Abstract
Misuse of antibiotics is largely responsible for the emergence of bacterial resistance. Children represent a subset of the population who frequently receive antibiotics. The objectives were to calculate the frequency of antibiotic prescriptions that do not comply with best practice recommendations in paediatrics primary care and to examine the thoughts and feelings of physicians and parents about antibiotic prescription and recommendations from the national health authorities. We included children admitted at the paediatric emergency room (PER) of the NANTES university hospital between June 2011 and October 2012 and who were under antibiotic drugs. Two independent experts evaluated the compliance with the national recommendations. Parents and general practitioner (GP) who prescribed the antibiotic before admission to PER were called to collect their thoughts and feeling about antibiotic prescription. The median age of the 88 included children was 2.8 years. The upper respiratory tract infection motivated the prescription of antibiotic in 59%. Seventy-six per cent of the prescriptions analysed were inappropriate, 72% of the antibiotics had a too broad spectrum, and one-third was not indicated. Ninety-one per cent of the interviewed parents thought that the antibiotic prescribed to their child was adequate. Among the 77 prescribing GP who were called, 33% agreed that they lacked time during consultation to explain to parents that no antibiotics were required. Antibiotic prescriptions were too often inadequate in this sample of children admitted in a French PER. Efforts have to be made with physicians and general public to optimize the antibiotic drug use.
Introduction
Growing bacterial resistance to antibiotics represents a threat to public health. Bacterial resistance is directly related to the misuse of antibiotics (Goossens et al., 2005). In recent years, a declining trend in prescription of antibiotics has been observed in most developed countries (Cosby et al., 2007; Goossens et al., 2005). In France, consumption of antibiotics decreased by 15–20% in primary care and by 10–15% in hospitals between 2003 and 2008. This is probably partly due to a national campaign led by the Ministry of Health from 2001 till 2007. However, since 2008, a new trend towards increasing use of antibiotics has been observed, placing France among the biggest consumers in Europe (Cosby et al., 2007).
Children represent a subset of the population who frequently receives antibiotics. During the first years of a child’s life, they will typically present with between 3 and 6 infectious episodes per year. Most often, these infections are caused by viruses. Craig and colleagues have shown that only between 5% and 10% of fever episodes are due to bacterial infections (Craig et al., 2010). One of the difficulties for doctors is to identify bacterial infections and to limit the use of antibiotics to those cases alone in order to preserve their efficacy.
The main objective of our study was to calculate the frequency of antibiotic prescriptions that do not comply with best practice recommendations in a sample of patients presenting to the PER of the Nantes University Hospital in France. The secondary objective was to examine the thoughts and feelings of physicians and parents about antibiotic prescription and recommendations from the national health authorities.
Method
We conducted an observational, prospective study from June 2011 to October 2012. We included patients aged 0–15 years who were admitted to the PER in NANTES Hospital and treated with an antibiotic started before the time of admission. Patient data included age and reason for consultation during which antibiotic was prescribed. Parents’ opinion about the current prescription of antibiotics for their child, and about prescription of antibiotics in France in general, were collected using a data sheet completed by parents during their stay in the paediatric emergency department. The questions concerning antibiotic prescriptions in general were in a multiple-choice format. Before participating, parents received written information and gave their verbal consent for their child to participate in the study. Data about the prescription – including diagnosis, type and dosage of antibiotic prescribed, prescription and results of diagnostic tests to justify antibiotic prescription, and known allergy to antibiotics – were provided by the physicians in the emergency department who examined the child. In a second phase, we conducted a telephone survey among primary health care physicians who prescribed antibiotics. We asked them questions about factors influencing their prescription in general and their opinion on the current recommendations concerning antibiotic prescriptions. There were six questions in a binary format (yes/no), and the physician was also able to make a free-text comment.
Compliance with recommendations was determined by two independent experts: one a university paediatric infectious disease specialist and the other a GP. In cases of disagreement, a third expert was asked to decide. In cases of inappropriate prescription of antibiotics, they classified the type of error as (i) antibiotic not indicated, (ii) choice of antibiotic inadequate, (iii) antibiotic dosage inadequate, (iv) antibiotic duration inadequate, (v) lack of biological or radiographical tests to confirm bacterial infection whilst indicated and (vi) diagnostic failure. Experts based their judgement on the French best practice recommendations (Agence Française de sécurité sanitaire des produits de santé, 2005, 2007, 2011).
