Skip Navigation LinksHome > May 2010 - Volume 85 - Issue 5 > Reporting-bias in Surveys of Sensitive Personal Information
Academic Medicine:
doi: 10.1097/ACM.0b013e3181d6c6ea
Letters to the Editor

Reporting-bias in Surveys of Sensitive Personal Information

Larson, Paul R. MD

Free Access
Article Outline
Collapse Box

Author Information

First-year fellow, University of Pittsburgh Family Medicine Faculty Development Fellowship Program, UPMC St. Margaret, Pittsburgh, Pennsylvania; larsonpr@upmc.edu.

Back to Top | Article Outline

To the Editor:

The study by Dunn et al1 sheds light on an important aspect of residents' well-being. However, while great care was given to an expansive discussion of the results' implications, the use of drugs and alcohol among the study population may have been a bigger problem than the authors were able to identify. This may represent a reporting bias that would undermine the validity of the survey findings.

Let me explain. In reviewing the residents' self-reported personal health concerns, the authors reveal that “almost no problems with alcohol, prescription drugs, or other drugs” were reported. However, among residents' concerns about current specific health problems that might jeopardize their training status if directors or supervisors were to learn of them, the three highest-rated concerns (shown in Figure 1) were problems with alcohol, prescriptions, or other drugs.1 Why should the residents have been concerned about disclosure of problems with alcohol, prescription drugs, or other drugs if they said that almost no problems with these substances exist? There are at least two possible explanations for this inconsistency.

First, the residents at the University of New Mexico (UNM) in 2001 may not have considered their alcohol consumption to be a personal health problem. This may represent psychological denial, as a higher rate of substance use would be expected than they reported, based on substance use of their peers in the general population, where 21% of 26- to 34-year-olds engaged in regular binge drinking behavior in 2001.2,3

A second possibility is that residents training at UNM in 2001 who responded to the survey were not completely forthcoming in presenting their self-reported personal health data. There may have been fear that the survey results would be reviewed by supervisors. Their fear would be compounded by the fact that all study participants were from the same school of medicine. This fear could also lead to a low survey response rate if those residents with more difficulties with alcohol or drug use were less likely to respond.

I suspect that the second alternative was more likely to have influenced the residents' responses. That is because fear of disclosure, even in a confidential survey, will often affect the accuracy of residents' responses to a survey. This can lead to systematic underreporting of sensitive personal information and undermine the validity of the survey findings.

Paul R. Larson, MD

First-year fellow, University of Pittsburgh Family Medicine Faculty Development Fellowship Program, UPMC St. Margaret, Pittsburgh, Pennsylvania; larsonpr@upmc.edu.

Back to Top | Article Outline

References

1 Dunn LB, Green Hammond KA, Roberts LW. Delaying care, avoiding stigma: Residents' attitudes toward obtaining personal health care. Acad Med. 2009;84:242–250.

2 Naimi TS, Brewer RD, Mokdad A, Denny C, Serdula MK, Marks JS. Binge drinking among U.S. adults. JAMA. 2003;289:70–75.

3 Roberts LW, Warner TD, Lyketsos C, Frank E, Ganzini L, Carter D. Perceptions of academic vulnerability associated with personal illness: A study of 1,027 students at nine medical schools. Compr Psychiatry. 2001;42:1–15.

© 2010 Association of American Medical Colleges

Login

Article Tools

Share