The principle of “first, do no harm” comes from the oath of Hippocrates: “I will follow that method of treatment which, according to my ability and judgment, I consider for the benefit of my patients, and abstain from whatever is deleterious and mischievous.”
An assertion by Miller1 is consistent with this oath: “The ethical requirements for the introduction of screening are stringent” because of the “risks associated with screening, which extend beyond the risks of the test itself, to those associated with invitations for screening, false reassurance for false negatives, the diagnostic process and the problem of false positives, and the overtreatment of those with borderline abnormalities.” Tuohey,2 however, argues that avoidance of harm is not enough and that the Hippocratic oath also includes the principle of benefit. The United States Preventive Services Task Force (USPSTF)3 acknowledges this balance in its statement about preventive counseling:
Given the safety and generally low cost of advising patients about health-related behaviors, the USPSTF recommends routinely addressing some health behaviors even when the long-term effectiveness of such counseling has not yet been definitively proven. In their updated recommendations, however, the USPSTF explicitly distinguished between recommendations based on good evidence of the effectiveness of counseling per se (eg, smoking cessation) and recommendations made primarily on the basis of the strong link between behavior and disease (eg, sexually transmitted disease prevention).
Is prevention counseling safe?
A key assumption supporting these recommendations is the safety of counseling for primary prevention. Unfortunately, the work of Bowman and colleagues4 in this issue of the Journal challenges the benign nature of recommending behavioral change to patients. In that study, patients whose physicians recommended a behavioral change had lower self-esteem, worse mental and social health, and more dysfunction from anxiety, anxiety/depression, and depression 3 months after the visit than patients whose physicians had not recommended this type of change.
To its credit, the USPSTF explicitly considers the adverse effects of screening tests in its recommendations about secondary prevention. The paper by Mold and coworkers5 in this issue of the Journal adds to the literature cited in the USPSTF’s second report. These authors found an apparent adverse effect of labeling, with hypertensive patients estimating recovery time from an upper respiratory tract infection that was nearly twice as long as that estimated by patients who did not have hypertension.
These 2 studies bother me. I am an unabashed prevention enthusiast. I teach second-year medical students the USPSTF methodology and recommendations, and I lead a monthly clerkship seminar on helping people change problem behaviors. I hound residents about prevention at every turn: a session on the checkup visit for each stage in our lifecycle core curriculum, waxing on about prevention during precepting in the clinic, organizing the checkup visit by major categories of death and disability (cardiovascular disease, cancer, and so forth) instead of the traditional history and physical examination (history of present illness, past medical history, family history, social history), and a customized prevention template in our electronic medical record. For trauma prevention, I have even asked patients about handguns, despite the survey results from Shaughnessy and colleagues that state most patients believe gun safety should not be discussed during an office visit.6 But these 2 papers have led me to question whether my enthusiasm is misguided and may be a disservice to my patients. I wonder if the potential for an adverse effect of counseling or labeling is a widespread phenomenon. Is it a long-term problem? Can we do anything to minimize it?
What the literature shows
A quick search of the literature shows that the concern about possible adverse effects of screening and risk assessment cuts across all areas of prevention. In metabolic disease prevention, adults screened for carrier status in cystic fibrosis7-9 were anxious while awaiting the test results, but the anxiety resolved by 3 months and had no effect on perception of health. Individuals screened for risk of Huntington’s disease10-17 experienced anxiety and depression in the first 2 months after the test, which generally decreased in the following year, especially after a good test result. Those who had pessimistic risk perceptions or pretest depression responded more poorly to the information. Pregnant women having serum screening for congenital anomalies18 reported no adverse psychological effects at 24 weeks’ gestation. Those having false-positive screening results for gestational diabetes mellitus19 rated their health lower at 32 weeks’ gestation than those with negative results but did not differ on anxiety, depression, or concern levels about the health of their newborn.
Men screened for prostate cancer20 had high serum cortisol levels at the screening examination, that were back to normal in 2 weeks. The highest levels were from those getting biopsies, but they returned to normal when informed of the results, regardless of whether those results were abnormal. The authors suggested minimizing the interval between a test and informing the patient of the results to decrease the duration of stress. Breast cancer screening21-23 leads to short-term increases in anxiety, distress, and intrusive thoughts. In breast cancer susceptibility genetic testing,24-27 there was more distress if the person was tested first or if previously tested siblings were not carriers. In general, distress declined after testing, but women anticipated a negative psychologic impact of positive test results involving anxiety, depression, and impaired quality of life, while others still worried even with a negative test result. A false-positive screening result for ovarian cancer28 was associated with short-term distress but not severe or persistent distress. For cervical cancer,29 3 months after a positive Pap- anicolaou test result women had significant elevation in worries about cancer and impairments in mood, daily activities, sexual interest, and sleep patterns. These adverse effects were most pronounced in those who did not comply with colposcopy, for whom uncertainty was maintained by lack of compliance. Testing children for familial polyposis30 did not lead to adverse effects, as child and parent depression, anxiety, and behavioral problem scores remained normal 3 months after testing for children with both mutation-positive and mutation-negative results. Patients screened for colorectal cancer31 were distressed by an invitation letter, a positive test result, and delays in the process of screening (eg, wait of 10 days to colonoscopy), but reported that it was worthwhile to have had the test. In general cancer screening programs,32-34 anxiety, distress, and depression symptoms can be reduced by a brief orientation program at the clinic visit. It has been suggested that a moderate amount of anxiety about results may even be a motivating factor in getting subsequent checkups, but severe distress can interfere with adherence behavior.