Posts

Can Effective Antibiotic Stewardship Co-Exist with Telemedicine?

Last week, JAMA Internal Medicine published a research letter from Kathryn A. Martinez, PHD, MPH, et al on the intersection of telemedicine, patient satisfaction and antibiotic prescribing practice. The study found that approximately 66% of patients who had video visits for respiratory tract infections (RTI) through a major telemedicine service received an antibiotic prescription. The researchers also found that patient satisfaction, based on a star system of 1 (lowest) to 5 (highest), strongly correlated with antibiotic prescriptions.

So, the question is, can telemedicine, antibiotic stewardship and patient satisfaction co-exist?

Telemedicine, Outpatient Care and Antibiotic Prescribing

A 66% antibiotic prescription rate seems high for RTIs, particularly when the majority of RTIs are viral rather than bacterial, as noted by the authors. But, that doesn’t tell the whole story. RTIs are frequently used for measuring antibiotic prescribing rates, in part because they are common, in part because the prescribing guidelines are incredibly clear, and in part because most RTIs are viral in nature. That means, there is actually a significant body of research around best practices related to antibiotic prescribing for RTIs.

A 2016 report by the Pew Charitable Trusts stated that acute respiratory conditions account for 44% of antibiotic prescriptions given in outpatient settings. These include: sinusitis (25%), otitis media (22%), pharyngitis (20%), viral upper respiratory infections (12%), bronchitis (12%), and pneumonia/asthma/allergies (9%). Additionally, a 2015 study published in JAMA compared antibiotic prescribing for RTIs in telemedicine and physician office visits and found them roughly equal (58% for telemedicine visits vs. 55% for physician office visits).

Achieving Antibiotic Stewardship through Virtual Care

With all the above, you’d be forgiven for throwing your hands in the air and giving up entirely on antibiotic stewardship, but, at least for online care, there is a better way. While we can’t be certain, our team suspects that there are a couple factors in play:

Antibiotics and Patient Satisfaction

The key takeaway from last week’s research letter is the correlation between antibiotic prescribing and patient satisfaction. The researchers found that 72.5% of patients who did not receive an antibiotic prescription for their RTI visit rated the encounter at 5 stars, compared with 90.9% of patients who did receive an antibiotic prescription. The correlation was further cemented by finding that physicians who prescribe antibiotics more frequently have higher overall patient satisfaction ratings than those that do not.

We also see the link between patient satisfaction and prescriptions through our post-visit patient surveys at Zipnosis, as well. Looking at patient survey responses, we see a dramatic difference in satisfaction between those who received a prescription (not necessarily an antibiotic) and those who did not. When asked how the virtual visit met their expectations, 89% of patients who received a prescription said it met or exceeded expectations. And, when asked to rate overall satisfaction on a scale of 1-7 (with 1 low and 7 high), 84% rated their satisfaction at 6 or 7 when they received a prescription compared with 36% among those who did not get a prescription.

Graph showing patient satisfaction relative to prescriptions given

*Data includes satisfaction survey responses for virtual patient visits from January 1 through September 30, 2018; n=5,993

As healthcare becomes an increasingly consumer-focused industry, health systems and providers are under pressure to achieve positive patient ratings, creating an imbalance in incentives, choosing to either satisfy patients or provide appropriate care. This is potentially even greater in direct-to-consumer telemedicine companies who rely extensively on patient ratings and reviews to attract new customers.

Of note, the JAMA study found that patient satisfaction remained higher for visits where patients received a prescription that was not an antibiotic. And in our experience, there is little difference in satisfaction between patients receiving antibiotic and non-antibiotic prescriptions. Offering patients non-antibiotic prescriptions may be an option for supporting patient satisfaction without overprescribing antibiotics.

Antibiotics and Visit Time

A concurrent article by the same researchers published in the Annals of Internal Medicine, identified a correlation between visit time and prescribing. Telemedicine visits resulting in an antibiotic prescription were approximately 20 seconds shorter than those where nothing was prescribed and more than a minute shorter than those resulting in a non-antibiotic prescription. The researchers concluded that because telemedicine visits were already short and providers employed by major telemedicine companies are often compensated by volume, implementing antibiotic stewardship practices that increase visit length may be challenging.

Antibiotics and Mode of Care

The one thing that has been left out of all the studies about antibiotic prescribing and telemedicine visits is the difference made by mode of care. Dr. Martinez and her co-authors acknowledge that their data may not represent all telemedicine visits, due to studying visits from a single direct-to-consumer telemedicine service company. Previous studies about antibiotic prescribing rates in telemedicine visits have also focused solely on direct-to-consumer video visits where the providers are employees or contractors of a telemedicine service company, rather than delivering care to their own (or their health system’s) patients. And mode of care can make a difference.

At Zipnosis, we monitor antibiotic prescribing adherence for visits on the asynchronous platform very closely. This adherence criteria is one of the metrics tracked and discussed with our Clinical Quality Advisory Council, as well as shared with individual customers for continuous quality improvement. We set aggressive antibiotic adherence targets for our most common protocols used via the asynchronous online adaptive interview and aggregate adherence data to evaluate performance. The table below shows both targets and actual adherence for three of our most common visit types

Table indicating antibiotic stewardship through prescribing guideline adherence for the Zipnosis platform

A third JAMA-published study, this one from 2016, found that approximately half of antibiotic prescriptions for acute RTIs diagnosed in outpatient settings were inappropriate based on prescribing guidelines. If rates of antibiotic prescribing are analogous, it stands to reason appropriateness would follow suit. In comparison, the percentage of visits on the Zipnosis platform that result in an antibiotic prescription but don’t match prescribing guidelines is less than 8%.

Zipnosis Asynchronous virtual care has two key benefits when it comes to antibiotic stewardship:

  1. The clinical algorithms, which incorporate organic, embedded clinical decision support, give providers the tools to easily and effectively make diagnosis and prescribing decisions based on national best practice guidelines.
  2. The modality is unlikely to allow patient expectations to influence provider decision-making. (A 2017 American Psychological Association paper found that when patients clearly expect antibiotics, physicians are more likely to prescribe them.)

What’s Next for Antibiotic Stewardship and Telemedicine?

We, as a clinical leadership team, believe the type of research and study exemplified by Dr. Martinez and her team is vital to telemedicine and virtual care. It produces the type of questions that need to be asked. At Zipnosis, we strive to be transparent with our quality and guideline adherence reporting. Our team would welcome researchers interested in exploring antibiotic stewardship to consider incorporating our data in future studies. As an industry, we should all embrace transparency – after all, high quality care and positive patient outcomes should be our number one goal.