Qualitative variables were described as percentages with 95% confidence interval (95%CI). Quantitative variables were described using median or mean, with interquartile range (IQR) or standard deviation. We assessed the degree of agreement between the two experts by calculating the κappa coefficient interpreted with the Landis and Koch scale (Landis and Koch, 1997). Statistical analyses were made with the Stata Statistical software version 11 (StataCorp. 2009).
Results
Population
On the 92 included patients, 4 were subsequently excluded because of missing data. The median age of the 88 remaining children was 2.8 years (IQR: 1.1–5.3). The sex ratio was 1. Symptoms motivating medical consultation with a physician at the first instance of consultation were fever in 37% (95%CI [27–47]) of the cases, respiratory symptoms in 34% (95%CI [24–44]), pain in 8% (95%CI [2–14]), gastrointestinal symptoms in 5% (95%CI [0–10]) and others causes in 15% (95%CI [7–23]). The physician consulted at the office was the usual family doctor in 72% of cases (95%CI [62–82]), another GP in 18% (95%CI [10–26]) and a paediatrician in 10% (95%CI [4–16]). The diagnoses made by the physicians prescribing the antibiotics before the patients’ admission to ER were upper respiratory tract infection in 59% (95%CI [49–69]) of the cases, lower respiratory tract infection in 25% (95%CI [16–34]), urinary tract infection in 8% (95%CI [2–14]), cutaneous infection in 6% (95%CI [1–11]) and enteritis in 2% (95%CI [0–5].
Antibiotic prescriptions
Amoxicillin, amoxicillin/clavulanate and cefpodoxime axetil were the most frequently prescribed antibiotics, accounting for 32% (95%CI [22–42]), 31% (95%CI [21–41]), and 26% (95%CI [17–35]) of the prescriptions, respectively.
Seventy-six per cent (95% CI [66–87]) of the antibiotic prescriptions were not compliant with the recommendations. The most frequent reason for non-compliance was the inadequate choice of antibiotic in 72% (95% CI [62–82]) of the cases, followed by non-indication of antibiotic prescription in 31% (95% CI [20–41]; Table 1). The agreement coefficient between experts was good (k = .79).
Reasons for non-compliance with recommendations; 67 prescriptions were non-compliant, some prescriptions had more than one reason.
GAS: group A Streptococcus
Parents and prescribing doctor’s thoughts
Of 88 parents answered the questionnaire concerning their feelings about antibiotic prescriptions, 91% (95% CI [85–97]) thought that the antibiotic prescription given currently to their children was useful, 51% (95% CI [40–62]) thought that antibiotic use in France was adequate, 31% (95% CI [21–41]) that it was large, 9% (95% CI [3–15]) that it was dangerous for the community because of emergence of bacterial resistance, 4% (95% CI [0–8]) that is was insufficient, and 5% (95% CI [0–10]) had no opinion. Sixty-seven physicians, who prescribed the antibiotics to the included children, were called to collect their thoughts on current antibiotic prescription recommendations. Twelve per cent (95% CI [4–20]) of the interviewed physicians agreed that the recommendations are difficult to access, 96% (95% CI [91–100]) were concerned by the recommendations, 29% (95% CI [18–40]) agreed that the recommendations are irrelevant to their everyday practice, 30% (95% CI [19–41]) agreed that the parent’s attitude or stress influences on their prescription of antibiotics, 33% (95% CI [21–45]) agreed that they lacked time during consultation to explain to parents that no antibiotics were required, and 31% (95% CI [20–42]) agreed that they lacked time to update their knowledge.
Discussion
Main results
Our study shows that antibiotics are still used inappropriately in paediatric primary care. More than three-quarters of the prescriptions were mismatched with current national recommendations available in 2012. The use of a too broad spectrum concerned 72% of the inadequate prescriptions. A minority of parents was aware of the over-consumption of antibiotic in France and that prescribing antibiotics could be dangerous for the community (resistance selection). A large majority of the interviewed GP was concerned by the recommendations but one-third argued that they lacked time to explain parents that antibiotic prescription was not necessary.
