eTable 1. Oral vs Parenteral Antibiotic Classification
eTable 2. Diagnostic Categories by Tier With Corresponding ICD-9CM Codes
eTable 3. Clinical Practice Guideline Recommendations for Adults and Children by Diagnosis
eTable 4. Weighted Mean Annual Rate of Total Visits and Rate of Visits With Antibiotics Prescribed (95% CI) per 1000 Population, by Diagnosis by US Census Region and Age, NAMCS/NHAMCS — United States, 2010–2011
eTable 5. Sensitivity Analysis With Weighted Mean Annual Antibiotic Prescribing Rates in 2010–2011 NAMCS/NHAMCS and Estimated Appropriate Annual Antibiotic Prescribing Rates per 1000 Population Using Lowest Regional Rates for all Diagnoses Targeted for Reductions and No Changes Targeted for Pneumonia, UTI, and Miscellaneous Bacterial Infections (eg, Pertussis and Syphilis) by Age Group and Diagnosis
Customize your JAMA Network experience by selecting one or more topics from the list below.
Fleming-Dutra KE, Hersh AL, Shapiro DJ, et al. Prevalence of Inappropriate Antibiotic Prescriptions Among US Ambulatory Care Visits, 2010-2011. JAMA. 2016;315(17):1864–1873. doi:10.1001/jama.2016.4151
The National Action Plan for Combating Antibiotic-Resistant Bacteria set a goal of reducing inappropriate outpatient antibiotic use by 50% by 2020, but the extent of inappropriate outpatient antibiotic use is unknown.
To estimate the rates of outpatient oral antibiotic prescribing by age and diagnosis, and the estimated portions of antibiotic use that may be inappropriate in adults and children in the United States.
Design, Setting, and Participants
Using the 2010-2011 National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey, annual numbers and population-adjusted rates with 95% confidence intervals of ambulatory visits with oral antibiotic prescriptions by age, region, and diagnosis in the United States were estimated.
Ambulatory care visits.
Main Outcomes and Measures
Based on national guidelines and regional variation in prescribing, diagnosis-specific prevalence and rates of total and appropriate antibiotic prescriptions were determined. These rates were combined to calculate an estimate of the appropriate annual rate of antibiotic prescriptions per 1000 population.
Of the 184 032 sampled visits, 12.6% of visits (95% CI, 12.0%-13.3%) resulted in antibiotic prescriptions. Sinusitis was the single diagnosis associated with the most antibiotic prescriptions per 1000 population (56 antibiotic prescriptions [95% CI, 48-64]), followed by suppurative otitis media (47 antibiotic prescriptions [95% CI, 41-54]), and pharyngitis (43 antibiotic prescriptions [95% CI, 38-49]). Collectively, acute respiratory conditions per 1000 population led to 221 antibiotic prescriptions (95% CI, 198-245) annually, but only 111 antibiotic prescriptions were estimated to be appropriate for these conditions. Per 1000 population, among all conditions and ages combined in 2010-2011, an estimated 506 antibiotic prescriptions (95% CI, 458-554) were written annually, and, of these, 353 antibiotic prescriptions were estimated to be appropriate antibiotic prescriptions.
Conclusions and Relevance
In the United States in 2010-2011, there was an estimated annual antibiotic prescription rate per 1000 population of 506, but only an estimated 353 antibiotic prescriptions were likely appropriate, supporting the need for establishing a goal for outpatient antibiotic stewardship.
