Analytical decision model for the cost-effective management of solitary pulmonary nodules

S. S. Gambhir*, J. E. Shepherd, B. D. Shah, E. Hart, C. K. Hoh, P. E. Valk, T. Emi, M. E. Phelps

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

215 Scopus citations

Abstract

Purpose and Methods: Multiple strategies are currently being used to manage patients who present with indeterminate solitary pulmonary nodules (SPN). We have used decision-analysis models to assess the costeffectiveness of various strategies for the diagnosis and management of SPN. Four decision strategies were compared: a wait and watch strategy, a surgery strategy, a computed tomography (CT)-based strategy, and a CT-plus-positron emission tomography (PET) strategy. An incremental cost-effectiveness ratio (ICER) was used to compare all strategies to the wait and watch strategy. Results: A CT- plus-PET strategy was the most costeffective over a large pretest likelihood (probability of having a malignant nodule), with a range of 0.12 to 0.69. Furthermore, within this likelihood range, the potential cost savings of using the CT-plus-PET strategy over the CT strategy ranged from $91 to $2,200 per patient. This translates to a yearly national savings of approximately $62.7 million. Conclusion: Decision-analysis modeling indicates the potential cost-effectiveness of [18F]2-fluoro-2-deoxy-Dglucose (FDG)-PET in the management of SPN. Furthermore, the decision trees developed can be used to model various features of the management of SPN, including modeling the cost- effectiveness of other newly emerging technologies.

Original languageEnglish (US)
Pages (from-to)2113-2125
Number of pages13
JournalJournal of Clinical Oncology
Volume16
Issue number6
DOIs
StatePublished - Jun 1998

ASJC Scopus subject areas

  • Oncology
  • Cancer Research

Fingerprint

Dive into the research topics of 'Analytical decision model for the cost-effective management of solitary pulmonary nodules'. Together they form a unique fingerprint.

Cite this