[Skip to Content]
[Skip to Content Landing]
November 29, 1965

How Reliable Are Tumor Measurements?

Author Affiliations

From the Department of Statistics and the Medical School (Dr. Gurland) and the Division of Clinical Oncology and the Department of Surgery (Dr. Johnson), University of Wisconsin, Madison.

JAMA. 1965;194(9):973-978. doi:10.1001/jama.1965.03090220029008

Discrepancy between measurements of tumors made independently by different physicians can be substantial. There are also differences between measurements of tumors made by the same physician on different occasions, but some physicians are significantly more consistent than others. The techniques by which measurements are made, namely, readings and tracings in the case of lesions shown on x-ray film, affect this relative consistency—some physicians may be relatively consistent by one technique but relatively inconsistent by another. Probability considerations concerning the wrong decisions that can be made regarding change of size indicate that descriptions of changes in small lesions are very unreliable.