DOI: 10.1177/17562872261493698 ISSN: 1756-2872

Terminal digit preference: A potential bias in radiographic and pathologic evaluations of kidney cancer- retrospective single center study

Ofer N. Gofrit, Vladimir Yutkin, Liat Appelbaum, Amitay Lorber, Moti Duvdevani, Mohammad Zuaiter, Guy Hidas, Tzahi Neuman

Background

The tendency to prefer digits zero and five as the terminal digits under various conditions is termed terminal digit preference (TDP). TDP can potentially introduce a significant bias into the management of patients with kidney cancer.

Objectives

To investigate whether TDP occurs during radiographic and pathological evaluations of kidney cancer.

Design

Single-center, retrospective case series.

Methods

Patients with kidney masses were evaluated using contrast-enhanced computed tomography and then operated on. The maximal tumor diameter was recorded by a radiologist using electronic calipers and by pathologists using a ruler. In total, 905 patients who underwent surgery were analyzed.

Results

Zero and five were the terminal digits in 40.6% and 25.5% of the pathologic, and 44.8% and 19.8% of the radiographic measurements. This represents an extremely unlikely event compared with the expected even distribution of all digits (p<10 -13 ). TDP persisted at extreme statistical significance around the T-stage-determining values (4.0 cm, 7.0 cm, and 10.0cm).

Conclusions

This study found a strong TDP in both radiographic and pathological evaluations of kidney cancer. This phenomenon can bias correct stage assignment and expose patients to both over- and under-treatment. This bias may also reduce the power of statistical tests in research. Understanding the importance of TDP by radiologists, pathologists, clinicians, and researchers is the first step toward avoiding such misinterpretations. Since a high rate of TDP introduces a significant bias to the results, future studies in kidney cancer should monitor their database for TDP and report on it in their manuscripts.