Engaging patients and their families in treatment decisions is integral to achieving patient-centered care; however, this is easier said than done. With the increasingly complex nature of cancer care and its associated risks, come technical and statistical jargon that all too often leave patients without a complete understanding of their options and ill equipped to make informed decisions.
MDs frequently rely on risk data to aid patients in treatment decisions, however, this data is particularly difficult for the average patient to understand. In this JNCI Commentary, a University of Michigan team reviewed the literature to identify evidence-based tactics for risk communication that help to facilitate patient understanding and informed decision-making. The authors limited their review to include strategies for communicating high-quality evidence-based statistics, as this information is most relevant to patients making “preference sensitive decisions” (i.e. care decisions involving trade-offs, such as those between quality of life and length of life, or between different quality of life measures).
The authors offer up ten methods for improved communication of risk data, including using plain language with patients, explaining how treatment changes one’s risk from baseline risk, summarizing risk and benefits, and limiting information in order to prevent cognitive overload. Below I’ve pulled out three particularly interesting strategies for risk communication discussed in the review.
Present data using absolute risks and avoid relative risks and number needed to treat (NNT)
Physicians present risk data in a variety of ways, however, studies suggest these strategies are not created equal. For example, in order to communicate treatment risks of preventative tamoxifen therapy, physicians may present relative risk reduction (chance of breast cancer reduced 50%), absolute risk reduction (5-year risk of breast cancer reduced from 4% to 2%), or NNT (the number of patients who would need tamoxifen therapy in order to prevent one case of invasive breast cancer).
One study concluded that NNT was the most difficult risk communication format for patients to understand. Additional research suggests that the perceived reduction in risk created by the presentation of relative risk may lead patients to overestimate the efficacy of a given therapy.
Interestingly, the patient is not the only one prone to misinterpreting risk statistics. The authors cite studies in which medical students and practicing physicians were more likely to prescribe treatment when presented with relative risk compared to absolute risk. As such, the authors recommend communicating risk in terms of absolute risk to avoid misinterpretation and bias.
Present risk data as a frequency of given negative outcome, illustrate using pictographs
In a study of patients’ perceptions of risk data, patients perceived a lower risk of side effects after viewing percentage data (“10% of patients get a bad blistering rash”) as opposed to frequency data (“10 out of 100 patients get a bad blistering rash”). Low numerical literacy among patients and the abstract nature of percentages may make interpretation of percentage data difficult, and could explain the reduction in patients’ perceived risk. In order to prevent misinterpretation of risk data, the authors advocate for using frequency data over percentage data.
Since pictographs (see example below) communicate risk data by graphically illustrating the frequency of a given negative outcome, the authors suggest using pictographs to illustrate the risks of treatment.
Recognize that comparative risk information (e.g. what the average person’s risk is) is persuasive and not just informative
Cancer patients, like everyone else, are inclined to compare their individual risk to the risk of the general population. Importantly, women who were told they had an above average risk of breast cancer rated preventative treatments as more effective and indicated that they would be more likely to pursue care than those told they had a below average risk of breast cancer.
Since both cohorts viewed the same absolute risk data, the authors conclude that the perceived efficacy of preventative treatment and increased likelihood of engaging in treatment were a function of the patient’s belief that she was at greater risk for breast cancer than was the general population. In light of this, the authors caution against using comparative risk and encourage patients to make decisions based on their own valuation of risk and benefit.
As we transition to new models for cancer care delivery, patient involvement will be more important than ever and will depend on effective education and decision support. We discuss strategies for encouraging patient and family involvement in cancer care as well as strategies for enhanced patient education in our Oncology Roundtable 2011-12 national meeting series. Register now to secure your spot.