After treatment for prostate cancer, it is standard for the treating physician to follow the patient's PSA, and inquire as to QOL changes. Thus, physician-obtained QOL data are nearly universally available, and any reported series of radical prostatectomy or radiation therapy will include an estimate of urinary incontinence, erectile dysfunction, bowel toxicity, etc. However, as the science of survey research began to be applied to cancer patients, it became clear that physician estimates and patient estimates of QOL often disagree.38 Litwin etal.39 utilized the CAPSURE database to assess differences in physician and prostate cancer patients' estimates of QOL using the SF-36 and UCLA PCI. Substantial differences were seen in assessment of physical, urinary, bowel, sexual function, fatigue and pain, with urologists generally underestimating quality of life. This trend has been noted by other studies,40-42 and explains why most modern QOL studies distribute questionnaires with return envelopes addressed to research coordinators, so as to remove the physician from the process as much as possible.
How do spouses perform as estimators of quality of life? Kornblith etal.43 studied a large sample of patients and spouses with validated instruments and found that spouses reported greater psychological distress but less sexual dysfunction than patients. Sneeuw et al.,44 noting poor response rates in studies of metastatic prostate cancer, compared patients and spouses assessments of QOL for D2 prostate cancer. Using the EORTC survey, they noted that proxies rated patients with more impairment in physical and role function, sleep disturbance, weight loss and sexual satisfaction. There was a low level of agreement for sexual function. However, when they looked at the severity of disagreement, there was 60% exact agreement and most of the disagreements were off by only one response category. In their estimate, their proxies rated QOL more accurately than the urologists in the Litwin study,39 and proposed that spouses may be a reasonable method of collecting QOL data.
Was this article helpful?