Answers are grounded in this paper and cite their sources.
U patient-oriented, and quite properly so.Whether one is describing treatment or the life expectancy of patients with untreatable disease, one is concerned with the brute facts of survival or death.A loss is just as much of a loss if it is due to complications of treatment rather than to primary effects of the cancer.Treatment is successful just so far as it removes increased risk of death from the patient and promises him normal chances of further survival.The best-known methods of end-result reporting are designed with this in mind.'* 10Sometimes, however, one may wish to study the pathological process per se, to seek a view of the disease unobstructed by competing causes of death.One may wish to separate failures of treatment from costs of that treatment, such as radiation pneumonitis, pancytopenia, or ureteral fistulas.One may not be as interested in relative survival (survival compared to some normal group) as with actual cancer deaths that measure the progress of the disease.This is particularly true when one wishes to identify species of cancer by their different rates of growth and spread and to correlate these properties with their gross and microscopic appearance.One can bring mathematical and graphic aids to this problem but this can be done easily only with diseaseoriented data.For these purposes, modifications of the usual methods of end-result reporting and analysis are needed.These exist,ll but insofar as studies of cancer go, they have received little attention and less use, especially in the American literature.Only in Swedish re-portslE,20 have I found them used in a way similar to that advocated here.Although clinicopathological studies supposedly are descriptions of tumor behavior, they seem particularly to lack the type of presentation that describes From the Department of Pathology, Memorial Hospital for Cancer and Allied Diseases, 444 E. 68th St., New York 21, N.Y.
No takes yet. Share an insight, caveat, or question.
John W. Berg (1964) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: