Frequency and intensity of postoperative surveillance after curative treatment of pancreatic cancer: a cost-effectiveness analysis.

BACKGROUND: Few data exist to guide oncologic surveillance following curative treatment of pancreatic cancer. We sought to identify a rational, cost-effective postoperative surveillance strategy.
METHODS: We constructed a Markov model to compare the cost-effectiveness of 5 postoperative surveillance strategies. No scheduled surveillance served as the baseline strategy. Clinical evaluation and carbohydrate antigen (CA) 19-9 testing without/with routine computed tomography and chest X-ray at either 6- or 3-month intervals served as the 4 comparison strategies of increasing intensity. We populated the model with symptom, recurrence, treatment, and survival data from patients who had received intensive surveillance after multimodality treatment at our institution between 1998 and 2008. Costs were based on Medicare payments (2011 US dollars).
RESULTS: The baseline strategy of no scheduled surveillance was associated with a postoperative overall survival (OS) of 24.6 months and a cost of $3837/patient. Clinical evaluation and CA 19-9 assay every 6 months until recurrence was associated with a 32.8-month OS and a cost of $7496/patient, with an incremental cost-effectiveness ratio (ICER) of $5364/life-year (LY). Additional routine imaging every 6 months incrementally increased total cost by $3465 without increasing OS. ICERs associated with clinic visits every 3 months without/with routine imaging were $127,680 and $294,696/LY, respectively. Sensitivity analyses changed the strategies' absolute costs but not the relative ranks of their ICERs.
CONCLUSIONS: Increasing the frequency and intensity of postoperative surveillance of patients after curative therapy for pancreatic cancer beyond clinical evaluation and CA 19-9 testing every 6 months increases cost but confers no clinically significant survival benefit.