June 2002, Vol 24, No. 6 |
Original Articles
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Frequency and pattern of antibiotic prescrip-tion for common infections in a family practiceH C Chan 陳洪昭 HK Pract 2002;24:274-280 Summary Objective: To assess antibiotic use by patients prior to a general practice consultation and the antibiotic prescribing habits of doctors at the time of consultation in the setting of a Hong Kong general practice. Design: Cross-sectional clinical study. Subjects: 2231 patients were seen over a 2-month period (20/1/2000 to 19/3/2000) in two general practices in Hong Kong. Antibiotic use was assessed if the patients suffered from at least one of five common infectious diseases seen in general practice including upper respiratory tract infection (URTI), acute pharyngitis, acute bronchitis, acute diarrhoea, and uncomplicated urinary tract infection (UTI). 1160 (52%) patients suffered from at least one of these 5 common infectious diseases. Main outcome measures: The antibiotics used prior to the consultation and the antibiotics prescribed by a private family physician. Results: Of the 1160 patients presenting with at least one of the 5 infectious diseases, 892 (76.8%) were taking medications prior to their consultation. 96 (8.3%) were taking antibiotics. Antibiotic therapy was prescribed for 107 (9.2%) of the 1160 patients during the consultations. Beta-lactams were the most commonly prescribed antibiotic (69.2%), followed by tetracyclines (13.1%), macrolides (10.3%), nitrofurantoin (4.7%), cotrimoxazole (1.9%) and quinolones (0.9%). Conclusion: Self-initiated medications by patients is common in these selected general practices in Hong Kong. However, the number of patients requiring antimicrobial therapy was low, suggesting optimisation of antibiotic usage in the general practice setting in Hong Kong is feasible. Keywords: Antibiotics, resistance, optimisation, general practice 摘要 目的：評估病人看醫生前使用抗生素的情況和全科醫 生診症時開抗生素的習慣。 設計：橫切面形臨床研究。 對象： 2000年1月20日至3月19日的兩個月內，香港兩間全科診所的 2231個病人。評估患有以下五種傳染病中至少一種病患的病人使用抗生素的情況，包括上 呼吸道感染，急性咽炎，急性支氣管炎，急性肚瀉，和沒有併發症的尿道炎。共有 1160（52%）個病人患上至少一種常見的傳染病。 測量內容： 病人見醫生前使用的抗生素和家庭醫生處方的抗生素。 結果： 在患上最少一種傳染病的病患者當中，892（76.8%）個在來會診前已有服藥，其中96個（8.3%）正在服用抗生素，在1160名病人中共有107個在見醫生後被處方抗生素。 b-內胺（b-lactams）是最常處方的抗生素（69.2%），其次順序為四環素（Tetracyclines,13.1%），大環內酯（Macrolides, 10.3%），口夫喃妥因（Nitrofurantoin, 4.7%），（Cotrimoxazole, 1.9%）和口奎諾酮（Quinolones, 0.9%）。 結論：在香港，病人自行使用抗生素的情況在這兩間診所的所在地是很常見的。但是真正需要抗生素的病人數目是很少的。由此可見優化使用抗生素方面還有改善的空間。 主要詞彙：抗生素，抵抗力，最優化，全科醫療 Introduction Inappropriate antimicrobial use is a global problem.1-3 It results in antimicrobial resistance and increases the cost of health care.4 It also presents the physician and community with management dilemmas in terms of patient treatment and community resistance.5 The problem exists in both the community and hospital setting.6,7 Successful programmes have been described in the Hong Kong hospital setting to tackle this problem. A recent programme at the Queen Mary Hospital, Hong Kong formulated practice guidelines for antibiotic usage. Clinicians were encouraged to rationalise the use of empirical antibiotics and streamline antimicrobial agents. This measure was aimed at preventing the emergence of resistant organisms, such as vancomycin resistant enterococci, vancomycin resistant Staphylococcus aureus, and controlling the spread of extended spectrum beta-lactamase producing Enterobacteriaceae.8 In the community setting the abuse of antimicrobial therapy includes patient self-medication and the accumulation of antibiotics at home after doctor shopping. In order to investigate the abuse of antimicrobials in the community setting a study was conducted in two general practices in Hong Kong. The objective was to assess patient's antibiotic use prior to the consultation and the pattern of antibiotics prescribed by a family physician for five common infections seen in general practice. Methods Subjects and setting A descriptive study was conducted in two private general practice clinics of a single practitioner (one on Hong Kong Island and the other on Kowloon peninsula) over a two months period (20/1/2000 to 19/3/2000). Those patients who were diagnosed as suffering from one of the following infectious diseases: upper respiratory tract infection (URTI), acute pharyngitis, acute bronchitis, acute diarrhoea, or uncomplicated urinary tract infection (UTI), were recruited into the study. URTI was defined as an acute onset of sneezing, rhinorrhoea or nasal obstruction with or without fever or cough. Acute pharyngitis was defined as an acute onset of sore throat, with or without fever or cough. Acute bronchitis was defined as an acute onset of cough, with or without fever or sputum production and without clinical or radiological evidence of pneumonia. Acute diarrhoea was defined as a recent onset of more than 3 loose to watery stools over a 24-hour period. Uncomplicated UTI was defined as an acute onset of dysuria, frequency, or suprapubic discomfort. Method All the selected patients were specifically asked if they had solicited a previous medical consultation or taken any medication within the last week. They were also asked if any medication had been self-prescribed and how it had been obtained. If the patients admitted a history of prior medication for the same illness within a week, the medications were retrieved for identification. The exact names and regimens of any antibiotics were recorded. If the drugs could not be identified as an antibiotic the medication remaining was collected for antimicrobial testing. In the event that the medication was not retrievable, five millilitres of urine was collected for antimicrobial testing. At the end of the clinical consultation, the doctor decided whether the patient required antimicrobial therapy for the selected clinical syndromes. The indications for antibiotic prescriptions and the regimens were recorded for later analysis. Microbiological testing A bioassay in the form of a disc diffusion test was used to determine the presence of antibacterial activity in an unidentified medication or urine specimen. Medications were dissolved in 10 millilitres of sterile water. Ten microlitres of the solution or undiluted urine were added to a sterile paper disk (6 mm in diameter), which was placed on the surface of a Muller-Hinton agar streaked with 0.5 McFarland standard of Staphylococcus aureus (NCTC 6571). The plates were incubated at 35 Results There were 2231 outpatient consultations seen in the two private clinics during the study period (659 from the Hong Kong practice, and 1572 from the Kowloon practice). 