Implicit rationing in the physician-patient choice of lung cancer treatment
Date
1994
Authors
Arndt, Kevin Charles
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This study uses data collected from non-small cell lung cancer patients attending the Victoria Clinic of the British Columbia Cancer Agency to analyze implicit rationing in the physician-patient choice of lung cancer treatment. An ordered logit model is estimated to identify the impacts of disease states and personal attributes on the probability of a patient receiving alternative degrees of treatment aggressiveness. Logistic probability estimates for the various treatment strategies are linked to corresponding treatment costs, and expected cost-age profiles are calculated for an illustrative patient profile. Direct comparisons of expected cost-age profiles are conducted within a ceteris paribus framework to investigate implicit non-disease criteria in the rationing of lung cancer treatment.
The results indicate that non-small cell lung cancer treatment is implicitly rationed according to several personal attributes of a patient. Main empirical results are: (a) Significant differences in expected costs are attributable to patient age, with progressively lower levels of treatment expenditures among older age categories; (b) Patients who have never been married sustain 48 percent to 68 percent less in expected treatment costs than patients who are widowed, divorced, or currently married; (c) Physician perceptions of patient health status, as revealed by the Karnofsky Performance Index, contribute to differences among expected treatment costs for patients assigned a rank of 100 or 90, a rank of 80, or a rank of 70 or less; (d) Expected costs of treatment do not significantly differ with respect to patient gender. The study concludes that high cost lung cancer treatments may already be rationed implicitly to a much greater degree than is generally assumed.