Reducing Inappropriate Oral Antibiotic Use: A Prescription for Change

EXTENT OF INAPPROPRIATE ORAL
ANTIBIOTIC USE

It is difficult to estimate the percentage of oral antibiotic use in children that is inappropriate. Data from the National Ambulatory Medical Care Survey (NAMCS)¹⁰ suggest that antibiotic use will reach 128 million doses in 1998, an increase from 86 million prescriptions in 1980. Not only has there been an increase in overall antibiotic use, but there has also been a shift in the leading diagnoses for which antibiotics are prescribed. In 1980, the most common diagnosis for which antibiotics were prescribed included upper respiratory tract infection (URTI), followed by otitis media, bronchitis, pharyngitis and acne. In 1992, otitis media led the way, followed by URTI, bronchitis, pharyngitis, sinusitis (added to the reporting system in 1985), and acne. Based on the NAMCS data, we estimate that approximately 30 million prescriptions will be written for AOM this year, an increase from 12 million prescriptions in 1980.

Recently Gonzalez and others¹¹ reported that each year 12 million prescriptions are written for adults with colds, URTI, and bronchitis—diagnoses for which antibiotics are not clearly indicated. In a follow-up study from the same data set that focused only on children, Nyquist and others¹² found that antibiotics were prescribed to 44% of children with common colds, 46% with URTIs, and 75% with bronchitis. Although pediatricians were less likely than nonpediatricians to prescribe antibiotics for these diseases, because more children are seen by pediatricians, they account for the majority of prescriptions. In a case-scenario questionnaire, Schwartz¹³ reported that 71% of family practitioners and 53% of pediatricians would immediately prescribe antibiotics for infants with mucopurulent nasal discharge that lasted 1 day. Only about one third of family practitioners and pediatricians waited until day 7 of illness to prescribe antibiotics for these same symptoms. The general consensus is that 10 days of symptoms are necessary before assuming a patient has sinusitis.¹⁴

There are many complexities in determining the extent of inappropriate oral antibiotic use. For example, the increase in number of prescriptions for AOM is attributable to at least three factors. First, access to care has improved dramatically during the past 20 years. Close to 95% of parents now report access to care for their children.¹⁵ Second, more children attend day care at an early age.¹⁶ Day care increases the risk of URTI,¹⁷ which contributes to the development of AOM. Third, making the diagnosis of either AOM or sinusitis is difficult, particularly in young children. It is unclear when either AOM or sinusitis are coded as a diagnosis, if the diagnosis was correct, or the physician sensed parental desire for antibiotic and recorded a diagnosis for which antibiotics are usually prescribed. There may be temporal trends in parental pressure to dispense antibiotics.

DIAGNOSTIC INDICATIONS FOR ANTIBIOTICS

Recently this journal published a supplement¹⁸promoting the judicious use of antibiotics, detailing diagnostic indications for antibiotics for the following diseases: otitis media; pharyngitis; and acute sinusitis. URTI, cough illness/bronchitis, and the common cold were also discussed. Although the information in these articles is helpful, we believe that the emphasis on physician education, without acknowledgment that parental pressure figures prominently in the use of oral antibiotics, addresses only one side of the equation. As mentioned above, Nyquist¹² and Schwartz¹³ reported that many of us do not follow precise diagnostic indications and write prescriptions for illnesses such as colds and coughs. However, there are many subtle complexities to the clinical encounter. For example, many of us have seen a child on day 5 to 7 of illness with nasal discharge and cough, whom we suspect may have or be developing sinusitis. Is the proper action to observe this child for 3 more days, perhaps requiring an additional visit, or should a clinician prescribe an antibiotic, knowing that for some children it may not be indicated? Diagnostic certainty is often not possible, unless invasive and expensive diagnostic tests are performed. A clinician is often asked to balance the needs of a family—a parent's desire to return to work and a child's need to return to day care or school—with scheduling another visit for the family. We live in a microwave society. Parents are busy and want their children well yesterday.

As mentioned above, making the diagnosis of AOM is complex. In young children, visualizing the tympanic membrane and determining mobility, placement of landmarks, and color are critical.¹⁹ The diagnosis of AOM can be difficult to make in the crying, febrile infant. Dowell and others²⁰ highlight the importance of categorizing otitis media as AOM or otitis media with effusion (OME). They also point out that antimicrobials are indicated for the treatment of AOM but not for OME unless the effusion is prolonged. These recommendations are helpful. Unfortunately many children are seen when they have mild signs or symptoms of URTI and it is difficult to determine whether the crying child has AOM or OME associated with a cold.

PATIENT ROLE IN DISPENSING ANTIBIOTICS

Parents misunderstand appropriate indications for antibiotics, although they have very strong opinions about them. In a survey of 400 parents, we found that parents believed that antibiotics were always or sometimes required for throat infections (83%), colds (32%), cough (58%) and fever (58%).²¹ In our study comparing single-dose intramuscular antibiotic to standard 10-day oral therapy for the treatment of AOM, 85% of 648 parents at the time of enrollment preferred single-dose treatment.²² At the conclusion of treatment, 90% of parents whose children received intramuscular therapy stated they would prefer it in the future as well as 76% who received standard therapy. We also found that parents influence the choice of specific antibiotics. In a survey of 1000 pediatricians, 33% of respondents indicated that seven or more times in the previous month parents requested a specific antibiotic or a different one than the pediatrician recommended.

