Postradiotherapy morbidity in long-term survivors after locally advanced cervical cancer: how well do physicians' assessments agree with those of their patients?

Abstract

Descriptions of late morbidity after radiotherapy in cervical cancer survivors (CCSs) are usually based on observations made by physicians, and rarely by patients themselves. We describe and compare physician-assessed morbidity with patient-rated symptoms more than 5 years after pelvic radiotherapy. In 147 CCSs treated between 1994 and 1999 at The Norwegian Radiumhospital, morbidity data were regularly documented by physicians at least for 5 years after radiotherapy. Information on patient-rated symptoms was collected by a questionnaire from 91 (62%) of the 147 survivors after a median follow-up time of 96 months (65-131 months). The results were compared with physician-assessed morbidity scores recorded at 5 years, and to selected normative data using descriptive statistics. Physician-assessed morbidity data were modeled using Kaplan-Meier method. Agreement between physician data and patient data was expressed using weighted kappa statistics. The 5-year Kaplan-Meier estimates of physician-assessed intestinal, bladder, and vaginal morbidity Grade 3-4 were 15%, 13%, and 23%, respectively. The prevalence of patient-rated severe symptoms from these organs was much higher (intestines 45%, bladder 23%, and 58% vaginal discomfort among sexually active CCSs). Poor agreement was confirmed by low values of kappa: For bladder the concordance was slight (kappa = 0.16) and for intestine it was fair (kappa = 0.27). Stress incontinence, diarrhea, nausea, and sexual problems were significantly (p < 0.001) more prevalent when compared with a control sample from the general female population. Morbidity is common after pelvic radiotherapy. However, our data indicate that physicians underreport patients symptoms. It is important to incorporate patient-reported outcomes in the evaluation of treatment-related morbidity.