Antibiotic Overuse May Depend on How We Talk to Our Doctors

Most acute upper respiratory infections (URIs) in adults, like bronchitis, are caused by viruses, yet antibiotics that are ineffective in treating viral infections are commonly prescribed for them. The U.S. Centers for Disease Control and Prevention (CDC) estimates that about one-third of antibiotic prescriptions are inappropriate.  Surely, physicians and other highly trained prescribers must know that antibiotics are only effective for bacterial illnesses, so why year after year are there constant reports of so much inappropriate prescribing?

Critica Chief Medical Officer David Scales is both an internist and sociologist and has been following the work of fellow sociologist Tanya Stivers, a professor at UCLA who has studied the conversations people have with their doctors, including around the issue of antibiotic prescribing. Her findings highlight the ways in which these conversations can tip the balance toward inappropriate prescribing and suggest strategies clinicians can use to avoid the inappropriate prescribing pitfall.

Stivers points out that research has shown that when physicians are asked why they prescribe antibiotics for viral illnesses they often cite  “patient demand for antibiotics.” It isn’t, she explains, that patients usually explicitly demand an antibiotic but rather employ more subtle techniques, like offering a bacterial cause as the likely cause of the illness or questioning either a viral diagnosis or a treatment plan that does not include an antibiotic. This, she says, puts pressure on the prescriber to offer an antibiotic.

Study Shows How Patients Pressure Clinicians

In a  study in which Stivers and her colleagues listened to recordings from hundreds of adult and pediatric visits, she found that “In 37% of adult cases, clinicians prescribed antibiotics in the absence of clinical signs” that would indicate an antibiotic was the appropriate choice of treatment. The rate was much lower for pediatric visits (16%).

Stivers then listened in detail to these conversations and found evidence to support the notion that patients subtly pressure physicians to prescribe antibiotics, even when they are not indicated. This pressure fell into two main categories:

1. Pre-diagnostic priming: describing the problem in ways that are relevant to a bacterial diagnosis. An example is to describe a sore throat as a “strep throat” or to complain of “tons of green mucus.”
2. Post-diagnostic negotiation: any response that “questions or contests a clinician’s recommendation for nonantibiotic treatment.” This might be in the form of “that never helps, and I always wind up taking an antibiotic anyway.”

Stivers’ research indicates that these types of pressure are significantly associated with inappropriate antibiotic treatment and she urges clinicians to be aware of such interactions and prepared to deal with them. She offers a variety of strategies the clinician can use to ward off antibiotic pressure without offending the patient that include foreshadowing the viral diagnosis (e.g., “that can be a symptom of a virus”), affirmative counseling (i.e., saying “here’s what we can do for you” instead of “here’s what we cannot do”), and appropriate use of persuasion that includes careful and clear explanations of why a viral diagnosis is most likely and antibiotics not the right approach to relief. Stivers concludes that “patient communication with physicians matters for treatment outcomes.”

Collaborative Care and Unintended Consequences

In a paper in the  Annual Review of Linguistics, Stivers and Alexander Tate of the Department of Medicine at the University of Chicago, elaborated on the importance of better understanding the nature of physician-patient communication to understand things like inappropriate antibiotic prescribing. They argue that a “paternalistic communication style prevailed in clinical practice across the 1950s and 1960s” but that this model was “heavily criticized in the 1970s and 1980s.” The result has been an increase in the patient’s role in clinical decision-making, often characterized as “collaborative care.” They write that “In this care model, the decision-making process is conceptualized as collaborative: Physician and patient are involved in treatment decisions through two-way exchanges of information, preferences, and investment in the treatment decision.”

That is certainly in many ways a positive development. With the increased ease of obtaining medical information on the internet, people are now better informed and more able to act as advocates for themselves and their families. No one should be nostalgic about the “paternalistic” model.”

 Nevertheless, Stivers and Tate point to unintended consequences of the collaborative model. Patients now view direct-to-consumer advertising about medications, read up about illnesses on the internet, and discuss symptoms and treatments across social media platforms. This leads to the kind of pressure communications Stivers talks about when the issue is antibiotic prescribing across a broad range of illnesses and conditions. The problem, Stivers and Tate say, is not mainly that patients frequently demand specific tests or treatments, although they sometimes do so. Rather, it is the subtle ways that patients pressure clinicians to order these tests and treatments that are sometimes unnecessary or inappropriate, a phenomenon they call “resistance.” “Resistance,” they write, “generally entails disagreeing with or questioning physicians’ diagnoses or treatments.”

Now, even as we applaud increased patient involvement in treatment decisions, advocating for things that are not medically indicated and may even be dangerous is certainly not what collaborative care should be about. As Stivers and Tate note, “such behaviors are not likely the sort of patient engagement originally envisioned by proponents of patient-centered care policy.”

Better Communication All Around is Needed

Antibiotic prescribing is a good case in point here because inappropriate prescribing can lead to serious adverse consequences. Like all medicines, antibiotics do have adverse side effects; chronic, inappropriate antibiotic use can, for example, lead to serious gastroenterological infections like those caused by the bacteria C. difficile (also known as C. diff). On a broader basis, over-prescription of antibiotics leads to antibiotic resistance, the phenomenon by which bacteria that were previously sensitive to an antibiotic become resistant to it. Antibiotic resistance is a major source of morbidity and mortality worldwide. It is crucial that we reserve the use of antibiotics for clinical situations in which they are needed and known to be effective. Clinicians generally know this, of course, and therefore need to be resilient against pressure to prescribe antibiotics inappropriately.

How can they accomplish this? Stivers and Tate suggest the following based on empirical studies: “Physicians can preempt resistance by offering a running commentary of their findings—particularly unproblematic findings. When clinicians use this ‘online commentary’ during visits in which they are disinclined to offer antibiotics for upper respiratory tract symptoms, parents are less likely to resist the nonantibiotic treatment recommendations.”

Thus, if we are to improve on the collaborative care model, including when it comes to safe antibiotic prescribing, we need to pay much more attention to how doctors and patients conduct routine conversations about things like upper respiratory symptoms. As Stivers has done, we must analyze the ways in which patients subtly pressure clinicians to order tests or prescribe medications that aren’t needed and at the same time improve the way physicians communicate findings and diagnostic logic to patients. Saying nothing to the patient during the physical examination, for instance, is contrary to what Stivers believes to be a best practice. Rather, the examiner should communicate the findings as they proceed: tell the patient what the blood pressure is, that the lungs sound clear, that the reflexes are all normal. Then explain how the diagnosis was formulated and put the rationale for a treatment plan in affirmative terms.

Sociologists clearly have a lot to teach physicians about communication styles. We need more of this kind of analysis if we are to improve what goes on in the doctor’s office.