Antibiotic Use and Antibiotic Resistance

Antibiotics transformed the practice of medicine in the last half of the 20th century. Penicillin was even considered by many to be a miracle drug. Beginning in the 1940s, antibiotics seemed to be the key to the inevitable elimination of infectious disease as a serious public health problem. With antibiotics, common infections and injuries that would previously have caused death or debility could now be effectively treated and cured. With antibiotic use, however, some bacteria can adapt, which can result in the development of antibiotic resistance, a public health problem which has grown substantially the last several decades. In the United States each year, at least 2 million people acquire infections with antibiotic-resistant bacteria and 23,000 people die of such infections.¹ Although reasons are multifactorial, including the use of antibiotics in livestock, a key factor known to be contributing to higher rates of antibiotic resistance at a population level is high outpatient consumption of antibiotics.^1-3 To emphasize the need to curb the rise of antibiotic resistance as a public health priority, in September 2014, President Obama signed an Ex9ecutive Order that directs combative actions including advancing development of new diagnostics, antibiotics, vaccines, and other therapeutics, strengthening surveillance, and enhancing antibiotic stewardship strategies.⁴

Use of Antibiotics for Acute Respiratory Tract Infections

Acute respiratory tract infections (RTIs) account for approximately 70 percent of primary diagnoses in adults presenting for ambulatory care office visits with a chief symptom of cough.⁵ Acute RTIs include acute bronchitis, acute otitis media (AOM), pharyngitis/tonsillitis, rhinitis, sinusitis, and other viral syndromes. They do not include community acquired pneumonia or acute exacerbations of chronic obstructive pulmonary disease (COPD), bronchiectasis, or other chronic underlying lung diseases.⁶ Standard recommended management of acute RTIs is to focus on ruling out serious illness for which antibiotics are indicated, such as bacterial pneumonia, and providing education and symptomatic relief for uncomplicated illnesses that do not require antibiotics. Existing clinical guidelines for adults and children suggest that acute bronchitis and other acute RTIs that can be caused by either viruses or bacteria, and which are generally self-limiting, should usually not be treated with antibiotics unless certain clinical indications are present.⁶

Despite guidelines recommending no antibiotic treatment for uncomplicated acute RTIs, the majority of outpatient antibiotic prescriptions in the United States are for acute RTIs. In 2007 to 2009, the National Ambulatory and National Hospital Ambulatory Medical Care Surveys found that antibiotics were prescribed during 101 million annual ambulatory visits for patients aged 18 years and above.⁷ In 2010, approximately 801 outpatient antibiotic prescriptions were dispensed per 1,000 inhabitants in the United States.^8,9 A 2013 report regarding healthy adults visiting outpatient offices and emergency departments (EDs) for acute bronchitis revealed prescriptions for antibiotics were given at 73 percent of visits between 1996 and 2010,¹⁰ despite the fact that the majority of acute bronchitis cases are caused by viral pathogens for which antibiotics are not helpful. Similarly, A 2014 analysis of data from the National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey indicated that 60 percent of children diagnosed with pharyngitis in the United States between 1997 and 2010 were prescribed antibiotics,¹¹ despite that the fact that only about 37 percent of pharyngitis episodes are caused by bacteria. Clearly, there is a need to identify and promote strategies that can help to bring antibiotic use for RTIs in line with current evidence-based guidelines.

Strategies To Improve Appropriate Antibiotic Use

The reasons for overuse of antibiotics for acute RTIs are influenced by numerous, diverse, and complex factors, both internal and external, including geographic location, environment (i.e., clinic type, time and resources), patient demographics (i.e., children versus adults, gender, race/ethnicity), patient and/or clinician preferences, ability to obtain followup care, clinician specialty, knowledge, experience, clinical inertia, peer group influence and oversight or feedback from infectious disease experts, and provider-patient communication and relationship.^12-14 Consequently, strategies to reduce antibiotic use for acute RTIs have varied in their targets. Strategies may target clinicians who care for patients with acute RTIs in outpatient settings, adult and/or pediatric patients with acute RTIs, the parents of pediatric patients with acute RTIs, healthy adults and/or children in the general population without a current RTI, or groups whose attendance policies may indirectly affect the use of antibiotics (e.g., employers, school officials).