Strengths and limits
We analysed the antibiotic prescription in children who were admitted at PER. We could hypothesize that these children might have consulted because of failure of the antibiotic treatment due to an inadequate prescription and therefore we would have overestimated the frequency of inadequate prescriptions. Nevertheless, the most frequent reasons of inadequacy of antibiotic prescriptions were the use of a too broad-spectrum molecule and the unnecessary prescription of antibiotic, and not the prescription of an ineffective antibiotic for an actual bacterial infection. We also met some inclusion difficulties explaining our small effective and limiting the scope of our results. Despite our efforts to identify all eligible children, some children might have been missed during the vacation periods or night shifts. Nevertheless, to avoid this selection bias, we carefully explained to all physicians working at the PER that all children who were under antibiotic drug should be included and not only those for whom they had the feeling that the prescription was inadequate. The validity of our primary endpoint (compliance of the antibiotic prescription with the current recommendations) was ensured by a double independent evaluation with a good coefficient of agreement between experts. Moreover, the anonymization of records enhanced objectivity.
Interpretation of the results and comparison with existing literature
Non-compliance was represented mainly by an error in the selection of the appropriate antibiotic molecule (72%; CI 95% [60–83]). The broad-spectrum molecules (amoxicillin, clavulanic acid and cephalosporin) are prescribed most frequently, although they are not the most appropriate treatment for respiratory tract infections (Agence française de sécurité sanitaire des produits de santé, 2005). This problem represents an obstacle in reducing bacterial resistance, as shown in other studies in Europe and the United States (Fossum et al., 2013; Lee et al., 2014).
The non-indication for antibiotic prescription represents one-third (31%; 95% IC [20–43]) of non-compliant prescriptions. This is somehow higher than the 20% found by Craig et al. in a study conducted in Australia on 15,781 patients (Craig et al., 2010). Based on the difference of our study that analysed prescriptions made before admission to the hospital, Craig et al. analysed the prescriptions done at the hospital. We could then hypothesize that antibiotic prescription is more supervised within the hospital and therefore that antibiotic prescriptions were more often appropriate. In an Irish study, authors showed that 47% of the antibiotic prescriptions in primary care were not indicated. Respiratory tract infections represented 75% of these non-indicated antibiotic prescriptions.
In 30% of the inadequate prescriptions, the reason for inadequacy was the lack of diagnostic tests, especially of the rapid diagnostic test for pharyngitis. A study of Cohen et al. has shown the superiority of the rapid diagnostic test compared to the Mac Isaac test (Cohen et al., 2012). In France, an earlier study found that a third of pharyngitis patients were prescribed antibiotics without having a diagnostic test (Angoulvant et al., 2012). In the present study, physicians explain this by referring to a lack of time during consultations.
Ninety-one per cent of surveyed parents felt the doctor’s prescription was useful. This illustrates their confidence in their doctor. However, patients weren’t aware of the side-effects of antibiotics: only 10% thought antibiotics to be harmful. A survey on a large cohort in the United States showed that 36% of parents knew about bacterial resistance (Finkelstein et al., 2005). This shows that bacterial resistance is underestimated in the population. A third of the surveyed doctors admit they don’t comply with recommendations because of a lack of time. Yet they are key players in the patient’s education because of the confidence patients place in them. Surveys show that information and explanations given to parents are considered satisfactory and can even make them change their expectations and beliefs (Finkelstein et al., 2005; Welschen et al., 2004). Health education is therefore one of the necessary levers to change prescriber behaviours, but this can take some time and needs to be evaluated (Légaré et al., 2011).
Conclusion
As suggested by the World Health Organization’s guidelines about global strategy for containment of antimicrobial resistance, real efforts have to be made to improve the prescription of antibiotics and maintain their efficacy (WHO, 2001). Guidelines for physicians and education programs have a positive impact on the reduction of prescriptions (Angoulvant et al., 2012; Légaré et al., 2011), but they are not followed enough. Multifaceted interventions, where educational interventions occur on many levels, may potentially reduce the incidence of antibiotic resistance.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.