Newswise — A study conducted at LMU demonstrates that patients' ability to accurately evaluate risks is influenced by the way physicians communicate statistical information to them.
Understanding medical results can be challenging, and the accompanying risk statements can be even more difficult to comprehend. Effectively communicating statistical information to patients is a formidable undertaking.
In the journal PLOS ONE, a group of researchers consisting of medical professionals, medical education experts, and mathematics education experts from LMU have published a study that examines strategies for enhancing the effectiveness of communication between doctors and patients regarding real risks.
Indeed, comprehending the true significance of certain numerical values can be challenging. Karin Binder, a mathematics educationalist and one of the study's authors, explains, "Even doctors face difficulties in accurately determining the appropriate predictive value. Consequently, if the data is intricate for doctors to interpret, effectively conveying the information to patients in a manner they can understand becomes even more demanding."
Let's consider the following scenario as an illustration: A patient receives an abnormal sonographic finding on their thyroid. Does this automatically imply that they have thyroid cancer? Not necessarily, as there exists a probability that the examination result may be positive even if the patient doesn't have thyroid cancer.
In order to convey the statistical scenario to patients following a positive test result, two approaches are available. One approach involves utilizing creative thinking, while the other is significantly more accessible and easier for patients to comprehend, as demonstrated by the researchers.
Bayesian vs. diagnostic information
The conventional Bayesian approach involves starting with the number of patients who truly have the disease. The doctor begins by explaining the overall frequency of the disease, such as "out of 1,000 patients, 50 are diagnosed with thyroid cancer." Then the doctor presents two key pieces of information: a) the number of patients with thyroid cancer who receive a positive test result (20 out of 50), and b) the number of patients without thyroid cancer who still receive a positive test result (110 out of the remaining 950).
This information is typically within the knowledge or easy research reach of the doctor. The proportion of positive tests among individuals with the disease is commonly referred to as sensitivity, a term that gained familiarity during the Covid-19 pandemic, particularly in relation to quality assessment of rapid tests. However, there is a common confusion between sensitivity (positive tests as a proportion of people with the disease) and the proportion of people with the disease among positive tests. These two percentages can significantly vary depending on the specific situation.
What do the aforementioned numbers signify when it comes to an individual with a positive test result? How many of those who test positive actually have the disease? If you find yourself unsure of the answer, you are not alone. Without additional information, only 10% of participants were capable of determining the number of individuals with positive results who truly had the disease.
"The diagnostic" communication approach of conveying information follows a different path. Initially, the doctor explains the number of patients with positive test results, regardless of whether they truly have the disease or not. In this case, it would be 130 individuals with notable thyroid ultrasound results out of the total 1,000 people examined. Next, the doctor proceeds to explain the number of people with positive tests who actually have the disease (20 out of 130), as well as the number of people with negative test results who still have the disease (30 out of 870).
The pertinent information is presented directly here, eliminating the need for mental calculations. If my test result is positive, then the probability of actually having thyroid cancer is 20 out of 130. When communicated in this manner, 72% of the study participants were able to reach this conclusion, compared to only 10% with the Bayesian approach.
What is the best way of communicating statistical information?
"Furthermore, with Bayesian communication, participants took significantly more time to arrive at the correct result, if they managed to do so at all," explains Karin Binder. "In busy doctor's offices and hospitals, time is often a limited resource." Therefore, the team of authors strongly encourages doctors to employ diagnostic information communication more frequently in the future. This approach would help prevent confusion, misinterpretation, and erroneous decisions.
However, it would be even more advantageous to allocate the necessary time to provide patients with a comprehensive overview of the situation, incorporating both diagnostic and Bayesian information. Only through this approach can we elucidate the intriguing phenomenon where even a medical test with exceptional quality criteria can possess limited predictive capability under certain circumstances (such as routine screenings).
RSS