Interventions may also fall into any of several categories based on their approach. Educational strategies include educating clinicians about current treatment guidelines or providing information to patients or parents of patients about why antibiotic treatment is not recommended. Strategies to improve communication between clinicians and patients include interventions designed to improve shared decisionmaking around use of antibiotics for acute RTIs. Clinical strategies include delayed prescribing of antibiotics or use of point-of-care diagnostic tests (e.g., rapid strep). System level strategies include clinician reminders (paper-based or electronic), clinician audit and feedback, and financial or regulatory incentives for clinicians or patients. Furthermore, multifaceted approaches may include numerous elements of one or more of the aforementioned strategies.

Measuring Effectiveness of Strategies To Improve Appropriate Antibiotic Use

The primary goal of improving appropriate antibiotic use is to slow the evolution of antibiotic resistance. Unfortunately, measuring this outcome accurately would require large populations and long time periods, and these types of studies are largely unavailable.

Another potential benefit of reducing overall antibiotic prescriptions is the reduced exposure of patients to potential adverse side effects. Recent studies reported in the news have drawn attention to potential adverse effects of antibiotics beyond those more established side effects such as allergic reactions or gastrointestinal disruption. One such report indicated that children with four or more courses of broad-spectrum antibiotics in their first 2 years of life were more likely to be obese later in childhood.¹⁸ Another recent report discussed evidence that certain antibiotics might be associated with increased risks of death and serious cardiac arrhythmias during standard treatment durations.¹⁹ The cost to patients and the healthcare system of unnecessary antibiotics should also be considered. However these other important secondary goals, like antibiotic resistance, are understudied. Therefore it is necessary to consider intermediate outcomes to evaluate the effectiveness of interventions.

The most logical intermediate outcome would be changes in appropriate antibiotic use. However, although guidelines suggest when antibiotic use is warranted, defining and determining “appropriate” use for study purposes is often difficult because it is subjective and requires both access to adequate patient-level data and clinical knowledge. For these and other reasons, the effectiveness of various interventions on antibiotic overuse has generally evaluated the impact of interventions upon overall antibiotic use, based on the understanding that for certain clinical conditions the majority of antibiotic use is unnecessary. For example, studies find that half or more of antibiotic prescriptions for acute RTIs are not necessary.^15-17 Accordingly, measures of overall change in antibiotic prescription use in such conditions are a relevant proxy for changes in appropriate use. Similarly, while “prescription” and “use” are not synonymous, measuring actual use is much more difficult and resource intensive than counting prescriptions. The usefulness of overall prescribing as a proxy for appropriate prescribing may vary because the ratio of inappropriate to appropriate prescribing can range so widely based on patient, provider and setting factors. For example, one study found a rate of 80 percent inappropriate prescribing of antibiotics for acute RTI,¹⁶ suggesting overall prescribing is a fairly good proxy for appropriate prescribing, while another study that reported 50 percent inappropriate prescribing suggests a much lower level of confidence in the proxy measure.¹⁷ Nevertheless, it is the most widely studied outcome for these interventions.

A main concern with reducing overall outpatient use of antibiotics for RTIs is that it would increase the risk of under-treatment of patients for whom antibiotics would have been indicated and lead to increases in undesirable outcomes such as hospitalization, medical complications, clinic visits, time off work and/or school, and longer symptom duration. Depending on patients' expectations, patient satisfaction may also be affected. The interventions themselves also may require substantial time and resources. Therefore, occurrence of these negative outcomes must be weighed against any reduction in antibiotic overuse that might be associated with a particular intervention.

