Clinical decisions associated with positron emission tomography in a prospective cohort of patients with suspected or known cancer at one United States center.

PURPOSE: In 2001, Medicare approved reimbursement of F-18 fluorodeoxyglucose positron emission tomography (PET) for a variety of cancers. PET has been observed to be more accurate than other imaging in cancer patients, but the impact of PET on management in routine practice is uncertain. PATIENTS AND METHODS: We studied a prospective cohort having noninvestigational PET at one university center. Before and after PET, a questionnaire was administered to solicit information regarding each physician's preceding actions, intended management, and probability estimates.
RESULTS: Seventy-one physicians provided data on 248 patients, of whom 40% had new or suspected cancer and 60% were undergoing restaging or had suspected recurrence. Lung, lymphoma, and head/neck cancers accounted for two thirds of cases. Sixteen physicians made 64% of requests. Physicians changed their intended management in 61% of patients (95% CI, 54% to 66%). For individual physicians ordering at least 10 scans, the average kappa was 0.16 (range, -0.04 to 0.36), reflecting only slight level of agreement between their before and after PET plan. PET was associated with a change in 90 (79%) of 114 patients if the pre-PET intended plan involved more testing or biopsy. In 32% of cases, physicians changed to a treatment from a nontreatment strategy. The therapeutic goal and mode changed in 22 (7%) and 21 cases (8%), respectively.
CONCLUSION: This study confirms that physicians often change their decision making based on PET. This impact is likely due to combined effects of PET's improved accuracy and reduced physician uncertainty. Physicians may also be overconfident in interpreting PET and use it as the final arbiter after an extensive evaluation in lieu of tissue biopsy.