Antibiotic-resistant infections affect 2 million people and are associated with 23 000 deaths annually in the United States, according to the Centers for Disease Control and Prevention (CDC).1 Antibiotic use is the primary driver of antibiotic resistance1 and leads to adverse events ranging from allergic reactions to Clostridium difficile infections.2 Data from other developed nations suggest that 80% or more of antibiotic use (excluding agricultural use) occurs among outpatients.3 In the United States in 2011, 262 million outpatient antibiotic prescriptions were dispensed.4 Declining trends in antibiotic prescribing in the United States occurred in the 1990s and early 2000s but stabilized by 2010.5,6
Reducing inappropriate use is essential to reduce both antibiotic resistance and adverse events. National guidelines addressing when to prescribe antibiotics have been published over the last 2 decades for many common diagnoses.7-10 The CDC’s Get Smart: Know When Antibiotics Work program focuses on promoting appropriate outpatient antibiotic use.11 In March 2015, the White House released the National Action Plan for Combating Antibiotic-Resistant Bacteria, which set a target of reducing inappropriate antibiotic use in the outpatient setting by 50% by 2020.12Quiz Ref ID However, the fraction of antibiotic use that is inappropriate and amenable to reduction is unknown.13 Previous goals and national measures of appropriate antibiotic use (eg, Healthy People 2020 targets and Healthcare Effectiveness Data and Information Set [HEDIS] measures) have targeted specific age groups and conditions (eg, ear infections or acute bronchitis).14,15 To our knowledge, there is no current overall estimate of appropriate outpatient antibiotic prescribing that considers all ages and conditions that may receive antibiotics. The objectives of this study were to establish a baseline of the current rate of outpatient, oral antibiotic prescriptions by age and diagnosis and to estimate the overall rate of appropriate, outpatient antibiotic prescriptions in the United States to inform public health and antibiotic stewardship efforts.
Baseline antibiotic prescribing rates were estimated using the National Ambulatory Medical Care Survey (NAMCS) and National Hospital Ambulatory Medical Care Survey (NHAMCS). These cross-sectional surveys are administered annually by the CDC’s National Center for Health Statistics.16 NAMCS samples visits to nonfederally-employed, office-based physicians primarily engaged in patient care and selects visits using a 3-stage probability sampling design. NAMCS samples geographic regions, physicians, and visits during randomly assigned 1-week reporting periods. NHAMCS is a survey of emergency departments and outpatient departments of nonfederal general and short-stay hospitals.16 NHAMCS uses a 4-stage probability sampling design. NHAMCS samples geographic regions, hospitals, outpatient department clinics, and emergency service areas, and visits during 4-week reporting periods. Data collected include patient demographics, up to 3 diagnoses that are coded by NAMCS/NHAMCS staff using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), and up to 8 medications mentioned—meaning prescribed, continued, or provided at the visit.
Sampled visits in both surveys are distributed randomly throughout the year. Weights are assigned to visits so national estimates can be produced. The most recent 2 years of data available in both NAMCS and NHAMCS (2010-2011) were used. Unweighted response rates during 2010-2011 ranged from 54.1% to 58.3% for physicians in NAMCS, 66.7% to 73.6% for outpatient departments, and 80.4% to 87.5% for emergency departments in NHAMCS. Estimates from NAMCS are adjusted for physician and item nonresponse. Estimates from NHAMCS are adjusted for hospital nonresponse and for nonresponse at the level of the emergency departments and outpatient departments. The National Center for Health Statistics research ethics review board approved NAMCS/NHAMCS, with waivers of informed consent and Health Insurance Portability and Accountability Act authorization for patients. These analyses were based on publically available, deidentified data, and therefore were not subject to institutional review board requirements as determined in consultation with the human subjects advisor for the National Center for Emerging and Zoonotic Infectious Disease.
All visits during 2010-2011 were included unless resulting in hospital or observation unit admission or if parenteral antibiotics were mentioned without oral antibiotics (0.4% of visits). National estimates were produced from 184 032 unweighted visits in 2010-2011. Antibiotics were coded in terms of their generic components and therapeutic classes using Lexicon Plus (Cerner Multum).16 Data were not available on route of administration. Antibiotics supplied only in topical formulations were excluded. Oral antibiotics were defined as antibiotics with oral formulations, and parenteral antibiotics were defined as those antibiotics not available in oral formulations (eTable 1 in the Supplement). As outpatient, oral antibiotics are only available via prescription, medication mentions of antibiotics were assumed to be prescriptions, similar to previous studies.17-19
The mean annual rates per 1000 population of visits with antibiotic prescriptions by age group (0-2, 3-9, 10-19, 20-39, 40-64, ≥65 years) and US Census region (Northeast, South, Midwest, and West) for 2010-2011 were calculated. The Northeast region included Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, New York, New Jersey, and Pennsylvania. The Midwest region included Ohio, Michigan, Indiana, Illinois, Wisconsin, Missouri, Iowa, Minnesota, Kansas, Nebraska, South Dakota, and North Dakota. The South region included Delaware, Maryland, District of Columbia, Virginia, West Virginia, North Carolina, South Carolina, Florida, Georgia, Tennessee, Kentucky, Alabama, Mississippi, Louisiana, Arkansas, Oklahoma, and Texas. The West region included Montana, Wyoming, Colorado, New Mexico, Arizona, Utah, Idaho, Washington, Oregon, Nevada, California, Alaska, and Hawaii. The mean annual rate per 1000 population, number, and percentage of visits with antibiotics prescribed were also calculated by diagnosis and age group. For diagnosis-based analyses, larger age groups (0-19, 20-64, ≥65 years) were used to produce more reliable estimates.