As numerous patient, clinician, and setting factors may modify the comparative effectiveness of interventions to improve appropriate antibiotic use (e.g., type of RTI, patient demographics, clinician specialty, type of clinic, geographic location, etc.), a review of the evidence should seek to clarify whether there are particular subpopulations that are more or less likely to benefit.

Existing Systematic Reviews and Guidelines Addressing Antibiotic Use for Acute Respiratory Tract Infections

There are a number of existing systematic reviews and guidelines that have contributed to our understanding of what works for targeted populations, interventions, or diseases.^20-27 The most comprehensive review to date, a 2006 technical review by Agency for Healthcare Research and Quality (AHRQ), entitled “Closing the Quality Gap: A Critical Analysis of Quality Improvement Strategies Volume 4—Antibiotic Prescribing Behavior” concluded that some quality improvement strategies may be moderately effective in reducing inappropriate antibiotic prescription. While no single strategy is clearly superior, the report concludes that clinician education and delayed prescribing may be more effective in certain settings and that interventions targeting prescribing for all acute RTIs may be more effective than those that target a single type of RTI. However, because improving antibiotic prescribing has become an even more urgent public health priority, there is an important need for an updated comparative effectiveness review that comprehensively addresses a broad range of populations and interventions. Therefore, the goal of the present systematic evidence review is to assess the comparative effectiveness of a breadth of possible strategies for reducing antibiotic use when not indicated for acute RTIs in adults and children. In addition to providing evidence on the benefits and potential harms of strategies, the review identifies gaps in the literature and suggestions to guide future research.

Scope and Key Questions

The Key Questions and analytic framework used to guide this report are shown below. The analytic framework (Figure 1) illustrates the scope of this review, including the target population, interventions, comparison, and outcomes that were examined in this review.

Figure 1

Analytic framework for improving appropriate antibiotic use for acute respiratory tract infections. ED = emergency department, KQ = Key Question, OTC = over-the-counter, QoL = quality of life, RTI = respiratory tract infection

Key Question 1. For adults and children with acute respiratory tract infection, what is the comparative effectiveness of particular strategies in improving the appropriate prescription or use of antibiotics compared with other strategies or standard care?

1. Does the comparative effectiveness of strategies differ according to how appropriateness is defined?
2. Does the comparative effectiveness of strategies differ according to the intended target of the strategy (i.e., clinicians, patients, and both)?
3. Does the comparative effectiveness of strategies differ according to patient characteristics, such as type of respiratory tract infection, signs and symptoms (nature and duration), previous medical history (e.g., frailty, comorbidity), prior respiratory tract infections, prior use of antibiotics, age, ethnicity, socioeconomic status, and educational level attained?
4. Does the comparative effectiveness of strategies differ according to clinician characteristics, such as specialty, number of years in practice, type of clinic organization, geographic region, and population served?
5. Does the comparative effectiveness differ according to the diagnostic method or definition used, the clinician's perception of the patient's illness severity, or the clinician's diagnostic certainty?
6. Does the comparative effectiveness differ according to various background contextual factors, such as the time of year, known patterns of disease activity (e.g., an influenza epidemic, a pertussis outbreak), system-level characteristics, or whether the intervention was locally tailored?

Key Question 2. For adults and children with acute respiratory tract infection, what is the comparative effect of particular strategies on antibiotic resistance compared with other strategies or standard care?

1. Does the comparative effect of strategies differ according to the intended target of the strategy (i.e., clinicians, patients, and both)?
2. Does the comparative effect of strategies differ according to patient characteristics, such as type of respiratory tract infection, signs and symptoms (nature and duration), previous medical history (e.g., frailty, comorbidity), prior respiratory tract infections, prior use of antibiotics, age, ethnicity, socioeconomic status, and educational level attained?
3. Does the comparative effect of strategies differ according to clinician characteristics, such as specialty, number of years in practice, type of clinic organization, geographic region, and population served?
4. Does the comparative effectiveness differ according to the diagnostic method or definition used, the clinician's perception of the patient's illness severity, or the clinician's diagnostic certainty?
5. Does the comparative effect differ according to various background contextual factors, such as the time of year, known patterns of disease activity (e.g., an influenza epidemic, a pertussis outbreak), whether the intervention was locally tailored, system-level characteristics, or the source of the resistance data (i.e., population vs. study sample)?

