Bollettino epidemiologico nazionale

FIELD STUDIES - Evaluation of the Impact of a Training Program to Change Antibiotic Prescribing Practices

Pasquale Falasca, Franco Gianelli, Maria Vittoria Novi, Davide Resi – Local Health Authority, Ravenna, Italy

In Italy, the use of broad-spectrum antibiotics in general medicine is common. In recent years, rising levels of antibiotic resistance have led to the prioritisation of rational use of antibiotics as a public health problem (1).  Training programs for general practitioners may be useful in encouraging appropriate prescribing practices (2).

To improve prescribing practices, during the fall of 1999, six physicians from San Pietro in Vincoli who formed part of a group practice working under contract with the Local Health Authority of Ravenna (Emilia-Romagna) and who covered a patient population of 6900 underwent a six-hour course on rational antibiotic use. The training was given over the course of two evenings by an infectious disease specialist. Topics discussed included antibiotic resistance patterns of the organisms most commonly seen in general practice in the area, with special emphasis on common respiratory pathogens.  Information on local antibiotic resistance patterns was provided by the microbiology laboratory of the Local Health Authority of the area. In addition, the physicians also received guidelines on prescribing practices that had previously been distributed (at the beginning of 1997 and again in April 1999) to all physicians working under contract with the Local Health Authority. In these guidelines, the recommended first-choice antibiotic for respiratory infections was oral penicillin.

To evaluate the effects of the training program, antibiotic use was compared for the first trimesters of 1999 and 2000, the peak season for respiratory infections in the area. The prescribing practices of the group receiving the intensive training was compared with that of all 300 physicians working for the Local Health Authority.

The analysis was limited to antibiotics for systemic use (J01 in the ATC classification system). Outcomes compared between the two groups included number of patients receiving antibiotics (exposed/1000 enrolled patients), the volume of drugs prescribed (measured in daily drug dosage DDD)/1000 enrolled patients/day, and the costs per enrolled patient per day of treatment.

  The number of patients treated with antibiotics/1000 enrolled patients during the trimester) for the S. Pietro in Vincoli physicians was nearly identical to that of the Authority as a whole for the first trimester of 1999.  The prevalence rate declined for both groups during the first quarter of 2000, but the decline was considerably greater in the S. Pietro in Vincoli physicians (-21% versus –11%). Expenditures per enrolled patient in S. Pietro, which were initially higher than those for the Authority, declined 22% between 1999 and 2000; this compares with a decline of 6% for the 300 Authority physicians.

While the data on costs per enrolled patient is influenced by the number of persons treated, the expenditures per day of therapy is exclusively a function of the costs of the drugs used. As shown in the Table, the physicians in the S. Pietro in Vincoli network used less costly drugs in 2000 than they had in 1999.  Thus, the decrease in overall costs for the network was the result of both fewer patients treated and changes in prescribing practices, as evidenced by decreased use of cefalosporin, macrolides and quinolones, while the decline in the expenditures for the Authority resulted only from reductions in numbers of patients treated.

Paradoxically, the DDD/day increased over time in the group from San Pietro in Vincoli, while for the Authority, this value decreased. This apparent contradiction is due to a move away from injectable antibiotics by the S. Pietro in Vincoli towards oral medications; the former are available in mono-dose packages, while oral medications are supplied in multi-dose packages.

In conclusion, during the observation period, the physicians of the network of S. Pietro in Vincoli considerably modified their antibiotic prescribing practices. In particular, the frequency of antibiotic prescription decreased overall, as did the pattern of antibiotic use.  Such changes appear to be due to having participated in the training course since this was the only intervention that differed between the network and the Local Health Authority as a whole.

The results of our study suggests that continuing medical education may be an excellent tool when combined with normative guidelines to improve the therapeutic approaches of individual physicians and the rational use of antibiotics.

Commentary

Pierluigi Morosini, Laboratorio di Epidemiologia e Biostatistica, Istituto Superiore di Sanita’

Recent reviews appearing in the Cochrane Library (2) and elsewhere (3) have summarized the current knowledge of the efficacy of various measures to change medical practice. These reviews demonstrate that distribution of brochures or other information material have little effect. Participation in continuous improvement of quality programs and in audits (evaluation followed by feedback without other specific interventions) can have a minor but non-negligible efficacy. The greatest improvements are obtained with recommendations actively supported by a leader in the field, with so-called outreach visits whereby specially trained personnel visit physicians’ offices.  This method is similar to that used by pharmaceutical companies, although in this case the personnel are trained and sent by university centers. It also includes the distribution of summary resource documents or computer programs to reinforce the message.

The work of Falasca, Gianelli, Novi, and Resi, conducted in the Local Health Agency of Ravenna is a good example of the improvement in antibiotic prescribing practices obtained with a continous quality improvement with some elements of outreach visits and the involvement of a leader in the field, in this case an infectious disease specialist who provided a training course for a limited group of physicians.

In programs designed to improve quality, most evaluations are limited to a pre- and post-training comparison of practices. One of the strengths of this study is that it not only compares the pre- and post-training performance of the group of physicians who underwent training but also compares the changes in their performance over time with that of physicians from the same Local Health Agency who received only printed material.  This use of a comparison group largely eliminated one of the major potential sources of confounding in the interpretation of the results and permitted the authors to demonstrate that the rational use of antibiotics by the physicians who underwent more intensive training was in large part due to their participation in the training program rather than to non-specific changes over time.  It should be noted, however, that the other physicians of the Local Health Agency also changed their practices during the study period, albeit to a lesser extent.  This may be due to the epidemiologic phenomenon known as “contamination”, whereby the other physicians may have improved their knowledge through contact with the six intensively trained physicians in the program.

1.   Traversa G. Notiziario Ist. Super.Sanità 2001; 14(1): iv.

2.   O'Brien Thomson MA, Oxman AD, Davis DA, et al. Audit and feedback versus alternative strategies: effects on professional practice and health care outcomes The Cochrane Library, Issue 1, 2001. Oxford: Update Software.

3.   Bero L, Grilli R, et al. BMJ, 1998; 317: 465-8.