Caverly TJ, Prochazka AV, Cook-Shimanek M, Pawlak M, Stickrath C. Weighing the potential harms of computed tomography: patient survey. JAMA Intern Med. Published online March 4, 2013. doi:10.1001/jamainternmed.2013.2918
eTable. Demographic characteristics for patients in the outpatient computed tomography (CT) scan waiting room (n=271)
Caverly TJ, Prochazka AV, Cook-Shimanek M, Pawlak M, Stickrath C. Weighing the Potential Harms of Computed Tomography: Patient Survey. JAMA Intern Med. 2013;173(7):588-590. doi:10.1001/jamainternmed.2013.2918
Author Affiliations: Division of General Internal Medicine, Department of Medicine (Drs Caverly, Prochazka, and Stickrath) and Department of Preventive Medicine (Drs Cook-Shimanek and Pawlak), University of Colorado Anschutz Medical Campus, Aurora; and Denver Veterans Affairs Medical Center, Denver, Colorado (Drs Caverly, Prochazka, and Stickrath).
Up to 1 in 3 imaging tests in the United States are ordered in situations when the expected benefits do not sufficiently exceed the risks.1 Unfortunately, studies suggest that clinicians are not well informed about the risks of medical imaging.2,3 Efforts to improve the risk communication skills of clinicians are a strategy to reduce imaging overuse (imaging that is inappropriate or discretionary). When patients are fully informed, they often opt for fewer tests and less aggressive care.4 However, the impact of current risk communication practices on patient knowledge is not well understood.
We undertook a survey to understand the frequency of risk communication discussions prior to undergoing computed tomography (CT) and how these discussions informed patients of potential scanning harms.
We gave a self-administered questionnaire to 286 consecutive patients undergoing outpatient CT at the Denver Veterans Affairs Medical Center (VAMC) from November through December 2011. Respondents answered questions within the following 4 domains: (1) demographics, (2) presence of risk communication, (3) preference for more information, and (4) knowledge of potential harms.
We assessed knowledge with a free-response question asking about general harms of CT. Respondents also gave a subjective ranking of the radiation exposure associated with chest radiography (CXR), magnetic resonance imaging (MRI), CT, and living 1 year in Colorado.
We defined 2 groups to help assess basic knowledge as an outcome variable: (1) those who knew that a CT scan is associated with a higher exposure to radiation than CXR and (2) those who did not know that a CT scan is associated with a higher exposure to radiation than CXR. To understand the effect that risk communication had on this basic knowledge, 2 additional groups were defined: (1) those who reported having a discussion of both the risks and benefits of undergoing the CT scan and (2) those who did not report having a discussion of both the risks and benefits of undergoing the CT scan. Analyses were performed using Epi Info statistical software (version 7; Centers for Disease Control and Prevention). Associations between groups were analyzed using the Pearson χ2 test, with P < .05 considered statistically significant.
Of 286 invited individuals, 271 completed the survey (94.8% response rate). Most of the respondents were older than 50 years (86%) and male (92%). Twenty-seven percent had a high school education or less, and 92% had undergone at least 1 previous scan, with 38% reporting more than 5 previous scans (see eTablefor patient characteristics).
A majority of respondents (62%) believed that the final decision to undergo CT was mainly the physician’s. A minority (35%) said they discussed the potential risks of the test with their health care provider. Only 17% (n = 46) reported all of the following prior to undergoing the CT scan: having a shared final decision, discussing the potential benefits, and discussing the potential risks with their health care provider.
Responses to the knowledge questions are given in the Table. Thirty-seven percent knew that CT was associated with more radiation than CXR. Those who reported discussing both risks and benefits with their health care provider were no more likely to know that CT was associated with more radiation than CXR than were respondents not reporting a risk-benefit discussion. (P = .60)—a result that did not change with adjustment for age, education, sex, number of previous scans, or ordering clinician in multivariate analysis.
Our study indicates that most decisions to undergo outpatient CT are made by physicians and risk communication is infrequent. The risk communication that took place had limited impact: respondents who recalled discussing the benefits and risks of imaging did not have better knowledge.
A minority of patients could state potential harms from the test. No respondents mentioned downstream consequences of incidental findings as a potential harm. Because the risk of developing cancer from the radiation associated with a single CT scan has led to calls for mandatory informed consent,5 we asked specific questions about radiation. Less than half of the respondents knew that CT was associated with more radiation exposure than CXR.
Our findings are limited by the fact that this was a single center study in an older, male VAMC population. Findings in this population may not generalize to other populations. Nonetheless, our findings are in line with findings from several previous patient surveys in non-VAMC populations demonstrating poor patient knowledge of radiation risks and limited shared decision making.3,6- 8
We believe it is problematic when the potential harms of CT are not adequately conveyed. Ignoring downsides can lead to imbalanced decision making in favor of overuse. Organizations seeking to reduce scanning overuse and also address underuse will likely consider order entry decision support9 among other methods. Correcting the lack of knowledge and the lack of communication about the potential harms of imaging tests—thereby enhancing shared decision making—should be part of any attempt to curb imaging overuse. It is time to begin empirically testing risk communication methods and translate these methods into routine clinical practice.
Published Online: March 4, 2013. doi:10.1001/jamainternmed.2013.2918
Correspondence: Dr Caverly, Division of General Internal Medicine, Department of Medicine, School of Medicine, University of Colorado, University Physicians Inc, COR/COHO, 13199 E Montview Blvd, Third Floor, Ste 300 Aurora, CO 80042-1719 (Tanner.Caverly@ucdenver.edu).
Author Contributions:Study concept and design: Caverly, Prochazka, Cook-Shimanek, Pawlak, and Stickrath. Acquisition of data: Caverly, Cook-Shimanek, and Pawlak. Analysis and interpretation of data: Caverly, Prochazka, and Stickrath. Drafting of the manuscript: Caverly. Critical revision of the manuscript for important intellectual content: Caverly, Prochazka, Cook-Shimanek, Pawlak, and Stickrath. Statistical analysis: Caverly and Prochazka. Obtained funding: Caverly, Cook-Shimanek, and Pawlak. Administrative, technical, and material support: Caverly and Stickrath. Study supervision: Caverly, Prochazka, and Stickrath.
Conflict of Interest Disclosures: None reported
Funding/Support: Dr Caverly is supported by institutional grant T32HP1006 from the National Research Service Award (NRSA). Drs Cook-Shimanek and Pawlak were supported in part by Health Resources and Services Administration (HRSA) grant D33HP02610 to the University of Colorado Preventive Medicine Residency Program. Dr Cook-Shimanek was supported in part by a Physician Training Award in Preventive Medicine (grant PTAPM-96-156-16) from the American Cancer Society (ACS).
Disclaimer: The contents are solely the responsibility of the authors and do not necessarily represent the official views of HRSA or ACS.
Previous Presentation: The results of this analysis were presented in poster form at the Society for General Internal Medicine 2012 meeting; May 10, 2012; Orlando, Florida.