It is clear that parents pressure pediatricians to dispense antibiotics. It remains unclear how often we comply with these requests. Data from our surveys suggest that one third of pediatricians at least occasionally dispense antibiotics when they are not indicated.²¹ The data from the NACMS certainly confirm that pediatricians dispense antibiotics inappropriately.¹²It is uncertain what percentage of antibiotic overuse is because of parental pressure.

The majority of work on the influence of patients on prescribing patterns has been done in England. Macfarlane and colleagues,^23-25 in a series of elegant studies conducted in primary care sites, have shown that three quarters of adults with upper respiratory tract illness receive antibiotics even though their general practitioners assess that antibiotics are indicated in only one fifth of cases. Patients who do not receive antibiotics are more likely to be dissatisfied and reconsult their physician. Nonclinical factors, particularly pressure from patients to prescribe antibiotics, are commonly cited by clinicians as important reasons for the excess use of antibiotics.

A number of studies have also been conducted in the United States that focus on physician-patient interaction. Vinson,²⁶ in a report from the Ambulatory Sentinel Practice Network, found in a study of 1398 children that if physicians perceived that parents expected a prescription for an antibiotic, the likelihood that a diagnosis of bronchitis would be made doubled and the likelihood that an antibiotic would be prescribed tripled. Hamm²⁷ reported that 65% of 113 adult patients with respiratory infection expected antibiotics. Antibiotics were prescribed for 77% of adults who wanted them, in contrast to 29% who did not.

A PRESCRIPTION FOR CHANGE

To reduce the use of inappropriate oral antibiotics, pediatricians will need to improve their diagnostic skills, understand therapeutic options, particularly with respect to AOM, and reeducate families about the risks and benefits of antibiotics.

Because AOM is the most common indication for antibiotics in children, making the diagnosis correctly is important, even before the consideration of whether to treat or not to treat a child with AOM. We believe that if we diagnose AOM more carefully, and do not use it as an excuse to dispense antibiotics, the number of prescriptions written for AOM will decline. It is only after we have improved our diagnostic skills, that we need to consider whether antibiotics are indicated for all children with AOM.

The issue of whether all children with AOM should be treated with antibiotics cannot be avoided in the 1990s.^28-30 With rising rates of bacterial resistance, it is possible that not treating children with antibiotics will result in a reduction in bacterial resistance, however, at what cost? It is clear that the treatment of AOM with antibiotics does little for the acute problem of pain. In a metaanalysis of all studies examining pain as an outcome, Del Mar and others³¹ found that only 14% of patients who are not treated with antibiotics will have ear pain beyond 24 hours of presentation. They estimate that 17 children must be treated at first presentation to prevent 1 child experiencing pain after 2 to 7 days. With respect to serious complications, such as mastoiditis, the risk among untreated children in developed countries has become rare.³¹ Experts are trying to define a group of children at low risk of serious sequelae who, at least initially, could be treated with symptomatic therapy rather than antibiotics.^{28 30-33}

As primary care physicians, and a peer-reviewer (H.B.) for many commentaries submitted to this journal that call on primary care pediatricians to do more during routine health care visits, we are sensitive to adding to the burden of seeing patients in the 1990s. There is certainly a perception that primary care physicians are already being asked to do more during routine visits. Many patients are in managed care health systems that increase our administrative workload. Is educating families about antibiotics as important as talking with families about violence in society, divorce, depression, abuse or discipline? We think so, but appreciate that other physicians may disagree. Virtually every child in the United States will receive antibiotics. Virtually all parents in the United States will see a physician because their child has a minor acute infectious illness and will wonder what role antibiotics play in helping their child recover. Finally, there are data that indicate when antibiotics are used appropriately, antibiotic resistance patterns can be reversed. In Finland, after a 2-year campaign to treat only group A streptococcal infections and not viral pharyngitis with penicillin, rather than a macrolide, the rate of pneumococcal resistance to erythromycin decreased by 53% (19% to 9%).³⁴

The campaign to educate families about antibiotics does not need to be carried out by physicians alone. We are impressed with the positive role that the media has played in this effort. In newspapers and parent magazines, the issue of appropriate indications for oral antibiotics and the role that antibiotics play in the emergence of bacterial resistance, has been responsibly discussed. Anecdotal information from colleagues and our experience in practice and with friends suggests that an increasing number of parents want to discuss the role of antibiotics in the treatment of minor infectious illnesses and limit their childrens' exposure to antibiotics. Recently, the AAP, the Centers for Disease Control and Prevention, and the American Society of Microbiology released a pamphlet³⁵ for parents that addresses the differences between viruses and bacteria, appropriate indications for oral antibiotics, and the role that overuse of antibiotics plays in the emergence of bacterial resistance. Currently, we are involved in producing an educational video promoting the judicious use of oral antibiotics.³⁶ The content of the video is based on focus groups conducted with both parents and physicians, and will include animation depicting the differences between viruses and bacteria, and have parents and physicians discussing appropriate indications for antibiotics.

Educating parents about antibiotics can impact on pediatric practice in two ways. First, an educated parent may pressure us less to dispense antibiotics when they are not indicated and second, educated parents may even question us as to why antibiotics are being prescribed in certain circumstances. There are data^{37 38} suggesting that physicians sometimes dispense antibiotics when they incorrectly perceive that parents want them. By educating families, we may be able to empower them to appropriately question us about antibiotic use. As mentioned previously, patient-mediated intervention is a proven effective strategy in changing physician behavior.⁹