Diagnostic categories for common outpatient infections (eTable 2 in the Supplement) were established and visits were classified into these categories. As NAMCS/NHAMCS do not contain links between diagnoses and medication mentions, visit diagnoses were classified based on the most likely indication for an antibiotic prescription in a tiered fashion. Tier 1 diagnoses were diagnoses for which antibiotics are almost always indicated: pneumonia, urinary tract infection (UTI), or miscellaneous bacterial infections (eg, pertussis and syphilis). Tier 2 diagnoses are diagnoses for which antibiotics may be indicated: sinusitis; suppurative otitis media; skin, cutaneous, and mucosal infections; pharyngitis; gastrointestinal infections; and acne. Tier 3 diagnoses were all other diagnostic categories for which antibiotics are not indicated or the indication was unclear. In assigning each visit a single diagnosis, priority was given to tier 1 diagnoses, then tier 2 diagnoses, then tier 3 diagnoses. If a visit had multiple diagnoses from a single tier, the first-listed diagnosis was assigned.
To estimate appropriate oral antibiotic prescribing in ambulatory care, the Pew Charitable Trusts convened a group of experts on outpatient antibiotic use to determine methods for estimating the fraction of antibiotic use that is appropriate. Each author reviewed available national guidelines from professional societies for common conditions (eTable 3 in the Supplement). Quiz Ref IDEstimated levels of appropriate antibiotic prescribing by age group were based on national guidelines for diagnoses for which national guidelines could be used to recommend specific antibiotic prescribing rates: pharyngitis; asthma, allergy; bronchitis, bronchiolitis; influenza; nonsuppurative otitis media; viral upper respiratory tract infection; viral and nonviral pneumonia; UTI; and miscellaneous bacterial infections (eg, pertussis and syphilis). Regional variability was used to produce estimates of appropriate antibiotic prescriptions for diagnoses for which guidelines could not be used to recommend specific antibiotic prescribing rates: sinusitis, suppurative otitis media, and the remaining other conditions. Regional analyses of antibiotic prescribing by age group were conducted to identify regions with the lowest regional antibiotic prescription rates, recognizing there is evidence of antibiotic overuse even in the lowest-prescribing regions.20 Estimates were combined using national guidelines and regional variability to calculate an overall estimate of appropriate antibiotic prescriptions per 1000 population.
Statistical analyses were performed using STATA 12 (STATA Corp) and accounted for the components of the complex survey design including patient visit weights, strata, and primary sampling unit design variables. Two years of survey data were combined for analyses to increase sample size by age and diagnosis. Ninety-five percent confidence intervals were calculated for all estimates. Antibiotic prescribing rates were compared using χ2 test for heterogeneity. Estimates were not calculated if based on fewer than 30 sampled visits or if the relative standard error was less than 0.3, as such estimates do not meet standards of reliability or precision.16 Significance was considered at 2-sided P value less than .05. Population denominators were based on the July 1, 2010, and July 1, 2011, set of estimates of the civilian, noninstitutional population of the United States, as developed by the Population Division, US Census Bureau.21 Additionally, a post hoc sensitivity analysis was performed using the lowest-prescribing region as the benchmark for all estimates of appropriate antibiotic prescribing for diagnoses targeted for reduction.
