Patient expectations

Patient expectations (or clinician perception of their expectations) may impact antibiotic prescribing.^{Reference Kohut, Keller and Linder10} In a recent study, only 19% of patients with ARI symptoms contacted their doctors, but more than half of these patients expected antibiotics.^{Reference McNulty, Nichols, French, Joshi and Butler11} Unfortunately, clinicians generally do not have a good sense of when these expectations are present. In a study of 287 outpatient visits for ARIs, parents and clinicians completed questionnaires that assessed expectations for antibiotics. Physicians initiated a discussion about antibiotics in 74% of encounters. Parents did so only 8% of the time, with mostly indirect mentions rather than requests for antibiotics. Nevertheless, physicians thought that parents expected antibiotics in these encounters, so they prescribed antibiotics 48% of the time (compared to 10% of the time if they perceived no expectations; P < .001).^{Reference Mangione-Smith, McGlynn, Elliott, McDonald, Franz and Kravitz12} Interestingly, parents who expected antibiotics but did not receive them had much higher satisfaction scores if the physician provided a contingency plan (eg, what to do if symptoms do not improve and promise of re-evaluation). In short, patients rarely ask for or expect antibiotics, but physicians think they do and are likely to prescribe them. Therefore adjusting communication strategies regarding antibiotics may have significant impact on antibiotic prescribing and patient satisfaction.

Risks of illness versus risk of antibiotic treatment

Clinician prescribing practices may depend more on the effect they believe antibiotics will have on their patients’ health than on their awareness of appropriate antibiotic indications. One study demonstrated that if emergency department clinicians believed the risk of antibiotic adverse effects were negligible, and there was even a slight possibility that antibiotics would make their patients feel better (i.e., “Why not take a risk?”), they were more likely to prescribe non–guideline-concordant antibiotics (odds ratio [OR], 1.28; 95% confidence interval [CI], 1.06–1.54).^{Reference Klein, Martinez, May, Saheed, Reyna and Broniatowski13} In contrast, clinicians who believed that antibiotics produce a small risk of harm to the patient (i.e., “antibiotics may be harmful”) were less likely to prescribe inappropriately (OR, 0.81; 95% CI, 0.67–0.98).^{Reference Klein, Martinez, May, Saheed, Reyna and Broniatowski13} These findings demonstrate why antimicrobial stewardship strategies should emphasize the potential harms of antibiotics to patients, especially where there is no benefit (eg, bronchitis or viral upper respiratory infection) or the benefit is slight (eg, sinusitis). Furthermore, despite fewer concerns about antimicrobial resistance, patients and parents of patients still have valid concerns about adverse drug effects and may actually prefer to avoid unnecessary antibiotic use.^{Reference Szymczak, Klieger, Miller, Fiks and Gerber14} Therefore, notwithstanding perceptions of their expectations, primary care clinicians should partner with patients and parents in shared decision making about appropriate antibiotic use as part of their antimicrobial stewardship efforts.

Role of pharmacists in outpatient antimicrobial stewardship

Given that >90% of Americans live within 5 miles of a pharmacy,^{Reference Klepser, Adams and Klepser29} pharmacists may be the most accessible healthcare professionals in the community. Numerous opportunities exist for pharmacists to improve antibiotic prescribing. The Society of Infectious Diseases Pharmacists recently outlined a number of potential pharmacist-led outpatient antimicrobial stewardship initiatives including antibiotic prescription reviews, patient counseling, and prescriber education, as well as more innovative initiatives such as point-of-care testing and initiation of treatment based on a physician–pharmacist collaborative practice agreement.^{Reference Blanchette, Gauthier and Heil28} A randomized pilot study demonstrated increases in guideline-concordant antibiotic prescribing and use of less costly antibiotic therapy by general practitioners after a 15-minute academic detailing visit by a clinical pharmacist that included antibiotic prescribing best practices, the distribution of laminated treatment recommendation charts, and encouraging prescribers to resist patient pressure.^{Reference Ilett, Johnson and Greenhill30} Similarly, a meta-analysis found that involvement of pharmacists in antibiotic stewardship interventions led to a significant reduction in antibiotic prescribing (OR, 0.86; 95% CI, 0.78–0.95) and improvement in the rate of antibiotic prescribing adherence (OR, 1.96; 95% CI, 1.56–2.45) by general practitioners.^{Reference Saha, Hawes and Mazza31} Pharmacist roles in these interventions included facilitating clinician huddles, creating workshops with clinicians, development and delivery of educational materials, and audit and feedback.^{Reference Saha, Hawes and Mazza31}

Pharmacists have also been engaged in developing treatment guidance for outpatient clinicians. A multidisciplinary team that included a pharmacist designed a clinical decision support tool to guide management of uncomplicated urinary tract infections.^{Reference Eudaley, Mihm, Higdon, Jeter and Chamberlin32} This collaboration led to significant reductions in the use of fluoroquinolone (27%) and trimethoprim/sulfamethoxazole (20%), and increases in nitrofurantoin use for cystitis (31%) and guideline-directed duration of therapy (32%).^{Reference Eudaley, Mihm, Higdon, Jeter and Chamberlin32} A multistate evaluation of a collaborative practice agreement between physicians and free-standing pharmacies found that pharmacists were able to accurately triage and manage patients presenting with influenza-like illness.^{Reference Klepser, Klepser, Dering-Anderson, Morse, Smith and Klepser33} In this study, 35% of patients had no primary care physician and 39% visited outside of normal clinic hours; these findings demonstrate the potential value of care using this model.

