Economic evaluations of gastric and pancreatic cancer

Anne Elixhauser, Michael T. Halpern

Research output: Contribution to journalReview articlepeer-review

22 Scopus citations


The total cost of cancer care in the US is about $146 billion, of which pancreatic cancer comprises $2.6 billion (1.8% of the total) and gastric cancer comprises $1.8 billion (1.3%). We have reviewed published studies presenting economic analysis of treatment or follow-up for patients with pancreatic or gastric cancer. Relatively few studies report on economic evaluations of pancreatic cancer care. There are also few economic studies for gastric cancer, although we identified three cost-effectiveness analyses. In general, economic analyses in these areas are relatively unsophisticated, relying on charge data or simple multipliers (e.g., average cost per day in the hospital multiplied by days in the hospital), and are often limited to in-hospital costs (particularly studies for pancreatic cancer). A wide range of costs is included in these studies and a variety of methodologies for assigning costs are used, making comparisons between studies difficult. Future health economics research in this area should evaluate the costs and effectiveness of alternative practice patterns for gastric and pancreatic cancer; conduct additional cost-effectiveness analyses of chemotherapeutic interventions; consider quality of life, survival, stage at diagnosis, patient-borne costs, and complications of therapy; and, take advantage of administrative data from large populations.

Original languageEnglish (US)
Pages (from-to)1206-1213
Number of pages8
Issue number26
StatePublished - 1999
Externally publishedYes


  • Cost of illness
  • Cost-benefit analysis
  • Cost-effectiveness analysis
  • Gastric cancer
  • Pancreatic cancer or neoplasm
  • Stomach neoplasm

ASJC Scopus subject areas

  • Hepatology
  • Gastroenterology


Dive into the research topics of 'Economic evaluations of gastric and pancreatic cancer'. Together they form a unique fingerprint.

Cite this