In 2010-2011, of the 184 032 sampled ambulatory care visits, 12.6% (95% CI, 12.0%-13.3%) were associated with antibiotic prescriptions (Table 1), with an estimated 506 antibiotic prescriptions (95% CI, 458-554) per 1000 US population annually (Table 2). Per 1000 population, antibiotic prescribing rates ranged from 423 antibiotic prescriptions (95% CI, 343-504) in the West to 553 antibiotic prescriptions (95% CI, 459-648) in the South (Table 3). The annual antibiotic prescription rate was highest among children aged 0 through 2 years at 1287 antibiotic prescriptions (95% CI, 1085-1489) per 1000 population.
Quiz Ref IDThe top 3 diagnoses associated with antibiotic prescriptions per 1000 population were sinusitis (56 antibiotic prescriptions [95% CI, 48-64] ), suppurative otitis media (47 antibiotic prescriptions [95% CI, 41-54]), and pharyngitis (43 antibiotic prescriptions [95% CI, 38-49]) (Table 2). Collectively, acute respiratory conditions, defined as sinusitis, suppurative otitis media, nonsuppurative otitis media, pharyngitis, viral upper respiratory tract infection, bronchitis and bronchiolitis, asthma and allergy, influenza and viral and nonviral pneumonia, were associated with an estimated 221 antibiotic prescriptions (95% CI, 198-245) per 1000 population annually.
For pharyngitis, national guidelines recommend antibiotic therapy only for streptococcal pharyngitis.7 Based on recent literature, 37% of children presenting for medical visits with sore throat tested positive for group A Streptococcus.22,23 However, 56.2% of visits (95% CI, 49.8%-62.4%) by children with pharyngitis were associated with antibiotic prescribing in 2010-2011 (Table 1). For adults presenting for medical visits with sore throat, about 18% have streptococcal pharyngitis (ie, meet clinical criteria and have a positive test result for group A Streptococcus),7,23 but 72.4% (95% CI, 66.8%-77.4%) of visits by adults aged 20 through 64 years with pharyngitis were associated with antibiotic prescribing.
National guidelines state that patients with bronchitis (excluding visits with diagnoses of chronic bronchitis, emphysema, or chronic obstructive pulmonary disease),24,25 bronchiolitis,26 viral upper respiratory tract infections,25,27 asthma and allergy,28 influenza,29 viral pneumonia30 and nonsuppurative otitis media31 should not receive antibiotics, and thus antibiotics for these conditions were considered inappropriate.
Sinusitis and suppurative otitis media sometimes warrant antibiotics; however, the appropriate rate of antibiotic prescriptions for these conditions is uncertain. Thus, the lowest regional rate of visits with antibiotic prescriptions was selected as the estimate of the appropriate rate (eTable 4 in the Supplement). For sinusitis, the estimated appropriate antibiotic prescription rates per 1000 participants by age group were as follows: 59 (95% CI, 32-86) for 0 through 19 years, 27 (95% CI, 17-36) for 20 through 64 years, and 37 (95% CI, 16-59) for 65 years and older. For suppurative otitis media, the estimated appropriate antibiotic prescription rates per 1000 participants by age group were as follows: 138 (95% CI, 96-179) for 0 through 19 years and 6 (95% CI, 4-9) for 20 through 64 years. Estimates could not be calculated for adults aged 65 years or older with suppurative otitis media due to the small number of sampled visits.
Antibiotic prescription rates were considered appropriate for pneumonia because the diagnosis almost always warrants antibiotic therapy.30
For other conditions, excluding acute respiratory conditions listed above, antibiotic prescriptions were considered appropriate for UTI and miscellaneous bacterial infections (eg, pertussis and syphilis), as these conditions almost always warrant antibiotic therapy.32 For the remaining other conditions (excluding acute respiratory conditions, UTIs, and miscellaneous bacterial infections), the lowest regional rate of visits with antibiotic prescriptions per 1000 participants by age group was selected as the estimated appropriate rate (eTable 4 in the Supplement): 137 (95% CI, 100-174) for 0 through 19 years, 180 (95% CI, 140-220) for 20 through 64 years, and 362 (95% CI, 272-452) for 65 years or older.
