who did not have three blood pressure measurements and were therefore excluded from the analysis.
Analysis
We wonder whether the analysis controlled for baseline characteristics? Can the authors speculate as to why diastolic WCE was higher than systolic? It is important to note that diastolic blood pressure is technically more difficult to measure and hence an inherent risk of inaccuracy could well account for these results.
Conclusions
For the above reasons, we contest whether this study shows WCE to be more common in women with breast cancer, than the general female population. The question as to whether quantifying the WCE would be useful in the oncology setting, other than to highlight the women's potential increased risk of a cardiovascular event (Mancia 2000), we believe requires more research, particularly through prospective multivariate analysis.