Patient surveillance after potentially curative treatment for ovarian cancer

Introduction: Patient surveillance after initial curative-intent treatment for ovarian carcinoma has important clinical, legal, and financial implications for patients, physicians, and society. The particular modalities used in surveillance by clinicians, and how frequently they are recommended, are not known at present. We sought to determine the current follow-up practice patterns of highly credentialed experts who treat patients with ovarian cancer and also provide care for them after initial treatment. Materials and Methods: We devised a survey featuring vignettes describing 4 patients with ovarian carcinoma of different clinical stages and questions based on the vignettes. Surveys were mailed to the 943 members of the Society of Gynecologic Oncologists (SGO). The survey requested demographic data. Those indicating that they performed ovarian cancer surgery and also participated in long-term follow-up care were asked how often they requested 11 discrete follow-up evaluations for their patients treated for cure in the first ten post-operative years. These 11 modalities include all those mentioned in the relevant medical literature. Simple descriptive statistics (mean, standard deviation, median, and range) were calculated. Results: Of the 943 gynecologic oncologists surveyed, 323 (34%) responded; 283 (30%) were evaluable. The most frequently recommended modalities for each year were office visit, pelvic examination and serum CA-125 level. There was marked variability in the recommended frequency of their use. For example, the number of office visits recommended in post-operative year 1 for a patient with stage III ovarian carcinoma with > 1 cm of residual tumor after surgery ranged from 2-12. No imaging studies were frequently recommended. The frequency of testing decreased with increasing post-operative years for all tests that were commonly recommended. For example, the median number of times the serum CA-125 level was recommended for a high-risk stage III patient was 4 during the first year, decreasing to 1 by year 5. Conclusion: The details of current surveillance testing in patients treated for ovarian cancer have not been documented previously. There is little agreement in the literature about the most appropriate surveillance strategy. Ours is the first data on this topic, to our knowledge. There is marked variability in the intensity of surveillance among SGO members, which is prima facie evidence of overuse and/or underuse and/or misuse of scarce medical resources. Further research should aim to identify the sources of this variability.