genomics & society

Health Literacy and Communicating Genomic Information

Posted by Myra I. Roche on November 1, 2011

In a previous post,I focused on how parents’ health literacy impacts their ability to understand and search for genetic information.  This post turns the tables and examines the influence of health literacy on the way health care providers communicate risks.  Does health literacy, or more specifically, numeracy, defined as the ability to use and understand numbers, affect the way physicians communicate risk information?  Using maternal serum screening for Down syndrome as the example, a recent article in Genetics in Medicine looked at how the perceived ease with numbers influenced the communication style of obstetricians and gynecologists who routinely offer this screening.  The article “Physicians’ communication of Down syndrome screening test results: The influence of physician numeracy” can be found here.   The questions asked were: how did these physicians present risks information and did their communication style correlate with their perceived (and actual) numeracy?

Pitfalls in interpreting numbers.
Risks can be presented quantitatively (in numbers) or qualitatively (in words) or both.   The way in which numbers are presented can affect patients’ understanding.  For example, risks presented as rates (e.g., 2.7 in 10,000) are understood better than those presented as proportions (e.g., 1 in 3703).  This difference is caused by a phenomenon called “denominator neglect” in which people judge the size of a number only by the size of the numerator while “neglecting” the size of the denominator.  They judged the number 1,286 out of 10,000 to be larger than 24.14 out of 100 because the former has a bigger numerator.  Of course, this example exacerbates the chance of misinterpretation because the numbers are not round and they have two different denominators making interpretation even more difficult.  For a fascinating summary of the best practices of portraying numerical risks, see this article by Lipkis, here.

But is using words any better?
Relying on qualitative terms, though, has its own hazards.  In one study, non-pregnant women were randomly assigned into fictitious high risk (12.5/1000) or low risk (2/1000) categories for a fetal chromosome abnormality.  After “taking” a “blood test” to clarify this risk, all participants were given the same revised result of 5 in 1000, regardless of their previously quoted risk.  Participants who were presented with this revised risk numerically (5 in 1000) reacted similarly, implying that they interpreted the risk the same.  But if this same result was instead labeled qualitatively either as “positive” or “abnormal”, participants reacted differently.  Those who were told their result was “abnormal” perceived themselves to be at higher risk, reported more worry, and were more interested in amniocentesis than those who received the same result labeled “negative” or “normal”.

How numeracy affects communication style.
Which methods did the Ob-Gyns in the study use to communicate the screening results?  Thirty-five percent used numerical values while 23% percent used words like “high-risk” and “low risk”.  Nearly one-quarter used the same loaded terms described above: “normal or abnormal (15%) and “positive or negative” (9%).  Use of numerical values was correlated with younger physician age but not with gender.  When patients were asked which format they preferred, they overwhelmingly chose frequency (72%) with many fewer choosing percentages (10%).  Those physicians who perceived themselves as having high numeracy were more likely to use numbers to explain the screening results but this tendency was related only to their perception.  Neither their actual ability nor their stated preference, both of which were independently measured, showed any correlation.

Learning how to communicate results.
What happens when a physician who perceives him or herself as having low numeracy communicates a screening result to a patient? They are likely to present the risk information in qualitative terms increasing the chance that patients will inflate their perceived risk.  Perhaps because the mean age in the study was 50, the overwhelming majority of the physicians in this study avoided the qualitative terms “positive” and “abnormal”.  But this may have also resulted from their past clinical experience.  A study of pediatric residents who had received no formal training in communicating results were asked to provide newborn screening results indicating carrier status for cystic fibrosis or sickle cell anemia to standardized patients posing as parents of a new baby.  An analysis of the resulting transcripts concluded that 41 of 59 contained “misleading” statements such as the screening test was “positive” and the mean time between the first “bad news” and the first “good news” statements was 28 statements.  On average, 1/5th of the session elapsed before the accurate message of “your baby is healthy” was communicated.  Although cause for concern, this may indicate that specific training in how to communicate screening results could improve a physician’s ability to focus on the most important messages, especially in those critical first few minutes.

Communicating genomic results.
While screening tests typically generate quantitative risks,  genomic tests results cannot always be translated into numerical risks.  The clinical use of chromosomal arrays is a good example in which a typical “result” involves the deletion or duplication of several contiguous genes, each which may have vital functions at the cellular level but which, individually and collectively, have unknown clinical significance at the organism level.  Assaying several genes (or, as is the case with sequencing, several thousands of genes) at once increases both the complexity and the ambiguity of the results.  For most genetic changes, there are simply insufficient data to quantify the risks of disease.  But even if they were available, would it really be in the patient’s best interest to be presented with scores of individual risk numbers for every possible condition that has an underlying genetic cause?   If not, it leaves us back at trying to find, or maybe invent, the right words so patients can “obtain, process, understand, and use” genomic information to make competent decisions.


2 Responses to “Health Literacy and Communicating Genomic Information”

  1. draganalassiter said

    I think there is a tendency to frame education in a caregiver-patient relationship as one way only. Most often, discussions about literacy presume that information is value neutral and the question is how to communicate such neutral information to patients or research participants. It is refreshing to see an article and this great discussion that destabilizes the neutrality of information by showing how caregivers’ experience (and, perhaps, the culture that differentially values numbers?) may shape the way they and their patients perceive risk and, ultimately, alter the course of their care.

  2. Karey said

    In Thedore Porter’s book Trust in Numbers: The Pursuit of Objectivity in Science and Public Life, he argues that the scientific community has become more reliant on numbers to both standardize results and communicate them to the public. People tend to trust and understand numbers or statistics more easily than words. With that being said, numbers can very well aid in discussion with patients about genetic risks and results, but a more careful look is needed on how to effectively communicate this information so that it is understood by the patient. The goal of effectively communicating genomic information is as you mentioned is “so patients can “obtain, process, understand, and use” genomic information to make competent decisions.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s