Role of nurses in outpatient antimicrobial stewardship

Because they function in multiple roles in outpatient settings and communicate with both patients and clinicians, numerous possibilities exist for nurses to affect antimicrobial use. For instance, nurse managers and nurse educators supported implementation of ASPs in 6 outpatient dialysis facilities, resulting in significant decreases in monthly antimicrobial doses while the rate of hospital admissions remained unchanged.^{Reference D’Agata, Lindberg and Lindberg34} In another study, a urine culture follow-up program was implemented in a pediatric urgent care network in which a nurse reviewed urine culture results, forwarded negative results from patients who received empiric antibiotics to clinicians for review, and notified patients or caregivers of plans to discontinue therapy.^{Reference Saha, Patel, Buckingham, Thornton, Barber and Watson35} This program resulted in an increase in antibiotic discontinuation from 4% to 84%, resulting in >3,000 antibiotic days avoided.^{Reference Saha, Patel, Buckingham, Thornton, Barber and Watson35}

Other potential areas of nurse involvement in outpatient antimicrobial stewardship include taking accurate antimicrobial allergy histories, instructing patients on how to provide appropriate specimens for culture, and providing patient education about antibiotics.³⁶

Acute respiratory infections

Antibiotic prescribing in the outpatient setting generally follows seasonal patterns, and more antibiotics are prescribed in the fall and winter seasons during the respiratory viral season.^{Reference Durkin, Jafarzadeh and Hsueh9} During the 2013–2014 and 2014–2015 influenza seasons, 41% of all patients presenting to outpatient clinics with ARI were prescribed antibiotics.^{Reference Havers, Hicks and Chung38} Based on the documented diagnoses, 41% of all antibiotics were not indicated.^{Reference Havers, Hicks and Chung38} Within that same cohort, of 440 patients who were prescribed antibiotics for pharyngitis, 38% had negative group A streptococcal testing, and 38% were prescribed antibiotics for acute sinusitis that were not indicated.^{Reference Havers, Hicks and Chung38} Numerous opportunities exist to both prevent unneeded antibiotic prescribing and to improve the use of guideline-compliant antibiotics.

Urinary tract infections

Urinary tract infections (UTIs) are another common indication for outpatient prescribing. Durkin et al^{Reference Durkin, Keller and Butler39} used a commercial insurance database to retrospectively evaluate outpatient management of UTIs in nonpregnant young women before and after the release of an IDSA clinical practice guideline. A slight decrease in inappropriate antibiotic utilization was observed in the postguideline period. The guideline recommended avoidance of fluoroquinolones as first-line therapy, but their use decreased only minimally in the postintervention period (44.04% vs 41.71%; P < .001), and they continued to be prescribed inappropriately (49.78% vs 47.85%; P < .001), with >75% prescribed for inappropriate durations.^{Reference Durkin, Keller and Butler39}

Skin and soft-tissue infections

Skin and soft-tissue infections (SSTIs) are another clinical syndrome in which antibiotics are often utilized. In a study of outpatient pediatric practices, 77% of antibiotics prescribed for SSTIs were for a duration of >7 days, and 10% of antibiotics prescribed were not first-line agents.^{Reference Jaggi, Wang, Gleeson, Moore-Clingenpeel and Watson40} Furthermore, compared with emergency medicine and family medicine clinicians, pediatricians in that study were more likely to prescribe first-line antibiotics for longer than necessary.^{Reference Jaggi, Wang, Gleeson, Moore-Clingenpeel and Watson40}

Behavioral interventions

Evidence shows that social and behavioral interventions can influence prescribing practices. For example, individuals are more likely to follow through with specific actions if they have publicly declared those intentions. Publicly displayed posters with physician signatures committing to responsible antibiotic prescribing led to a 19.7% reduction in prescribing rates.^{Reference Meeker, Knight and Friedberg42} One medical group created a patient wellness bag containing tissues, cough drops, hand sanitizer, and patient education materials that was distributed to patients with symptoms of bronchitis. This decreased inappropriate antibiotic prescribing from 84% to 8% in primary care clinics and from 67% to 54% in the emergency room.⁴³ Interventions in SSTI might include clinician education of guideline-concordant prescribing coupled with clinical decision support to limit the type and duration of antibiotics prescribed. Similar interventions have shown significant improvement in prescribing.^{Reference Jenkins, Knepper and Sabel44}

