Monitoring of Antibiotic Resistance

MONITORING OF ANTIBIOTIC RESISTANCE

To get the most out of a newly implemented steward-ship programme (or a change to an existing one) it is necessary to have the means by which to measure its effect. It is, therefore, necessary to plan in advance what parameters will be measured and what will be the baseline data against which the changes will be judged. Some of the parameters might include, but are not restricted to:

• Antibiotic consumption and costs, both in total and by specific drug class

• Costs associated with prescribing potentially toxic antibiotics, e.g. gentamicin and vancomycin blood level monitoring

• Rates of resistance to specific antibiotics by problem pathogens

• Pharmacy interventions to advise on inappropriate antibiotic use

• The incidence of hospital-acquired infections.

One of the problems that has dogged reviewers trying to assess the extent of the benefits of stewardship programmes is that in many of the cases reported in the medical literature multiple changes to an established programme have been introduced together, or they have overlapped in time so that evaluating the contribution of each change has been difficult. It is, therefore, worthwhile deciding in advance when a new policy or practice will be implemented and when its effect will be assessed. The input of information technology specialists and hospital epidemiologists to a stewardship management team becomes important since they, together, can decide how the data will be recorded and analysed to best effect.

The more sophisticated antibiotic control and information systems do not simply record data on antibiotic consumption, cost and resistance, but are capable of relating infection control data to antibiotic use and would be expected to draw attention to situations where a change in use of a particular antibiotic was associated ith increasing isolation of a particular pathogen. Such n association does not, of course, mean that one caused the other, but it does raise staff awareness of that potential. Computer-assisted surveillance of hospital-acquired infection (HAI) has been shown, in some cases, to be more effective than manual monitoring and reporting; as long ago as 1986 one study reported that 90% of antibiotic-resistant HAIs were detected by computer compared to 76% manually. There is also the potential to improve antibiotic prescribing by minimizing the risk of adverse effects when information systems provide patient-specific warnings on allergies, immune and renal functions and the potential for interaction with the patient’s other drugs.

It is not surprising, perhaps, that there are marked differences in antibiotic resistance patterns from one country to another. This is illustrated in Figure 15.2, which shows that penicillin resistance in Streptococcus pneumoniae can vary from less than 5% in some European countries to more than 50% in others. Smaller, but nevertheless significant, variations may also arise between different regions of a single country; Figure 15.3 shows the corresponding Strep. pneumoniae data for the UK from the national Heath Protection Agency. However, data on local resistance patterns are of paramount importance and well-structured monitoring programmes should be capable of identifying unforeseen consequences of changes in antibiotic use such as that arising when a preapproval policy for cephalosporins was introduced in a New York hospital in an attempt to control cephalosporin resistance in Klebsiella species. The policy did achieve a 71% reduction in ceftazidime-resistant Klebsiella isolated in intensive care units, but monitoring revealed a concomitant rise in imipenem use and a 69% increase in imipenem-resistant Ps. aeruginosa that was attributed largely to the preapproval policy.