Advertisement

Communication of false-positive tests—can it be improved?

      To the Editor:
      The recent article by Pinckney and colleagues (
      • Pinckney R.G.
      • Geller B.M.
      • Burman M.
      • Littenberg B.
      Effect of false-positive mammograms on return for subsequent screening mammograms.
      ) highlights the perception of risks in patients with a false-positive test. Despite confirmation of the absence of breast cancer, and adjustment for additional factors such as the use of hormone replacement therapy, education, and prior false-positive mammograms, patients with false-positive mammograms returned for subsequent screening mammography. There could be several reasons for patients’ tolerance of risk. Patients with false-positive tests undergo subsequent evaluations by radiologists, physicians, and surgeons who, in turn, have their own way of communicating the results to the patients. A perusal of the literature reveals that medical students and physicians find information involving risks presented as probabilities less clear than presentation of data in the “natural frequency format” (
      • Hoffrage U.
      • Lindsey S.
      • Hertwig R.
      • Gigerenzer G.
      Communicating statistical information.
      ,
      • Hoffrage U.
      • Gigerenzer G.
      Using natural frequencies to improve diagnostic inferences.
      ). Also, the perception of risks that are not under one’s voluntary control is perceived to be worse than when the person takes a risk willingly (e.g., smoking, snowmobiling) (
      • Slovic P.
      Perception of risk.
      ).
      Medical risk presented as sensitivity and specificity can often be confusing to physicians. The overall sensitivity of a first screening mammogram is about 90% and specificity is about 93% (
      • Kerlikowske K.
      • Grady D.
      • Barclay J.
      • Sickles E.A.
      • Ernster V.
      Likelihood ratios for modern screening mammography risk of breast cancer based on age and mammographic interpretation.
      ). However, the prevalence of breast cancer in women older than 40 years with a family history of breast cancer is 0.009% (
      • Kerlikowske K.
      • Grady D.
      • Barclay J.
      • Sickles E.A.
      • Eaton A.
      • Ernster V.
      Positive predictive value of screening mammography by age and family history of breast cancer.
      ). Physician interpretation of the positive predictive value of similar data has often been found to be inaccurate (
      • Hoffrage U.
      • Lindsey S.
      • Hertwig R.
      • Gigerenzer G.
      Communicating statistical information.
      ). It is possible that a patient with a false-positive result may receive varying and sometimes conflicting information from their physician. Presentation of such information in the “natural frequency format” could be stated as follows: “9 out of every 1000 women under 40 years with a family history of breast cancer have breast cancer. Of these 9 women, 8 will have a positive mammogram. Of the remaining 991 women without breast cancer, 70 women will have a positive mammogram. Therefore, in a study of women in this age group, only 8 of 78 women (9%) with a positive test will actually have the disease.” It remains likely that women who have had a false-positive mammogram could receive their medical information in a format that they can understand best. However, there are no studies to date demonstrating that information presented in the “natural frequency format” would reduce the anxiety levels and clinic visits of patients with false-positive mammograms.

      References

        • Pinckney R.G.
        • Geller B.M.
        • Burman M.
        • Littenberg B.
        Effect of false-positive mammograms on return for subsequent screening mammograms.
        Am J Med. 2003; 114: 120-125
        • Hoffrage U.
        • Lindsey S.
        • Hertwig R.
        • Gigerenzer G.
        Communicating statistical information.
        Science. 2000; 290: 2261-2262
        • Hoffrage U.
        • Gigerenzer G.
        Using natural frequencies to improve diagnostic inferences.
        Acad Med. 1998; 73: 538-540
        • Slovic P.
        Perception of risk.
        Science. 1987; 236: 280-285
        • Kerlikowske K.
        • Grady D.
        • Barclay J.
        • Sickles E.A.
        • Ernster V.
        Likelihood ratios for modern screening mammography.
        JAMA. 1996; 276: 39-43
        • Kerlikowske K.
        • Grady D.
        • Barclay J.
        • Sickles E.A.
        • Eaton A.
        • Ernster V.
        Positive predictive value of screening mammography by age and family history of breast cancer.
        JAMA. 1993; 270: 2444-2450