1 The safety in everyday clinical usage of three 4-quinolone antibiotics, (ciprofloxacin, norfloxacin and ofloxacin), was compared with similar data for azithromycin and cefixime, each agent being examined by Prescription-Event Monitoring (PEM) during the early post-marketing period.
2 In PEM the exposure data are derived from general practitioner prescriptions confidentially provided by the Prescription Pricing Authority. Outcome data are provided by questionnaires (green forms) on which the prescribing medical practitioner records event data. When necessary, further information is obtained from a number of sources which include follow-up of all pregnancies and the patients' life-time medical record.
3 The main outcome measures were demographic information, including the patient's date of birth and sex; the indication for prescribing the drug being monitored; the reason for stopping treatment; the start and stop dates of treatment and the events recorded during and after treatment.
4 The final cohort for each of the five antibiotics exceeded 11000 patients. The only event significantly related to the use of all five antibiotics was nausea/vomiting. This was also the most frequent adverse event causing treatment to be discontinued with norfloxacin, ofloxacin and azithromycin (relevant information was not requested in the studies of ciprofloxacin and cefixime). Vaginal candidiasis was significantly more frequently associated with the use of the three 4-quinolones than with azithromycin and cefixime but it was frequently delayed until the week or two after the cessation of therapy. Within each event, as recorded in these studies, the highest event rates (the number of events per 1000 patients) in the week following the start of therapy were: 9.2 for diarrhoea with cefixime; 4.9 for nausea/vomiting with ofloxacin; 2.4 for rash with azithromycin; 2.2 for abdominal pain with norfloxacin; 1.5 for headache/migraine with ofloxacin; 1.4 for malaise/lassitude with ofloxacin; 1.2 for dizziness with norfloxacin. Uncommon events (reported in less than 1:1000 patients) included rare cases of allergic phenomena, convulsions and pseudomembranous colitis. There were no reports of tendinitis, tenosynovitis or tendon rupture in children but tendon disorders were reported in the two months following the start of treatment in 20 adults. A total of 307 pregnancies were reported. Thirty-eight of the 55 women who received these drugs during the first trimester of pregnancy gave birth to healthy babies. No congenital abnormalities were reported. Apart from one case of unconfirmed pseudomembranous colitis, none of the other 2468 deaths that occurred in these studies was attributed to the antibiotics.
5 These five antibiotics are acceptably safe antimicrobial agents when used in general medical practice. PEM is an effective method for monitoring the safety of recently introduced antimicrobial agents.