To the Editor
In a recent issue of the Journal, Conte et al (
1
) reported a prognostic rule for elderly patients hospitalized with community-acquired pneumonia. The rule consisted of a simple staging system that used five readily available predictor variables to stratify patients according to the risk of hospital mortality into one of four stages. The authors suggested that the system may be useful for standardization of case mix when comparing interhospital mortality rates or when evaluating therapeutic agents for patients over the age of 65 years.The study shows that this discriminant rule identifies risk for geriatric patients more accurately than the British Thoracic Society rule (
2
), a rule developed using data for all adults. We believe these results should be interpreted with caution, because the British rule was intended to identify only the sickest patients for admission to an intensive care unit. The prognostic rule of Conte et al has yet to be compared with a rule that accurately stratifies patients into risk categories from the lowest through the highest probability of death. Additionally, the accuracy of the rule has not been evaluated prospectively or using outcomes other than hospital mortality (eg, 30-day mortality, hospital length of stay, intensive-care-unit admission, and repeat hospitalizations). The latter outcomes may be an important consideration when assessing hospital performance for patients with pneumonia.An attractive feature of this prognostic rule for geriatric patients is how easily it can be applied to identify risk of hospital mortality quickly. However, we think it is important to note that the rule was not intended to be used prospectively to assist clinicians with initial site of treatment decisions for older patients. Among other limitations, the 4% risk of hospital mortality for patients in the lowest risk stage (no risk factors) is unacceptably high to consider treating these patients with an initial course of outpatient therapy. A prognostic rule such as the Pneumonia Severity Index that accurately identifies adult patients at very low risk of an adverse medical outcome may be a better choice for hospital admission decisions (
3
).References
- A prognostic rule for elderly patients admitted with community-acquired pneumonia.Am J Med. 1999; 106: 20-28
- Community-acquired pneumonia in adults in British hospitals in 1982–1983.Q J Med. 1987; 62: 195-220
- A prediction rule to identify low-risk patients with community-acquired pneumonia.NEJM. 1997; 336: 243-250
Article info
Publication history
Published online: August 16, 2004
Identification
Copyright
© 2000 Excerpta Medica Inc. Published by Elsevier Inc. All rights reserved.
