In a recently published, fictionalized account of a case of suspected scientific misconduct involving the study of a drug anticipated to have anticancer effects, evidence of the drug's positive tumor effects in mice were purposefully withheld from scientific peer review.1 Withholding the data was a choice made by a senior researcher because sharing the findings would have been complicated personally and professionally for him. One outcome of these data being kept silent was that a drug with potential benefit for cancer treatment became a casualty. Although fictionalized, the account was informed by a collection of actual experiences.

In a recent, nonfictional academic example, a study team chose to withdraw its data from review because of team discomfort that the data would be offensive to colleagues who might, as a result of the offense, be less collaborative in the future. Being silenced, the data did not risk offending others, but patients' suffering related to the topic of study was not formally addressed either.

In another recent example, researchers quietly shared the unrelenting pressure to show scholarly productivity in terms of funded grant submissions. To respond to that pressure, already collected data were not being fully analyzed, and instead, energies were directed toward preparing the next grant. The unanalyzed data were thereby silenced as assuredly as the previous 2 examples of withheld and withdrawn data.

Data, by their very existence, are an intervention because they can prevent, describe, explain, or predict illness and treatment outcomes. Data can and do affect human behavior, ideas, thoughts, and relationships. In this way, data demonstrate their powerfulness and commonly create these effects in ways that often surprise. Data can be offensive and burdensome. Data can immediately or overtime directly clash with established beliefs or even revered and unquestioned ideals. Data can cause us to collide with our own beliefs but with now-altered logic, which leaves us little option but to reconsider our world views.

Data can be wrong or can be erroneously interpreted, 2 possibilities that risk causing harm. But if we withhold, withdraw, or underanalyze data, we cannot know the full benefit or harm of those data, which are outcomes that in their own rights are informative. Data are meant to have scientific and lay public scrutiny. Both data and their interpretations are meant to be peer reviewed an unending number of times, over time, and by very disparate individuals. Data kept silent serve no health cause and no cancer patient.

Most certainly, we, as researchers, ought not to speak prematurely about data or at least not without carefully denoting the data's limitations and the caveats and cautions to interpreting the data at an early point. But speaking at an early point may be more defensible than silencing data for personal and professional reasons of discomfort. If a researcher allows data to be silent, what might this action mean for a society that values discussion, debate, and disagreement as paths to knowledge? Could silencing data be contagious for other forms of silencing? On what basis do we, who have been trained to seek new knowledge in reliable and valid ways, defend the choice of withholding, withdrawing, or otherwise silencing data?

Data by themselves do not have full voice. Cancer-related data will always need to be matched with responsible, sage researchers, clinicians, patients and others if data and their interpretations are to illuminate both immediate and future care and treatment possibilities. Let us risk personal and professional discomfort secondary to data, and with wisdom, help our cancer data to speak.

My very best to you,

- Pamela S. Hinds, PhD, RN, FAAN

Editor-in-Chief, Cancer Nursing

An International Journal for Cancer Care

Department of Nursing Research and Quality Outcomes

Children's National Medical Center

School of Medicine and Health Sciences

Department of Pediatrics

The George Washington University

Washington, DC

Reference