Study finds age, race, provider type tied to improper prescribing

New research conducted at a large healthcare network shows that patient and provider age are among the characteristics most significantly associated with inappropriate antibiotic prescribing for upper respiratory infections, with provider type and race also playing a role.

The study by researchers with Carolinas Healthcare System, published yesterday in Infection Control and Hospital Epidemiology, found that in both adults and children in outpatient settings, the risk of receiving inappropriate antibiotics for four common upper respiratory conditions grew with age, and that older providers were more likely to prescribe antibiotics than younger ones. The study also indicated that white patients were more likely to receive antibiotics than patients of other races, and that advanced practice practitioners, such as nurses and physician assistants, were more likely to prescribe them.

The study adds to a growing body of research on outpatient prescribing that indicates patients are too often receiving antibiotics for ailments that don't require them. It also suggests that when it comes to inappropriate prescribing in doctors' offices and urgent care clinics, many factors are at play, and the authors of the study say that means that stewardship interventions will have to be multifaceted and tailored to different settings.

"This isn't going to be a one-size-fits-all solution," Lisa Davidson, MD, an author of the study and the medical director for the Antimicrobial Support Network at Carolinas Healthcare System, said in an interview. "What you need in family medicine versus what you need in urgent care are actually going to be different types of interventions."

Nurses, PAs 15% more likely to prescribe

In the study, Davidson and her colleagues looked at antibiotic prescribing for four upper respiratory conditions that are most commonly caused by viruses and do not require antibiotics: acute upper respiratory infection, bronchitis, bronchiolitis, and non-suppurative otitis media (ear infection). Using electronic medical record data, they reviewed prescribing patterns for 448,990 outpatient visits across 246 practices and 898 providers in Carolinas Healthcare System from January 2014 through May 2016.

To evaluate the factors associated with antibiotic prescribing, the researchers looked at patient age, race, gender, comorbidities, and insurance status, along with the age of the provider and the type of practice (internal medicine, family medicine, urgent care, or pediatrics). Providers were divided into physicians and advanced practice practitioners. Factors associated with antibiotic prescribing for pediatric and adult patients were evaluated separately.

The overall prescribing rate for the four conditions was 407 prescriptions for 1,000 patient visits, with acute bronchitis being the most common indication for which antibiotics were prescribed (703 prescriptions per 1,000 visits). At the practice level, family medicine had the highest rate of prescribing, followed by urgent care, internal medicine, and pediatrics. The most frequently prescribed antibiotic was azithromycin (46.6%), followed by amoxicillin (18.1%) and amoxicillin-clavulanate (11.8%).

After the investigators controlled for confounding factors, the results for pediatric patients showed that patients 3 to 9 years of age were 25% more likely than those aged 2 years or younger (the reference group) to receive antibiotics for one of the four conditions, while patients 10 to 19 years had a 31% greater risk. In addition, providers aged 51 to 60 were more than four times more likely to provide an antibiotic than those under the age of 30.

Those patterns were also seen in adults, with patients aged 40 to 64 4% more likely to receive an antibiotic than those aged 20 to 39, and providers 51 to 60 years old 92% more likely to prescribe an antibiotic than physicians age 30 and younger. The results also revealed that nurses and physician assistants were 15% more likely to prescribe an antibiotic to an adult patient than physicians were.

The authors suggest the reason that older physicians are more likely to prescribe antibiotics for these conditions could be attributed to how views on antibiotics have changed in recent years. "The hypothesis is that antibiotic stewardship, particularly outpatient antibiotic stewardship, is really a more recent concept, so older providers might not have incorporated it into their practice as much," Davidson said.

"For a long time, there was not a recognition that prescribing antibiotics had any potential harm," co-author Melanie Spencer, PhD, executive director of Carolinas Healthcare System's Center for Outcomes Research and Evaluation, added. "People tend to practice as they've always practiced."

Patient pressure, other factors

But the results suggest that patient pressure may be playing a role in prescribing patterns as well. Previous research indicates that when patients believe they need an antibiotic for an ailment, that expectation can influence providers, even if it's clear that an antibiotic is unnecessary. Although the study did not delve into the underlying reason why older patients were more likely to receive an antibiotic, the authors hypothesize that working-age patients may be pressuring providers and advance practice practitioners for prescriptions.

"The most common person coming in [to urgent care] is the person who has work or has to take care of a loved one and comes in because they haven't been feeling well…and they really need to get back to work or whatever their responsibilities are," Davidson said. "That puts a lot of pressure on the clinician, because we're here to make people feel better."

This cultural expectation could explain overprescribing of azithromycin, which requires a much shorter course than amoxicillin. "When we talk with providers, they tell us people often come in and say 'just give me a Z-pack,'" Spencer said, referring to a form of the brand-name medication Zithromax. "They have enough experience that they know what they want, and they don't want anything else."

The results also indicate that there may be a racial component to antibiotic prescribing, though the reasons are not well understood. African-American and Asian children were 14% and 31% less likely, respectively, to receive an antibiotic than white children, while African-American and Asian adults were both 15% less likely to receive an antibiotic than white adults. In addition, medical practices in urban areas were 36% more likely to prescribe an antibiotic than those in rural areas, a finding the authors say has not been reflected in other studies.

Changing the culture

So how do you change both the patient and provider culture around antibiotic prescribing?

"This issue can't just be addressed when the patient's in the room with the doctor," Davidson said. "It has to address the expectations of the entire interaction."

This means educating patients from the moment they make an appointment all the way through the appointment, and having a consistent message about antibiotics the entire time. Davidson said Carolinas Healthcare System is using a variety of resources to convey this message, including a website, educational videos, and patient checklists. The system also provide clear guidelines explaining when patients should call back to discuss worsening symptoms.

It also means providing physicians and advance practice practitioners with data on their prescribing practices. "Everybody thinks that they do a good job, and everyone wants to do the right thing, but until you have the data in front of you and actually know what your prescribing practices are, that's very hard to do," Davidson said.

Ultimately, Davidson and Spencer hope that identifying patient and provider-level factors associated with inappropriate prescribing will help them create more effective stewardship interventions that acknowledge how different types of settings, providers, and patients can influence prescribing patterns.

"When you're trying to put together an antibiotic stewardship program, knowing about those differences helps you tailor what you're doing," Spencer said.

See also:

Jan 30 Infect Control Hosp Epidemiol study

Newsletter Sign-up

Get CIDRAP news and other free newsletters.

Sign up now»


Unrestricted financial support provided by

Bentson Foundation 3MAccelerate DiagnosticsGilead 
Grant support for ASP provided by


  Become an underwriter»