1160 (52%) patients were diagnosed as having one of the five surveyed infections. There were 557 male and 603 female patients with an overall median age of 32 (range 0.5-87) years. The distribution of patients with the defined infectious diseases was shown in Table 1. Sources and prevalence of prior medications of the subjects were shown in Table 2. Eight hundred and ninety-two (76.8%) patients were taking some kind of medications prior to the consultation and 241 patients brought along their previous medications to the attending doctor, of which 57 patients' medications were identified to include an antibiotic. Of those 651 patients who could not produce the medications for identification, 6 patients could remember exactly they were given antibiotics and 11 patients were found to have been taking medications with antibacterial activity when tested for antibiotics. Of 180 patients who underwent urinary testing, antibacterial activity was demonstrated in 22 specimens. Overall, 96 (8.3%) out of 1160 patients suffering from at least one of the selected five infections had been taking antibiotics prior to consultations (Table 3). Seventy-one percent (68/96) of these patients were suffering from URTI (Table 4).
During the consultation antibiotics were prescribed for 107 (9.2%) of the 1160 subjects. The percentage of patients given antimicrobial therapy for their illness were URTI 1.4% (13), acute pharyngitis 36.1% (44), acute bronchitis 67.8% (40), acute diarrhoea 3.2%(2) and uncomplicated UTI 100% (8) (Table 5). Beta-lactams were the most commonly prescribed antibiotics representing 69.2% of the prescriptions (amoxicillin or amoxicillin-clavulanate 40; penicillin V 33; cephalexin 1). (Table 6) Paediatric patients (age
Discussion There is worldwide concern over the development of antibiotic resistance.1 This is not only a problem limited to the hospital setting. Resistant organisms previously unique to the hospitalised patients such as Methicillin-resistant Staphylococcus aureus are now also found in community acquired infections.6,7 The problem of antibiotic resistance within a community is related to both the prescribing habits of hospital doctors and the community's general practitioners. Magee et al conducted a study on urinary isolates involving 190 general practices. The investigators demonstrated a correlation between antibiotic resistance in urinary coliforms and the use of antibiotics by the general practices and suggested geographically localised effects from antibiotic use occurs in community.9 This was further exemplified in the study by Arason et al showing that community-wide use of antibiotics correlated with the carriage of penicillin-resistant pneumococci in Iceland.10 Although quite a lot is known about prescribing and antibiotic consumption patterns in the hospital setting in Hong Kong,8 knowledge of such patterns in the community setting is quite limited. Some of the factors which contribute to the difficulties in accessing such information include the fact that in Hong Kong the user directly pays for the service, which means that there is no centralised access to patient details. Also the family physician dispenses medication making centralised analysis of prescriptions impossible. This study is unique in that it is the first study to investigate the antibiotic prescribing patterns in the general practice setting in our locality and to assess antibiotic consumption prior to consultation. Five commonly encountered clinical syndromes including URTI, acute pharyngitis, acute bronchitis, acute diarrhoea and uncomplicated UTI were selected for detailed study. These diseases were chosen because many doctors may think that a high proportion of antibiotics were likely to be prescribed for these entities. They were also chosen for the fact that URTI and acute bronchitis are almost always viral in origin and rarely require antibiotic therapy, and there is growing evidence that antibiotics are often unnecessary for the management of most patients with sore throat.11-13 This study found that 76.8% of patients were taking medications prior to the consultation. Undoubtedly much of this is cultural with the frequent use of home remedies and herbal medicines. Indeed herbal and proprietary medications represented 10.2% of all the pre-consultation medications that were taken. In a study conducted by Lam et al, the investigators found that of 1068 people interviewed by a randomised telephone survey over a 2 weeks period, 32.5% of respondents were on self-medication. Of those respondents complaining of illness, Western medicine was used as often as Chinese medicine.14 A further study involving 563 university students in Hong Kong, the practice of self-medication was found to have a prevalence of 94%.15 This study found that 8.3% of patients with at least one of the designated infections were already on antibiotic therapy prior to the consultation. This is alarming given that only 0.7% of the patients, the