What is a good clinical prediction tool for pneumonia diagnosis?

From the Guidelines

The CRB-65 score is a highly effective and recommended clinical prediction tool for pneumonia diagnosis and severity assessment, as it is simple, easy to calculate, and has been validated in several studies 1.

Key Factors of CRB-65 Score

The CRB-65 score evaluates four key factors:

• Confusion
• Raised respiratory rate (30 breaths/min or more)
• Low blood pressure (systolic less than 90 mmHg or diastolic 60 mmHg or less)
• Age 65 or older Each factor scores one point, with a total score ranging from 0-4.

Interpretation of CRB-65 Score

Patients are stratified for risk of death (within 30 days) as follows:

• 0: low risk (less than 1% mortality risk)
• 1 or 2: intermediate risk (1 to 10% mortality risk)
• 3 or 4: high risk (more than 10% mortality risk)

Clinical Judgment and CRB-65 Score

Clinical judgment is the most effective way to assess a patient’s risk of severe disease, and the CRB-65 score should be used in conjunction with clinical judgment to inform decisions about hospitalization and antibiotic therapy 1.

Comparison with Other Tools

The Pneumonia Severity Index (PSI) is another valuable tool, though more complex with 20 variables. For primary care settings where laboratory testing may be limited, the CRB-65 provides a simplified alternative 1.

Recent Studies

Recent studies have validated the use of the CRB-65 score in primary care settings, and have shown that it is a useful tool for predicting mortality and guiding treatment decisions 1.

Limitations

While the CRB-65 score is a useful tool, it is not perfect and should be used in conjunction with clinical judgment and other diagnostic tests, such as chest radiography and laboratory tests, to guide treatment decisions 1.

From the Research

Clinical Prediction Tools for Pneumonia Diagnosis

Several clinical prediction tools have been proposed and validated for the diagnosis and management of community-acquired pneumonia (CAP). These tools aim to predict patient outcomes, guide initial management, and identify patients at low or high risk of mortality.

Commonly Used Prediction Tools

• Pneumonia Severity Index (PSI): a well-established clinical prediction rule that categorizes patients into five risk classes based on demographic and clinical factors 2, 3, 4.
• CURB-65 score: a six-point score that assesses confusion, urea, respiratory rate, blood pressure, and age to predict mortality in CAP patients 2, 5, 3, 4.
• A-DROP score: a prediction rule proposed by the Japan Respiratory Society, which assesses age, blood urea nitrogen, SpO2, confusion, and systolic blood pressure to predict mortality in CAP patients 2.
• CRB-65 score: a simplified version of the CURB-65 score that omits the blood urea measurement 5.
• Other scores, such as the ATS 2001, ATS/IDSA 2007, SCAP score, and SMART-COP, have also been proposed and validated for predicting ICU admission and mortality in CAP patients 6.

Performance of Prediction Tools

• The PSI and CURB-65 scores have been shown to have excellent long-term prognostic accuracy, with a stepwise increase in mortality per risk class 3.
• The A-DROP score has similar power to the PSI score in predicting short-term mortality and is slightly more accurate in identifying patients at low risk than the CURB-65 score 2.
• The addition of blood biomarkers, such as pro-adrenomedullin and pro-atrial natriuretic peptide, can further improve the prognostic capabilities of the PSI and CURB-65 scores 3.
• The PSI and CURB-65 scores remain adequate for predicting mortality in clinical practice, even in patients with SARS-CoV-2 CAP 4.

Key Findings

• Clinical prediction tools, such as the PSI and CURB-65 scores, are useful for predicting mortality and guiding management in CAP patients 2, 5, 3, 4.
• The choice of prediction tool may depend on the specific clinical context and patient population 6.
• The addition of blood biomarkers and other clinical factors can improve the prognostic performance of these tools 3.