Antibiotic Policy Fails to Halt Increased Antibiotic Prescribing

Abstract & Commentary

Synopsis: Despite seven years of antibiotic policy suggestions, the use of antibiotics increased from 12% to 19%; spending more than doubled as a result of prolonged prophylaxis in surgery, immunosuppressed patients, and empiric broad-spectrum therapy as a substitute for diagnosis or "just in case."

Source: Gould IM, Jappy B. Trends in hospital antibiotic prescribing after introduction of an antibiotic policy. J Antimicrob Chemother 1996;38:895-904.

Trends in antibiotic prescribing in the hospitals of the Grampian region of Northeast Scotland (population, 500,000) were monitored prospectively for seven years from 1986 using computerized ward stock lists and laboratory data relating to all in-patient and out-patient treatments. The main outcome measures were the number of antibiotics available for routine and restricted uses, annual expenditure, and defined daily doses (DDDs) of high expenditure antimicrobial agents. An antibiotic committee introduced a policy and formulary in the third year of the study which offered three choices of core antibiotics for uncomplicated infection, three for primary treatment failure or holding therapy while awaiting the results of culture, and three for cases of penicillin allergy. Guidance was also given on surgical prophylaxis, the avoidance of endocarditis, artificial joint infection during invasive procedures, and for avoiding potentially serious infections in immunocompromised patients. A list of reserved drugs was selected for the hospital formulary, which included only seven of the 30 new antibiotics that became available during the study and included ciprofloxacin, teicoplanin, and fluconazole. In 1991, a revised edition of the policy was introduced following widespread consultation leading to the choice of core drugs being extended to 41 oral and/or systemic agents, of which 13 were intended for restricted use reflecting an expansion of prophylactic use and the introduction of new agents.

The committee met quarterly, and the formulary and policy were publicized in at least three venues. During the time, there were no significant changes in acute bed services, bed occupancy, and deaths, although acute in-patient discharges increased by 15.4%, outpatient attendences were up by 17.4%, and long-stay bed numbers halved. However, the antibiotic policy had only limited success in controlling antibiotic prescribing since DDDs increased 46% to 627 per 1000 patient days, the proportion of the drug budget consumed by antibiotics increased from 11.9% to 18.7%, and gross expenditure on antibiotics more than doubled from $427,000 to $1,014,000. Two-thirds of the increase was due to the use of new drugs, particularly ciprofloxacin, fluconazole, and teicoplanin, but older antibiotics such as co-amoxiclav, cefuroxime, and cefotaxime were also more widely used with a shift toward more oral formulations. Only the use of amoxicillin and cephradine went against the trend. An audit of 104 prescriptions showed that half did not conform to policy and that 28 of the 52 prescriptions recommended by microbiologists did not conform to the policy. The three doses recommended for surgical prophylaxis were given in only 50% of 74 orthopedic operations and 24% of the 38 general surgical operations, and specimens were sent for culture in only 24% of such operations, with only one-quarter of cultures yielding significant growth (i.e., 6% of all operations). Sustained feedback to the surgeons over a period of two years failed to persuade them to alter their habits. Traditional control measures seem to have gone about as far as they can, and we still need to understand the reasons for overuse of antibiotics, one of which might be the system under which clinicians practice.


Everyone agrees that too many antibiotics are prescribed for too long, often for trivial reasons, but little seems to change. Task forces are assembled, working parties convened, and, as in this case, Antibiotic Committees are established and formularies are promulgated—but all apparently to little or no avail. The Grampian Region is to be commended for its efforts and the authors congratulated for attempting to shed light on the manifest resistance of medical practitioners to alter their prescribing habits. Surgeons and those caring for immunosuppressed patients insisted on prolonging prophylaxis; broad-spectrum drugs were preferred to those with a narrow spectrum, and performing invasive diagnostic tests for diagnosis and "spiralling empiricism" seemed to be the order of the day. Only drugs on the reserve list were generally prescribed properly. The authors suggested that prescribing might have been further improved by establishing computer links between pharmacy and microbiology, restricting reporting of susceptibility tests to a few drugs, and including the cost of treatment on the reports. However, tools, no matter how effective, can only work if they are used. Clearly, prescribers feel safer in treating "just in case" with agents that offer broader cover, especially when under immense pressure to cut bed occupancy and send patients home sooner. They also feel it simpler and easier to treat empirically rather than first establish a diagnosis before starting treatment, not least because the two processes run at different tempos on parallel tracks and are often totally oblivious to each other’s existence. Besides, as hospital admissions become more and more focussed on the very ill, empirical treatment is bound to become the norm both in hospital to keep infectious mortality to a minimum and for out patients to deal with more minor cases to minimize the risk of having to admit them. Since in-patients will be treated almost exclusively with parenteral preparations, it would be hard to resist the temptation of using the newest and, therefore, most expensive, drugs since these are always perceived to be the best available. This will also apply to the oral formulations given to out patients.

The authors concluded that antibiotic policies as practiced in the United Kingdom restricted access to new broad-spectrum agents and expensive preparations such as the lipid formulations of amphotericin B. However, this approach seems impotent in breaking the vicious circle of more prescribing leading to more resistance, which, in turn, encourages the use of the newer and more expensive drugs. Moreover, antibiotic policies seem bound to fail as they are based on a narrow understanding of the rational use of antibiotics, namely that they should only be used to treat infectious diseases caused by susceptible pathogens. Besides the fact that Western Society has little time for the masterly inactivity this demands, clinicians have widened the definition of "rational" to include a reasonable substitute for diagnosis, a basis for security, a means of defensive medicine, and so forth. Surgeons, always a breed apart, probably overplay their prophylaxis because, in the memorable words of the Wakely Prize winner in Lancet, December 21, 1996, "they regard even the most trivial wound infection as a slight on their skill." And, why not? Antibiotics are among the safest drugs (ironically their Achilles heel), and a course of antibiotics is often both cheaper and more reliable than microbiological diagnosis. Moreover, failure, as measured by genuine persistence or relapse, is actually quite rare. If, with the help of antibiotics, patients can be nursed through complex surgery, brought back from a near-dead experience by intensive care, and sustained during a period of life-threatening immunosuppressive treatment, all of which are hugely expensive anyway, where’s the harm? Epidemiologists and other specialists will shriek in horror and wag an admonishing finger, but the fact remains that antimicrobial resistance is generally perceived as a remote possibility, or at least the lesser of two evils, until it happens. Then it becomes someone else’s problem. Prescribers are voting with both feet and may actually only be responding intelligently to the complexities of a system under which they are forced to practice; neither understand nor have any real incentive for doing so. It is easy to blame the sinner rather than understand the sin, but, if we are serious about addressing the problem of inappropriate prescribing properly, we must at the very least go back to the drawing board, widen the terms of inquiry beyond the laboratory and the office, and take the time to be true scientists and look at all the facts, especially the awkward ones.