eight patients suffering from UTI, had a definite indication for antibiotic therapy. The figure may actually be higher allowing for the fact that our method of antibiotic susceptibility testing may be insensitive in detecting the actual number of patients on antibiotic therapy. The detection method used would be affected by reduced activity of expired medications, undetectable levels remaining in urine, metabolism of the active drug and non-urinary routes of excretion. The situation of excessive use of antibiotics in the community is not unique to Hong Kong and tends to reflect a worldwide trend. In the USA over half the patients seen with colds or URTIs are prescribed antibiotics16 and more than one fifth of all antibiotic prescriptions are written for bronchitis or URTIs.16,17 In this practice antibiotics were prescribed for 9.2% of patients with the designated infections and this corresponds with an overall antibiotic prescription rate of 7.9% for all consultations. Recently, members of the CDC and the American Association of Paediatrics have drawn up guidelines for the judicious use of antibiotics for common paediatric respiratory infections. The guidelines state that antimicrobial agents should not be given for the common cold even in association with mucopurulent rhinitis. For sinusitis, antimicrobial agents are justified providing diagnostic criteria are fulfilled. A diagnosis of bacterial sinusitis requires prolonged nonspecific upper respiratory signs and symptoms of rhinosinusitis and cough without improvement for 10-14 days or more, severe upper respiratory tract signs, and symptoms including fever of 39 Beta-lactams represented 69.2% of antibiotics prescribed by the practitioner. The prescription rate for URTI in the current study is low. Even if pharyngitis is included within the entity of URTI as in some definitions,19 the prescription rate is 5.5% (57/1031). This is much lower than Australian general practitioner prescribing rates of 57% for URTI in urban areas. Australia has a high rate of antibiotic consumption, which is similar to that in other countries such as United Kingdom and the United States.20 The author believes that a reduced antibiotic prescription rate can be achieved if a family physician is willing to prescribe antibiotics judiciously and if he has a well-established relationship with his patients. Petursson described a reduction of up to 23% in the number of antibiotics prescribed in his general practice.21 Reductions of this magnitude may be difficult to achieve in some practices, however rationalisation of antibiotic prescribing has been shown to be realistically achievable through the use of antibiotic workshops.22 It has also been shown that bacterial resistance can be reduced if antibiotic prescribing habits are changed. In Finland, the isolation of macrolide resistant group A streptococci was approximately halved when macrolide use was reduced after a successful campaign.23 The result of this study may be biased by a small sample size conducted in two private clinics by one family physician. However, it serves as a pilot study to facilitate further investigation in this area. I foresaw difficulties if a large number of practitioners were recruited in the initial study as they might find difficulty in complying with guidelines involving study protocol. Another foreseen difficulty was the lack of resources to perform the relatively labour-intensive antimicrobial-susceptibility testing. Conclusion This study shows that the problem of self-medication and inappropriate use of antibiotics by patients is common in Hong Kong. However, more rational use of antibiotics is possible as demonstrated in my study. One of the limiting factors for promoting the appropriate use of antibiotic is patient expectations. Lam et al by means of a telephone-structured questionnaire also found that 76% of Hong Kong people expected a prescription after a medical consultation. This study also indicated that many general practitioners have been conditioned to comply with patient's demands or perceived demands. When patients were questioned about their most recent consultation, nearly 100% indicated that the consultation resulted in a prescription.24 Patient education and the establishment of prescribing guidelines for infections is an essential part of dealing with patients' anxieties and providing doctors with the knowledge and confidence of a reasonable and appropriate standard of antibiotic prescribing.25 It is our duty to our patients and society to try to reduce antibiotic use when bacterial infection is unlikely. Acknowledgement I would like to acknowledge Miss W L Woo and Mr Y C Lam for their laboratory technical support. Key messages
H C Chan, MBBS(HK), FRACGP, PDipID (HKU), FHKAM(Family Medicine) Correspondence to : Dr H C Chan, G/F, 27, Berwick Street, Sham Shui Po, Kowloon, Hong Kong. References
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
in air for 20 to 24 hours. The presence of a clear zone of inhibition around the disc (
8mm) was interpreted as positive for antibacterial activity. Plain disks inoculated with 10 microlitres of sterile water were used as a negative control on each agar plate. 
12 years) accounted for 191 of the total patients with one of the designated infections in these 2 practices, and accounted for eighteen antibiotic prescriptions. The prescription rate in the paediatric group was 9.4%. The paediatric prescription rate for URTI was 3.6% compared with the adult rate and overall rate of 0.9% and 1.4% respectively. The highest paediatric prescription rates were seen for cases of UTI (2 cases) and bronchitis (5 cases), both with prescription rates of 100%. In adult, the prescription rates were 100% and 64.8% respectively for these two illnesses. (Table 5)