Office-based physicians in the United States write approximately 1 million prescriptions for antibiotics annually. About one half of these prescriptions are inappropriate because they are used for conditions caused by viruses, such as colds and chronic coughs, according to the Centers for Disease Control and Prevention (CDC).¹ To promote awareness of proper antibiotic use, the CDC, the U.S. Food and Drug Administration (FDA), and major national health organizations have launched a new public health campaign, Get Smart: Know When Antibiotics Work.² The FDA also is requiring new labeling on systemic antibiotics, beginning in February 2004, advising that such drugs should be used only to treat bacterial infections and encouraging physicians to counsel patients about proper antibiotic use, including taking medications exactly as directed.³

Antimicrobial resistance hinders our ability to treat infections such as otitis media, sinusitis, and pneumonia.^4,5 Two studies have documented a decline in antibiotic prescribing in office-based practices, suggesting increased awareness of inappropriate use.^6,7 Between 1989 to 1990 and 1999 to 2000, antibiotic prescriptions for children declined 47 percent⁶; between 1991 to 1992 and 1998 to 1999, the proportion of office visits involving a prescription for an antibiotic, including those for acute respiratory tract infections, significantly decreased among children (from 33 to 22 percent; P<.001) and adults (from 13 to 10 percent; P<.001).⁷

Obstacles remain, however. Despite an overall decline in antibiotic prescriptions, broad-spectrum agents are still overused. In one study,⁷ prescriptions for broad-spectrum antibiotic agents were found to have increased from 23 to 40 percent in children and from 24 to 48 percent in adults between 1991 and 1999. Strong predictors of the prescription of a broad-spectrum antibiotic were physician specialty and practice region, with internists in the northeast and south using broad-spectrum agents in 76 percent of the regimens they prescribed. These antibiotics were less likely to be prescribed if the patient was black, a member of a health maintenance organization, or uninsured.⁸

Physicians cite diagnostic uncertainty, time pressure, and patient demand as the primary reasons for their tendency to overprescribe antibiotics.² To determine how patient demand affects prescribing behavior, Scott and colleagues⁹ analyzed data from nearly 300 diectly observed outpatient visits for acute respiratory tract infection at 18 family practices in Nebraska. Antibiotics were prescribed in 68 percent of the visits. However, according to CDC guidelines, 80 percent of those antibiotic prescriptions were unnecessary.

Patient behaviors that led to a prescription for antibiotics fell into three categories: explicit requests, presentation of chief complaints, and appeals to nonmedical circumstances. Patients were significantly more likely to receive an unnecessary antibiotic by portraying their illness as severe (P<.001) or by mentioning a prior positive experience with an antibiotic (P<.001). Lack of knowledge and misguided beliefs about antibiotics may explain why patients pressure physicians. In a recent study,¹⁰ 58 percent of patients were unaware of possible health dangers associated with antibiotic misuse (e.g., increased infection risk or colonization with resistant bacterial pathogens).

Steps that physicians can take to educate patients on judicious antibiotic use have been published.^11,12 By continuing this education and by prescribing more targeted therapeutic agents, the goal of appropriate antibiotic use is feasible. However, to employ targeted therapy, knowledge of emerging resistance patterns is essential. The new FDA labeling guidelines recommend that culture and susceptibility information be used when available to select or modify antibacterial therapy. In the absence of such data, empiric selection should consider local epidemiology and susceptibility patterns.³

Susceptibility patterns are monitored by surveillance databases that document resistance rates, identify major resistant pathogens, correlate relationships between antibiotic use and resistance, and alter empiric therapy based on common pathogen susceptibility. The National Nosocomial Infections Surveillance System collects data from more than 300 hospitals,¹³ and state health departments also frequently report such data to the CDC. The Antimicrobial Resistance Management (ARM) Program, Focus Technologies, Inc., SENTRY, and the Alexander Project have established independent databases to track antimicrobial susceptibility.

For example, the ARM Program has shown that pneumococcal isolates are more susceptible to ceftriaxone than to cefotaxime, especially in the southeastern United States.¹⁴ Knowing about this difference in resistance can help physicians select more targeted therapy which, in turn, helps conserve the effectiveness of antimicrobial agents within a community.

It is critical to educate patients about the appropriate use of antibiotics and to continue to improve databases for antibiotic resistance patterns so that physicians can prescribe targeted therapy. Only in these ways will we stop the overuse of antibiotics and emergence of multidrug-resistant bacteria.