The actual and estimated appropriate antibiotic prescription rates by age group for each condition and overall are shown in Table 4. The estimated annual appropriate antibiotic prescription rate for acute respiratory conditions per 1000 population was 111 antibiotic prescriptions for all ages vs 221 antibiotic prescriptions (95% CI, 198-245) annually for acute respiratory conditions in 2010-2011, a 50% reduction from the point estimate of the 2010-2011 annual rate. The estimated annual appropriate rate per 1000 population for other conditions (excluding acute respiratory) was 242 antibiotic prescriptions vs 284 (95% CI, 256-313) annually in 2010-2011, a 15% reduction from the point estimate of the 2010-2011 annual rate. When all conditions were combined, the estimated appropriate annual antibiotic prescription rate for all conditions in all ages in the United States per 1000 population was 353 vs 506 (95% CI, 458-554) annually in 2010-2011, a 30% reduction from the point estimate of the 2010-2011 annual rate. The results of the sensitivity analysis using the lowest-prescribing region as the estimate of appropriate antibiotic prescribing for all diagnoses identified for reduction are shown in eTable 5 in the Supplement.
Quiz Ref IDThese analyses describe antibiotic prescribing practices during ambulatory care visits in the United States by age group and diagnosis in 2010-2011, with an overall estimated annual rate of 506 antibiotic prescriptions per 1000 US population. In the United States, an estimated 154 million prescriptions for antibiotics were written in ambulatory care settings annually during 2010-2011. Half of antibiotic prescriptions for acute respiratory conditions may have been unnecessary, representing 34 million antibiotic prescriptions annually. Collectively, across all conditions, an estimated 30% of outpatient, oral antibiotic prescriptions may have been inappropriate. Therefore, a 15% reduction in overall antibiotic use would be necessary to meet the White House National Action Plan for Combating Antibiotic-Resistant Bacteria goal of reducing inappropriate antibiotic use in the outpatient setting by 50% by 2020.12 This estimate of inappropriate outpatient antibiotic prescriptions can be used to inform antibiotic stewardship programs in ambulatory care by public health and health care delivery systems in the next 5 years.
This estimate of inappropriate outpatient antibiotic prescriptions is based on deriving an estimate for the rate of ambulatory care visits during which antibiotics are prescribed when not indicated and for diagnoses that are overused (eg, sinusitis). To derive this estimate, a conservative, mixed approach was used including (1) no reductions for conditions almost always warranting antibiotics, (2) application of clinical guidelines for nonbacterial respiratory conditions, (3) an estimate for group A streptococcal prevalence for pharyngitis, based on evidence-based guidelines, and (4) use of geographic variability by focusing on regions with the lowest antibiotic prescribing rates for suppurative otitis media, sinusitis, and remaining conditions. Previous studies estimating potential estimates of inappropriate outpatient antibiotic use have relied solely on estimates of bacterial prevalence and have concluded that more than 50% of antibiotics for acute respiratory tract infections are unnecessary.17,18 Although knowledge of bacterial pathogen presence could be informative, routine performance of sinus aspiration or tympanocentesis is impractical. Current treatment guidelines for sinusitis and otitis media advocate using stringent clinical criteria for diagnosis and to identify candidates for watchful waiting and antibiotic therapy,8-10,33 but the fraction of diagnoses that meet these criteria is unknown.
Quiz Ref IDA substantial amount of antibiotic overuse is likely driven by overdiagnosis of certain conditions (eg, sinusitis diagnosis without meeting criteria). Evidence supporting the problem of overdiagnosis, in particular for sinusitis and otitis media, includes variability in diagnosis by race and by clinician.34 Although the health of the populations in each region and thus the need for antibiotic use may vary, the regional analyses from which the estimated appropriate rates are taken are based on large regions with populations ranging from 55 million (Northeast) to 113 million (South). To our knowledge, no data suggest worse outcomes for these conditions in low-prescribing regions due to undertreatment; in fact, there is evidence of antibiotic overuse even in low-prescribing regions.20 For pneumonia, UTI and miscellaneous bacterial infections, no specific goal was recommended at this time, even though overdiagnosis of these conditions likely occurs, particularly with asymptomatic bacteriuria diagnosed as UTI. For this effort, the group focused on conditions for which the diagnoses often or sometimes do not require antibiotics, rather than misdiagnosis of conditions almost always requiring antibiotics. Misdiagnosis will present important opportunities in the future, especially for infections like UTI and community-acquired pneumonia.
