CDC Grand Rounds: Getting Smart About Antibiotics

Alicia Demirjian, MD; Guillermo V. Sanchez, MPH; Jonathan A. Finkelstein, MD; Shari M. Ling, MD; Arjun Srinivasan, MD; Lori A. Pollack, MD; Lauri A. Hicks, DO; John K. Iskander, MD

Disclosures

Morbidity and Mortality Weekly Report. 2015;64(32):871-873. 

In This Article

Introduction

Each year in the United States, approximately two million persons become infected with antibiotic-resistant bacteria, at least 23,000 persons die as a direct result of these infections, and many more die from conditions complicated by a resistant infection.[1] Antibiotic-resistant infections contribute to poor health outcomes, higher health care costs, and use of more toxic treatments.[2] Although emerging resistance mechanisms are being identified and resistant infections are on the rise, new antibiotic development has slowed considerably.[2]

Inappropriate antibiotic prescribing is an important and modifiable contributor to antibiotic resistance and is a problem in all health care settings.[1] Inappropriate antibiotic use contributes to excess health care costs, promotes antibiotic resistance, and contributes to preventable adverse drug reactions. Antibiotics cause approximately 142,000 adult emergency department visits annually for adverse drug reactions; almost four out of five of these visits are for allergic reactions.[3] Antibiotics also contribute to both health care- and community-associated Clostridium difficile infections, which are associated with considerable costs to patients and the health care system.[1,4] In 2009, approximately $10.7 billion was spent on antibiotic therapy in the United States, including $6.5 billion, $3.6 billion, and $526.7 million in the outpatient, inpatient acute, and long-term care settings, respectively.[5] The cost of antibiotic resistance to the U.S. economy is an estimated $20 billion annually in excess direct health care costs, with an additional $35 billion in lost productivity.[1]

Antibiotic prescribing must be tracked to understand and improve antibiotic use. Several data sources and surveillance systems have been employed to examine antibiotic prescribing in hospitals and the community. These include the National Ambulatory Medical Care Survey, the National Hospital Ambulatory Medical Care Survey, the National Healthcare Safety Network, claims data from health plans and insurance companies, and data from private vendors.[6] An accurate assessment of antibiotic prescribing, regardless of clinical setting, is important to identify opportunities to improve prescribing and maintain provider accountability.

An estimated half of antibiotic prescriptions given during pediatric ambulatory care visits are inappropriate, and over one quarter of adult prescriptions are for conditions for which antibiotics are rarely indicated.[6,7] Health care providers prescribed 262.5 million courses of antibiotics in 2011 (842 prescriptions per 1000 persons), and prescriptions per 1,000 persons vary markedly according to geography.[8] The highest prescribing states in 2011, Kentucky and West Virginia, had a rate more than twice that of the lowest prescribing state (Alaska). Why such variability exists is unclear, but this variability is unlikely to be explained by differences in population distribution and extent of infectious diseases.

Inappropriate antibiotic use is not limited to the outpatient setting. A recent evaluation of prescribing for inpatients in two specific scenarios (urinary tract infections in patients without indwelling catheters and treatment with intravenous vancomycin) identified that antibiotic use could have been improved in 37% of cases.[9] Frequency of antibiotic prescribing among inpatients varies considerably among hospitals. A recent study of 19 hospitals that had completed data validation and submitted antibiotic use data from one or more patient care settings, found threefold differences in usage rates among 26 medical/surgical wards.[9]

Visits for acute respiratory tract infections lead to more inappropriate antibiotic prescribing than visits for any other group of diagnoses. For example, antibiotic treatment for acute uncomplicated bronchitis is not recommended, and despite decades-long, widespread efforts to curb antibiotic prescribing, in 2010, 71% of all outpatient visits for this condition resulted in an antibiotic prescription.[10] Similarly, overprescribing for pharyngitis is common. Only 5%–10% of pharyngitis cases among adults are caused by group A Streptococcus, for which antibiotic treatment is recommended, yet antibiotics are prescribed for approximately 60% of ambulatory care visits for adult pharyngitis.[7] Outpatient antibiotic prescribing for children with acute respiratory tract infections has been decreasing since the mid- to late-1990s, but the rate of decline has slowed and might have reached a plateau.[11] Several factors have been hypothesized to have contributed to this decrease, including the increased use of pneumococcal conjugate and influenza vaccines, national education campaigns to promote appropriate antibiotic use, and increasing concern among both the general public and health care professionals about antibiotic resistance.

In addition to the problem of overuse, antibiotic selection is often inappropriate. Prescribers often choose second- or third-line antibiotics, which are typically broad-spectrum drugs, despite established clinical practice guidelines recommending more targeted agents. Overuse of broad-spectrum antibiotics (e.g., second- or third-generation cephalosporins, fluoroquinolones) is especially problematic because of their potential for increased selection of resistant bacterial populations and their role in treating serious infections. Among U.S. ambulatory care visits during 2007–2009, broad-spectrum antibiotics accounted for 74% of antibiotics prescribed to patients during visits for respiratory conditions.[7] Among hospitalized patients, 56% received an antibiotic during their stay and 30% received at least 1 dose of a broad-spectrum antibiotic.[9]

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....