NEW YORK (Reuters Health) – Data from the UK General Practice Research Database suggest that failure to prescribe antibiotics when patients present with a lower respiratory tract infection may lead to higher rates of hospitalization and mortality. Of the 14% of the study cohort not treated with antibiotics on the day of presentation, 33% died.

These findings, published in the May issue of Chest, prompted the author of a linked editorial to question if we have gone too far in limiting antibiotic use in lower respiratory tract infections?

The real answer may be that clinicians are giving antibiotics to the wrong patients.

“Most cases of pneumonia result from bacterial infection, and antibiotic therapy remains the mainstay of pneumonia management,” note Dr. Christopher C. Winchester and co-investigators at the University of Aberdeen. However, routine use of antibiotics in primary care is being discouraged to prevent antimicrobial resistance.

To study these issues, the researchers identified 151,088 patients from 346 primary care practices who presented for treatment of a lower respiratory tract infection in 2004.

Excluding those who died on the index date, 128,580 (85.8%) were prescribed an antibiotic. “The only factor to be associated with a reduced risk of both hospital admissions and death related to respiratory infection … was antibiotic prescription on the day of diagnosis.”

In the 3 months following the first lower respiratory tract infection, 13.4% were admitted to a hospital, 5.6% of these patients had respiratory infections. Antibiotic prescription on the index date was associated with a significantly reduced likelihood of hospitalization among patients 18 to 64 years of age (hazard ratio 0.61), but not among patients 1 to 17 years of age or those age 65 or older.

Antibiotics were associated with significantly reduced risk of death among patients 18 to 64 years old (HR 0.21) those 65 years of age or older (HR 0.31). The overall association remained significant when excluding patients hospitalized on the index date (HR 0.37).

More than half of the deaths occurred on the index date, suggesting that “patients with a lower respiratory tract infection may be consulting a primary care physician too late for such adverse outcomes to be avoided,” Dr. Winchester’s group speculates.

They point out that the number of patients needed to treat to prevent death or hospital admission was well over 1000, so clearly, for most patients, antibiotics may not be necessary. “A cautious approach may be needed,” they say, when deciding which patients should be treated, “particularly in high-risk and vulnerable groups.”

However, editorialist Dr. Grant W. Waterer at the University of Western Australia, Perth, cites the lack of information about confounding factors in this study, and notes that Winchester et al. present no evidence that prescribing antibiotics more frequently would have led to more favorable outcomes.

He concludes: “Prospective observational studies can and should be conducted to both confirm the observations of Winchester et al. and to tease out the causal relationships if they exist.”

Reference:
Chest 2009;135:1118-1120,1163-1172.