The geographic variation nationally and globally indicates that a 30% reduction in antibiotic prescribing is achievable. Per the sensitivity analysis, if national antibiotic prescriptions rates for targeted diagnoses for each age group were at the rate of the lowest-prescribing region, prescribing would be 19% lower than the 2010-2011 rate. However, when low-prescribing regions are used to produce estimates for appropriate prescribing, more antibiotic prescriptions are considered appropriate than when estimates are based on national guidelines.
Another study examining dispensed antibiotics found the lowest-prescribing 5 states dispensed 36% fewer antibiotic prescriptions than the United States overall; this study found a similar pattern to our study of higher prescribing in the South vs the West.4 Sweden dispensed 328 antibiotic courses per 1000 population in 20143 compared with 877 antibiotic courses dispensed per 1000 population in the United States in 2011.35 Sweden has very low rates antibiotic-resistant infections.36 By reducing antibiotic prescribing for common infections, 1 potential concern is increasing suppurative complications. However, in Sweden, acute otitis media diagnoses and antibiotic use decreased during 2000-2005 by 50%, likely due to stricter diagnostic criteria and educational campaigns, with no increase in mastoiditis,37 all prior to introduction of pneumococcal conjugate vaccine.38 Effective interventions to reduce inappropriate prescribing, such as clinician and patient education, audit-and-feedback, academic detailing, communication training, rapid diagnostics, clinical decision support, and delayed prescriptions, can be used in ambulatory care settings to improve appropriate antibiotic use.39
Strengths of NAMCS/NHAMCS include national representativeness and inclusion of both diagnoses and therapy. However, the use of NAMCS/NHAMCS in this study also has the following limitations. First, the NAMCS and NHAMCS rely on clinician diagnoses and lack data allowing for validation of diagnoses and links between medications and diagnoses. ICD-9-CM codes, although assigned by NAMCS/NHAMCS staff, lack specificity to differentiate all diagnoses of interest and may not fully represent the clinician’s diagnosis. NAMCS/NHAMCS allow for 3 diagnoses, and if the visit lacked antibiotic-appropriate diagnosis, it was assumed that none existed. However, all 3 diagnoses fields were used and antibiotic prescribing was considered appropriate if there was any antibiotic-appropriate diagnosis.
Second, estimates in NAMCS and NHAMCS are adjusted for nonresponse for a limited number of variables, and thus nonresponse bias may be present.40
Third, data from NAMCS and NHAMCS represent visits, not illness episodes, such that multiple prescriptions for 1 person would be captured separately, and therefore the overall estimated rates of antibiotic prescribing per population and inappropriate prescribing may be overestimated. For instance, in 2010-2011 in NAMCS/NHAMCS, an estimated 1.2 billion ambulatory care visits occurred annually from a population of 305 million people in the United States, suggesting that, on average, approximately 4 ambulatory care visits occurred for each person in the United States. In reality, some individuals had more than 4 ambulatory care visits, and some had fewer or no visits. Of these projected 1.2 billion visits annually, an estimated 12.6% were associated with antibiotic prescriptions, which would equate to a projected 154 million antibiotic prescriptions written annually in these settings or an estimated population-based rate of 506 antibiotic prescriptions per 1000 US population (ie, every person in the population would have received 0.5 antibiotic prescriptions per year from an office or emergency department–based visit). However, some patients received no antibiotic prescriptions, and some likely received more than 1 antibiotic prescription. In addition, because some individuals had multiple visits and would have been sampled multiple times, it is almost certain that complex patients (who would have many visits annually) are overrepresented in the study population compared with truly “typical” patients, and far more than healthy patients, who may not have had any visits during the sampling frame. It also is possible that for these complex patients, what may seem like inappropriate prescriptions could have been appropriate. Moreover, the more times an individual sees a physician, the more likely that patient will get a prescription, independent of whether the prescription was appropriate or not. However, these multiple exposures for an individual are treated as “on average” a single exposure for multiple individuals and the data were not adjusted for comorbidities or for the number of times an individual is included in the sampling frame, suggesting that the overall estimated rate of antibiotic prescribing per population may be biased upward.