Other efforts include “accountable justification” in the electronic medical record (EMR), whereby the prescriber was prompted to justify antibiotic prescription out of guidelines; “suggested alternatives” where an EMR flag based on ARI diagnosis coding displayed treatment alternatives based on diagnosis coding; or “peer comparison” where clinicians who were not in the top decile of appropriate prescribers received an email telling them they were “not a top performer,” along with their rates compared to top performers.^{Reference Meeker, Linder and Fox45} A study comparing these 3 approaches found that mean antibiotic prescribing rates decreased by 5% (95% CI, 7.8%–0.1%; P = .66) for suggested alternatives, decreased by 7% (95% CI, 9.1%–2.9%; P < .001) for accountable justification, and decreased by 5.2% (95% CI, 6.9%–1.6%; P < .001) for peer comparison.^{Reference Meeker, Linder and Fox45} Furthermore, after discontinuing the behavioral interventions, only the prescribers in the peer comparison group demonstrated lasting decreases in inappropriate antibiotic prescribing.^{Reference Linder, Meeker and Fox46} All 3 behavioral interventions required minimal change in prescriber workflow with no diagnosis shifting or documented harm. Another study focused on antibiotic prescribing in pediatric practices and, utilizing education and peer comparison, these researchers demonstrated significant improvements in appropriate antibiotic prescribing for URI, acute bacterial sinusitis, and acute otitis media. However, in this study, the authors suggested that some diagnosis shifting may have occurred.^{Reference Clegg, Bean, Ezzo, Hoth, Sheedy and Anderson47}

Financial incentives

Pay-for-performance initiatives are sometimes used to drive better outcomes from best practices.⁴⁸ In the United Kingdom, the National Health Services (NHS) Quality Premium program provided a 10% quality premium reward to clinical commissioning groups (comprising local general practitioners, and responsible for commissioning healthcare for the community), thereby incentivizing “improved antibiotic prescribing (ie, fewer total prescriptions, fewer broad-spectrum antibiotics, and better prescription data validation) in primary and secondary care.”⁴⁹ Balinskaite et al^{Reference Balinskaite, Johnson, Holmes and Aylin50} evaluated antibiotic prescribing before and after this incentive program and found that financial incentives were associated with decreased overall and broad-spectrum antibiotic prescribing rates, with no adverse effects to patients or community.^{Reference Balinskaite, Bou-Antoun, Johnson, Holmes and Aylin51}

Laboratory testing-guided stewardship

C-reactive protein (CRP) has been associated with the prediction of bacterial pneumonia, and Cals et al^{Reference Cals, Butler, Hopstaken, Hood and Dinant52,Reference Cals, Schot, de Jong, Dinant and Hopstaken53} trained physicians in the use of CRP to predict pneumonia in respiratory illnesses, resulting in a 20% reduction of antibiotic prescribing rates. A multinational European trial randomized practices to receive training in the use of point-of-care CRP testing in respiratory illnesses, or enhanced communication training, or both.^{Reference Little, Stuart and Francis54} The combination of CRP training and enhanced communication training resulted in the highest reduction in antibiotic use compared with either alone. Procalcitonin may be used to guide prescribing for acute respiratory syndromes. One study randomized a group of physician practices to procalcitonin-guided antibiotic prescribing and usual care; the group with procalcitonin-guided treatment had a 72% reduction in antibiotic prescribing compared to the usual care group.^{Reference Briel, Schuetz and Mueller55} However, this approach required procalcitonin measurements during the visit and within 24 hours after the visit and close follow-up, which may not be practical in practice settings in the United States where the no-show rate is very high.

Laboratory testing-guided stewardship is also another area where pharmacists can have an impact. Previously mentioned pharmacist collaborations involved clinical decision support and triaging that resulted in improved stewardship outcomes for UTI^{Reference Eudaley, Mihm, Higdon, Jeter and Chamberlin32} and influenza-like illness.^{Reference Klepser, Klepser, Dering-Anderson, Morse, Smith and Klepser33} Pharmacist prospective review of cultures (predominantly urine) whenever outpatient oral antibiotics were prescribed in a New York Veterans’ Affairs system found that ~20% of prescription encounters with cultures resulted in pharmacist intervention and that 74% of interventions required a change in the antibiotic regimen.^{Reference Wattengel, Sellick and Mergenhagen56} Pharmacist interventions led to reduced rates of 30-day treatment failure (5.4% vs 28.3%; P < .0001) or 30-day hospital admission (0.7% vs 10.9%; P = .0005), when their interventions were accepted.^{Reference Wattengel, Sellick and Mergenhagen56}