European Surveillance of Antibiotic
A Study of the European Society of Biomodulation and Chemotherapy
Antibiotic use data, although preliminary, gives some interesting insights. The low aminoglycoside and extended spectrum b -lactam resistance rates in Pseudomonas, Acinetobacter and Enterobacteriaceae from Aberdeen are associated with low use of these agents. However, the high MRSA rates, in common with virtually all UK hospitals, may well be associated with the extremely high co-amoxiclav use in Aberdeen, compared with the other centres in this project.
The high use of imipenem in Leipzig may be a factor in the high isolation rate of Enterococci Alerts as previously suggested. The lowest rates of Enterococci were seen in Aberdeen, possibly because co-amoxiclav is active against 99% of such strains in Aberdeen at the moment.
Antibiotic exposure was undoubtedly high. Where data were available, approximately three-quarters of patients had received at least one antibiotic and almost one third had received two or more agents.
There were surprising and wide variations in antibiotic use. Leipzig, with its problems of enterococcal Alerts, had a very high use of vancomycin, although Aberdeen had the lowest use (1/18 that of Leipzig) despite a major MRSA problem.
The high use of ciprofloxacin has not, as yet, been associated with major problems, at least as detected by the project.
The marked differences between centres in isolation rates for certain species, such as Haemophilus, pneumococcus, pseudomonas and anaerobes, may reflect not only specimen differences, but also processing differences. The specimen differences may be due both to patient differences or sampling differences and it will be interesting to collect data on such variables.
Whilst many Alerts remained at lower numbers than originally anticipated from previous studies, there was an overall increase in Alerts from 3.3 to 4.8% of all Denominator organisms. In particular, Alert pneumococci increased from 4.1% in 1999 to 12.2% in 2000. Whilst the rate in Warsaw actually went from 16.5% in 1999 to 1.9% in 2000, the situation in Bratislava NCI was completely different, increasing from 4% in 1999 to 16% in 2000. However, there was no high level resistance. These large swings indicate, most probably, outbreaks as well described in the literature. Alternatively, they could possibly be due to sampling artefacts and well illustrate the problems of drawing conclusions from snapshot surveys over short periods of time.
b -lactamase producing H. influenzae remained low although rates increased from 1.6 to 5%. The high rates of Alert Enterobacteriaceae in Warsaw reflect the inclusion of the National Reference Laboratory in this project. It has a particular interest in ESBL production.
Comments on antibiotic use and MRSA Alerts in Aberdeen have been made elsewhere in this report. Another issue concerning this is the poor state of cleanliness in UK hospitals which is currently being addressed.
The drop in ciprofloxacin resistance in Campylobacter may reflect changes in antibiotic use patterns in animal husbandry.
Glycopeptide resistance in CNS remained very rare despite the large numbers of patients being treated for malignancies and the high use of glycopeptides. Only one isolate of glycopeptide resistant S. aureus was reported although no centres are screening routinely for such Alerts and routine detection methods are unlikely to detect them.
The high number of intensive care and cancer patients is reflected in the high numbers of Alert Pseudomonas and Acinetobacter, the latter seeming particularly well adapted to colonisation and spread in this environment.
Outcome data shows higher mortality than would be expected from the literature but, how much is attributable to the Alert organisms is, of course, only speculation. Whilst cure rates may be considered disappointing, a number of factors such as chronicity of underlying disease and colonisation rather than infection may explain such data.
The use of quinolones for prophylaxis seems questionable, particularly in view of the resistance problems. More data on antibiotic consumption including defined daily doses (DDDs) should be available next year.
© ESBiC Datacenter Munich, 2001