Fourth, NAMCS/NHAMCS also have significant time delays, limiting the ability to use data more recent than 2011. The same analysis of 2008-2009 data showed that antibiotic prescribing patterns were stable from 2008-2011 and that 31% of 2008-2009 outpatient antibiotic prescriptions were unnecessary. More recent data indicate that overall outpatient antibiotic courses dispensed have remained stable from 2011 through 2013 with 877 and 849 antibiotic courses dispensed per 1000 population, respectively.35 The rate of antibiotic prescriptions in this study is just more than half the rate of antibiotics dispensed in outpatient settings in the United States in 2011.4,35 It is not known how these rates in this study based on data from 2010-2011 compare or reflect the rates of prescribing in 2016.
Fifth, antibiotics prescribed at urgent care and retail clinic, federal facilities, hospital discharges, telemedicine encounters and for long-term care residents are not included in NAMCS/NHAMCS. In addition, visits to physician assistants and nurse practitioners are not systematically sampled and may be underrepresented.
Sixth, these data reflect medication mentions, not antibiotics dispensed or consumed, and delayed prescriptions cannot be differentiated from standard prescriptions. Appropriate antibiotic selection was not addressed, but a separate study is under way to identify opportunities to improve selection.
In the United States in 2010-2011, there was an estimated annual antibiotic prescription rate per 1000 population of 506, but only an estimated 353 antibiotic prescriptions were likely appropriate, supporting the need for and establishing a goal for outpatient antibiotic stewardship.
Corresponding Author: Katherine E. Fleming-Dutra, MD, Centers for Disease Control and Prevention, 1600 Clifton Rd, MS A-31, Atlanta, GA 30329-4027 (firstname.lastname@example.org).
Author Contributions: Drs Fleming-Dutra and Hicks had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Fleming-Dutra, Hersh, Hicks.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Fleming-Dutra.
Critical revision of the manuscript: Hersh, Shapiro, Bartoces, Enns, File , Finkelstein, Gerber, Hyun, Linder, Lynfield, Margolis, May, Merenstein, Metlay, Newland, Piccirillo, Roberts, Sanchez, Suda, Thomas, Woo, Zetts, Hicks.
Statistical analysis: Shapiro.
Obtained funding: Hyun, Zetts.
Administrative, technical, or material support: Fleming-Dutra, Hyun, Zetts, Hicks.
Study supervision: Fleming-Dutra, Hersh, Hicks.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr File reports receiving grant funding from Cempra and Pfizer and personal fees from Actavis, Melinta, Merck, MotifBio, Nabriva, and Tetraphase. Dr Lynfield reports being co-editor for a book published by Wiley-Blackwell on Infectious Disease Surveillance for which royalties are donated to the Minnesota Department of Health. Dr Merenstein reports giving expert advice to Proctor and Gamble, Bayer, and Nestle. Dr Newland reports receiving grant funding from Pfizer and personal fees from RPS Diagnostics. No other disclosures are reported.
Funding/Support: This project was made possible through a partnership with the Centers for Disease Control and Prevention (CDC) Foundation. Support for this project was provided by Pew Charitable Trusts.
Role of the Funder/Sponsor: The CDC participated in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Pew Charitable Trusts participated in the interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Pew Charitable Trusts sponsored in-person and telephone author meetings and supported some author travel to in-person meetings.
Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC or the Department of Veterans Affairs.
Additional Contributions: We thank Tia Carter, MS, Allan Coukell, BSc Pharm, and Elizabeth Jungman, JD, MPH (all from Pew Charitable Trusts), for their assistance in convening author meetings. We also thank Allan Coukell, BSc Pharm (Pew Charitable Trusts), for his review of the manuscript. The above named individuals did not receive any compensation for their role in the study.
Create a personal account or sign in to: