Pediatricians have become familiar with the growing problem of bacterial resistance to antibiotics, especially as it relates to pneumococcal disease, including acute otitis media (AOM).^1,2 Otitis media is the most common diagnosis during an ill visit to a pediatrician. More than 70% of children have had at least one middle ear infection by the age of 3 years and 33% have had three or more episodes.³ The incidence of AOM is increasing, as evidenced by the doubling of prescriptions written for AOM in the last decade from 12 million in 1980 to 24 million in 1992.⁴ This increase in antibiotic usage is likely due to a number of factors, including increase in real disease, due in part to widespread day care attendance,^5,6 improved access to care,^7,8 and overprescription of antibiotics.

Increased prescription of antibiotics may be contributing to the emergence of bacterial resistance.^4,9 Hofmann et al¹⁰ found a higher incidence of infection with resistant pneumococci among white and suburban children and postulate that this finding may be due to an increased use of antimicrobial agents in an affluent population. Zenni et al¹¹ also showed an increase in the incidence of unresolved otitis media associated with nasopharyngeal colonization with penicillin-resistant pneumococci.

Many pediatricians have had the experience of parents demanding antibiotics for illnesses such as viral upper respiratory infections, nonspecific diarrhea, or sore throats. However, recently there has been increasing discussion in the print media about the dangers of antibiotics (Newsweek, March 28, 1994:47-51; Time, September 12, 1994: 62-69, Consumer Reports, July 1995:492-493; Mothering, Fall 1992:45-49), and some parents have begun to question the necessity of antibiotic therapy.

We undertook this study to determine parents' range of knowledge and understanding of antibiotics and to determine the experience of pediatricians with respect to prescribing oral antibiotics. Specifically, we were interested in parents' experiences and concerns regarding antibiotics, their knowledge of the indications for antibiotics, whether there was antibiotic use without physician knowledge, and whether parents were concerned that their children had received antibiotics unnecessarily. We wanted to compare parents' views with pediatricians' interpretation of parents' understanding and concerns.

METHODS

During the study, a research assistant approached one parent from each family before the child's visit with the physician and asked the parent to participate in a study about his or her opinions and experiences with antibiotics for children. Parents were interviewed by the research assistant in the waiting room and were asked questions related to demographic information and about their experiences with antibiotics. Most questions were yes/no or Likert scale in type. Several open-ended questions were also included about parents' perceptions of possible antibiotic side effects and illnesses that required an antibiotic prescription. Sample questions included: "Tell me how frequently you believe that antibiotics are helpful in treating these illnessesear infection?" (possible responses, always, sometimes, never, don't know); and "Have you ever felt that any of your children needed an antibiotic when the doctor did not prescribe one?" (possible responses, yes or no, and if yes, how frequently, one to three times, four to six times, more than six times). ² analysis was used to analyze categorical variables. The study was approved by the Human Studies Committees of Boston City Hospital and Boston University School of Medicine and also by appropriate committees at each of the practices.

In a second, related survey, 100 pediatricians across Massachusetts were asked about their opinions regarding parents' views on antibiotics, using a mailed questionnaire. We chose names of physicians from the Fellowship Directory of the American Academy of Pediatrics, selecting one pediatrician from each town, and choosing the first, alphabetically, excluding resident fellows, emeritus fellows, and subspecialty fellows until 100 physicians had been chosen. The exclusion criteria served to select a sample of practicing general pediatricians. Each physician was mailed a 2-page questionnaire and was requested to complete it and return it in a self-addressed, stamped envelope. The questions contained in the physician survey dealt mostly with physician perception of parents' concerns and were designed to mirror those in the parent survey. Sample questions included: "How many times in the past month has a parent requested an antibiotic for his or her child when you did not feel one was indicated?"; and "How many times in the past month has a parent requested a specific antibiotic or requested that you prescribe a different one than you were going to prescribe?". Questions related to how often an event had occurred in the past month were quantified as follows: never; one to three times; four to six times; seven or more times. ² analysis was used to analyze categorical variables. The study was approved by the Human Studies Committees of Boston City Hospital and Boston University School of Medicine.

RESULTS

Parent Survey

The demographics were identical in the two private practices (PP) and have been combined (Table 1). The parents from the PP were largely white (84%) and had completed college (81%). They were older than the health center parents, had a higher family income, and were more likely to have commercial insurance for their children. The parents in the CHC were mostly black (80%) and had not completed college (91%).

Table 1. Demographic Characteristics of Parents [View Table]

The majority of parents (78%) had experience giving their children oral antibiotics. Fifty-three percent had done so in the past 6 months and 73% in the past year. With respect to the most recent antibiotic that parents had given any of their children, 89% (262/296) were satisfied with the ability of the antibiotic to cure their child's illness, and 78% said they were able to give their child every single dose. The responses to these questions did not differ between parents from the PP and the CHC.

When parents were asked whether they were worried that any of their children were receiving too many antibiotics, 29% answered yes (32% in the PP compared with 20% in the CHC, P = .054). Eighty-five percent of parents believed that there could be problems with receiving too many antibiotics. In an open-ended question in which parents were asked to list potential problems with receiving too many antibiotics, 55% mentioned immunity or resistance. Ineffectiveness was mentioned by 15% of parents, and other responses such as allergies, vomiting, and diarrhea were listed by <10% of parents. Despite these concerns, 18% of parents said they had given an antibiotic at home before consulting a physician.

We asked parents if they believed that antibiotics were helpful in treating a variety of illnesses (Table 2). Most parents answered that antibiotics were always or sometimes helpful in ear infections and throat infections, but a large number also responded that antibiotics were useful in treating colds, coughs, and fever. More parents in the PP indicated that antibiotics were helpful for ear infections (95%) and throat infections (87%) than parents in the CHC (88% and 71%, respectively). More parents from the CHC believed that antibiotics were useful in treating colds (59%) than did parents from the PP (23%). Fifty-eight percent of parents, overall, responded that antibiotics were useful for cough and fever, and responses for these symptoms did not differ between the PP and CHC.

Table 2. Parents' Understanding of Antibiotic Use [View Table]

Table 3 describes parents' interactions with physicians around antibiotic prescriptions. When parents were asked "Has your child ever needed an antibiotic when the doctor did not prescribe one?" 14% said yes. More parents from the CHC (22%) responded affirmatively to this question than parents from the private practices (12%, P = .014). Ear infections (32%) and colds (26%) were the most common reasons for parents to believe that their child required an antibiotic.

Table 3. Parent/Physician Interaction Concerning Antibiotic Issues (%) [View Table]

Parents whose children had received antibiotics in the past were asked whether they believed that an antibiotic had ever been prescribed unnecessarily and whether they had ever requested that a physician prescribe a specific antibiotic. Significantly more parents from the PP (12%) believed that an unnecessary antibiotic had been prescribed, in comparison with the CHC families (3%, P = .019). Again, the most common reasons were ear infections (50%) and colds (10%). Thirty percent of parents had at some time requested that their child's physician prescribe a specific antibiotic (PP = 34%, CHC = 19%, P = .021). Eighty-two percent of these parents said that their doctor had prescribed the antibiotic requested (PP = 83%, CHC = 77%, P = .783). Forty-seven percent of parents from the PP had read an article about the use of antibiotics in children compared with 19% of parents from the CHC (P = .000).

Parents were given a list of five antibiotic factors (cost, dosing schedule or how many times a day the antibiotic is given, side effects, strength or effectiveness, and taste) and asked to choose the two that they believed were the most important. Parents from both the PP and the CHC selected side effects (82%) much more frequently than any other factor. Forty-one percent of parents believed that the strength or effectiveness of an antibiotic was important, 37% chose taste, and 33% thought that the dosing schedule was important. Cost was chosen least frequently (7%), probably because most parents reported that they had insurance that covered part or all of the cost of medications (see Table 1).

Physician Survey

Table 4. Demographic Characteristics of Physicians (N = 61) [View Table]

We asked the physicians if they believed that any parents in their practice were worried that their children were receiving too many antibiotics. Fifty-eight percent of pediatricians reported that some, many, or most of the parents were worried that their children were receiving too many antibiotics. Seventy-one percent of physicians said that at least four times in the past month (23%four to six times in the past month, 48%seven or more), a parent had requested an antibiotic when the physician did not believe one was indicated (Table 5). Thirty-five percent said that at least occasionally they went along with the request. Sixty-one percent of pediatricians reported parents requesting a specific antibiotic at least four times in the past month (31%four to six, 30%seven or more), and physicians were more likely to agree to parents' requests in this situation (30% responded often or most of the time). Fifty-four percent of physicians reported a parent requesting an antibiotic over the phone at least four times in the previous month (26%four to six, 28%seven or more), but 81% of physicians rarely or never agreed to this request. There was no difference among physicians, based on length of time in practice, in their willingness to prescribe an antibiotic when they believed it was unnecessary or to prescribe a specific antibiotic requested by parents.

Table 5. Physician Responses to Questionnaire (N = 61) [View Table]

Physicians were asked to choose two of five antibiotic factors that they believed were most important to parents (cost, dosing schedule, side effects, effectiveness, taste). Pediatricians were most likely to believe that parents were concerned about dosing schedule (63%). Fifty-three percent believed parents thought effectiveness was important, 37% chose cost, and 32% responded that parents were concerned about taste. Only 15% of physicians believed that parents were concerned about side effects. Twenty-seven percent of pediatricians believed that children received the entire 10-day course of antibiotics as prescribed always or most of the time.

DISCUSSION

We are unaware of any other data published in peer-reviewed journals about parent and physician beliefs, concerns, and experiences with antibiotics. Our parent survey suggests that parents have misconceptions about appropriate indications for antibiotics and often give antibiotics without physician knowledge. It is possible that parents misunderstood the questions about specific signs of illness and indications for antibiotics. For example, if they had prior experience with a child having a cold, had sought care, and a physician had made the diagnosis of an ear infection, parents may have responded that antibiotics are indicated in treating a cold. In some regards, this scenario illustrates the complexity in understanding what parents know and believe about antibiotics. With respect to the physician survey, physicians reported that parents request antibiotics when they are not indicated. Both surveys suggest that parents have impact on which antibiotics are prescribed.

Prescriptions for AOM are likely to reach 30 million in 1996, an increase of 150% since 1982.⁴ The reasons for this change include an increase in the incidence of AOM, improvement in access to care, and possible unwarranted use of antibiotics. The number of young children attending day care has increased significantly over the past decade. Day care attendance is associated with frequent respiratory infections, some of which are accompanied by AOM.^5,6 Several studies have shown that access to care for children has improved with approximately 90% of parents indicating that their children have a regular source of care.^7,8 We believe that improved access to care is accompanied by increased diagnoses of minor infections and increased oral antibiotic use. Unnecessary prescription of antibiotics also occurs. How much is unclear, although in our survey, one-third of physicians reported at least occasionally prescribing antibiotics when they were not indicated and 19% at least occasionally prescribe antibiotics over the phone. Which of these three factorsincrease in real disease, improvement in access to care, or unwarranted antibiotic useis contributing most to the increase in antibiotic prescriptions is unknown.

Anecdotal information from our colleagues in PP and those who practice at Boston Medical Center suggests that parents exert pressure on physicians to dispense antibiotics. Marcy,¹⁵ in an address at a meeting on improving antibiotic use practices sponsored by the National Managed Health Care Congress, stated that the primary cause of overuse of antibiotics is parental coercion. Although we would like to believe that clinicians prescribe antibiotics only when indicated, it is often difficult in a busy practice or emergency room setting to explain to a parent why an antibiotic is not indicated. It is less time-consuming to write a prescription than to engage in a lengthy discussion with parents about the natural history of an upper respiratory infection, diarrhea, or sore throat. In addition to parental coercion and time constraints, concerns of malpractice and litigation may also contribute to physician prescribing practices. Finkelstein and Platt¹⁶ recently presented data on nurse-practitioner and physician management of febrile children in an office setting and showed wide variation in prescription of antibiotics among clinicians in the study. Nonspecific fever in young children is always a cause for concern because of the risk of meningitis. The variation demonstrated by Finkelstein and Platt¹⁶ may reflect concern of litigation or simply variation in prescribing patterns.

This study represents an initial step in understanding the complex relationship between parents and physicians as it relates to antibiotics. There are some limitations to this study. First, our population represents only 400 parents from two distinct socioeconomic groups and 61 pediatricians in Massachusetts; our results may not be able to be generalized to different populations of parents or physicians. Second, 39 physicians did not respond to the questionnaire. It is possible that their responses would have been different from those of the responders. Third, we relied on self-report. It is not known how accurately parents and physicians recall actual experiences with antibiotics. In studying parents' attitudes about antibiotics, other methods, such as focus groups could be used to assess opinions and beliefs. We previously conducted focus groups with adolescents about their understanding of acquired immunodeficiency syndrome and found them useful in exploring a broad range of responses from participants and in uncovering unexpected information not elicited during structured interviews.¹⁷ With respect to physician prescribing patterns, a prospective observational study would be helpful in determining if actual practice is similar to self-report.

Physician prescribing patterns are influenced by physician and parental knowledge about antibiotics. The literature on antibiotic compliance sheds some light on how prescribing patterns could be changed. Many studies have documented an increase in parent compliance with antibiotic treatment using various written, visual, and telephone reminders for parents.¹⁸ Maiman et al²¹ have also shown that a continuing medical education intervention aimed at physicians to improve their compliance-teaching strategies has a positive effect on parent compliance. Recently, Wall et al²² reported the effectiveness of a brief office-based intervention, delivered during a well-baby visit, in decreasing maternal smoking. Whether any of these strategies would be useful in altering the parent-physician interaction to allow more appropriate prescription of antibiotics is unknown.

CLINICAL IMPLICATIONS