[Lung diseases in the elderly. Assessment of guidelines for the probabilistic prescription of antibiotics in a department of geriatric care].

Objective: We developed a prescribing guideline containing recommendations for the initial empirical antibiotic therapy in community or nosocomial pneumonia. The aim of the present study was to examine the impact of this measure.
Method: The prescribing guideline was implemented in May 1999. We retrospectively reviewed the charts of all patients>65 years with community-, or nursing home- or hospital-acquired pneumonia hospitalised in our department of acute geriatric care between May 1999 and November 2000. The criteria assessed were: consistence with the guideline, clinical effectiveness within 72 hours, adequation with the isolated germs and intra-hospital mortality.
Results: Data were collected on 112 patients (63 women et 49 men; mean age=80 +/- 8 Years). The pneumonia was community-acquired in 52 cases (46%), nursing home acquired in 25 cases (22%) and hospital-acquired in 35 cases (31%). Antibiotic prescription was consistent with the guideline in 64 cases (57%). When the antibiotic therapy was consistent, the patients were more likely to improve within 72 hours (45/64 versus 23/48; p=0.01). Despite a tendency, the number of antimicrobial treatments adapted to the isolated microorganisms was not significantly higher in the consistent group (22/36 adapted treatments versus 10/20). The intra-hospital mortality (25%) was similar in the two groups consistent and not consistent with the guideline. SARM was the most frequent multiresistant bacteria that was isolated.
Conclusion: The use of a prescribing guideline might improve the efficiency of empirical probabilistic antibiotic therapies. The impact of the guideline use on overall antibiotic costs and microbiological flora remains to be determined.