Key Question 3. For adults and children with acute respiratory tract infection, what is the comparative effect of particular strategies on medical complications (including mortality, hospitalization, and adverse effects of receiving or not receiving antibiotics) compared with other strategies or standard care?

1. Does the comparative effect of strategies differ according to the intended target of the strategy (i.e., clinicians, patients, and both)?
2. Does the comparative effect of strategies differ according to patient characteristics, such as type of respiratory tract infection, signs and symptoms (nature and duration), previous medical history (e.g., frailty, comorbidity), prior respiratory tract infections, prior use of antibiotics, age, ethnicity, socioeconomic status, and educational level attained?
3. Does the comparative effect of strategies differ according to clinician characteristics, such as specialty, number of years in practice, type of clinic organization, geographic region, and population served?
4. Does the comparative effectiveness differ according to the diagnostic method or definition used, the clinician's perception of the patient's illness severity, or the clinician's diagnostic certainty?
5. Does the comparative effect differ according to various background contextual factors, such as the time of year, known patterns of disease activity (e.g., an influenza epidemic, a pertussis outbreak), whether the intervention was locally tailored or system-level characteristics?

Key Question 4. For adults and children with acute respiratory tract infection, what is the comparative effect of particular strategies on other clinical outcomes (e.g., health care utilization, patient satisfaction) compared with other strategies or standard care?

1. Does the comparative effect of strategies differ according to the intended target of the strategy (i.e., clinicians, patients, and both)?
2. Does the comparative effect of strategies differ according to patient characteristics, such as type of respiratory tract infection, signs and symptoms (nature and duration), previous medical history (e.g., frailty, comorbidity), prior respiratory tract infections, prior use of antibiotics, age, ethnicity, socioeconomic status, and educational level attained?
3. Does the comparative effect of strategies differ according to clinician characteristics, such as specialty, number of years in practice, type of clinic organization, geographic region, and population served?
4. Does the comparative effectiveness differ according to the diagnostic method or definition used, the clinician's perception of the patient's illness severity, or the clinician's diagnostic certainty?
5. Does the comparative effect differ according to various background contextual factors, such as the time of year, known patterns of disease activity (e.g., an influenza epidemic, a pertussis outbreak), whether the intervention was locally tailored or system-level characteristics?

Key Question 5. For adults and children with acute respiratory tract infection, what is the comparative effect of particular strategies on achieving intended intermediate outcomes, such as improved knowledge regarding use of antibiotics for acute respiratory tract infections (clinicians and/or patients), improved shared decisionmaking regarding the use of antibiotics, and improved clinician skills for appropriate antibiotic use (e.g., communication appropriate for patients' literacy level and/or cultural background)?

Key Question 6. What are the comparative nonclinical adverse effects of strategies for improving the appropriate use of antibiotics for acute respiratory tract infections (e.g., increased time burden on clinicians, patients, clinic staff)?

Analytic Framework

The analytic framework below (Figure 1) illustrates the population, interventions, outcomes, and adverse effects that guided the literature search and synthesis and their relationship to the Key Questions.

Organization of This Report

For each Key Question, results are organized into subsections for each intervention category. We arranged the subsections to match the ordering of the intervention categories as listed in the inclusion criteria. Within each intervention category subsection, evidence was further grouped by specific intervention type (i.e., delayed prescribing, specific point-of-care tests for the clinical section) and ordered based on volume of evidence (most to least).