Diagnostic model appears to be more effective than clinical judgment in detecting serious bacterial infection in young febrile children

Article Outline

• Question
• Design
• Setting
• Participants
• Intervention
• Outcomes
• Main Results
• Conclusions
• Commentary
• Copyright

Craig JC, Williams GJ, Jones M, Codarini M, Macaskill P, Hayen A, et al. The accuracy of clinical symptoms and signs for the diagnosis of serious bacterial infection in young febrile children: prospective cohort study of 15 781 febrile illnesses. BMJ 2010;340:c1594.

Back to Article Outline

Question 

Among young children presenting with a febrile illness but suspected of having serious bacterial infection (SBI), is there a way to distinguish SBIs from self-limiting non-bacterial illnesses?

Back to Article Outline

Design 

Two year prospective cohort study.

Back to Article Outline

Setting 

The emergency department of The Children's Hospital at Westmead, Westmead, Australia.

Back to Article Outline

Participants 

Children < 5 years presenting with a febrile illness between July 2004 and June 2006.

Back to Article Outline

Intervention 

A standardized clinical evaluation by physicians, which required entry of 40 clinical features into the hospital's electronic record keeping system. Physicians were also asked to estimate the probability of the patient having 10 different diagnoses. SBIs were confirmed or excluded using standard radiological and microbiological tests and follow-up.

Back to Article Outline

Outcomes 

Diagnosis of one of three key types of SBI (urinary tract infection, pneumonia, and bacteremia), and the accuracy of both the clinical decision making model and clinician judgment in making these diagnoses.

Back to Article Outline

Main Results 

Follow-up data was available for 93% of the 15 781 instances of febrile illnesses recorded during the study period. The combined prevalence of any of the three infections of interest (urinary tract infection, pneumonia, or bacteremia) was 7.2% (1120/15 781, 95% confidence interval (CI), 6.7% to 7.5%), with urinary tract infection the diagnosis in 543 (3.4%) cases of febrile illness (95% CI, 3.2% to 3.7%), pneumonia in 533 (3.4%) cases (95% CI, 3.1% to 3.7%), and bacteremia in 64 (0.4%) cases (95% CI, 0.3% to 0.5%). Almost all (>94%) of the children with SBIs had the appropriate test (urine culture, chest radiograph, or blood culture). Antibiotics were prescribed acutely in 66% (359/543) of children with urinary tract infection, 69% (366/533) with pneumonia, and 81% (52/64) with bacteremia. However, 20% (2686/13 557) of children without bacterial infection were also prescribed antibiotics. On the basis of the data from the clinical evaluations and the confirmed diagnosis, a diagnostic model was developed using multinomial logistic regression methods. Physicians' diagnoses of bacterial infection had low sensitivity (10-50%) and high specificity (90-100%), whereas the clinical diagnostic model provided a broad range of values for sensitivity and specificity.

Back to Article Outline

Conclusions 

Emergency department physicians tend to underestimate the likelihood of SBI in young children with fever, leading to undertreatment with antibiotics. A clinical diagnostic model could improve decision making by increasing sensitivity for detecting SBI, thereby improving early treatment.

Back to Article Outline

Commentary 

Fever is one of the most common problems in pediatrics. A number of studies have particularly targeted approaches to the management of fever in young infants. In addition, the value of an accurate history and physical exam has received renewed attention in medical education. Despite these emphases, it is unclear how accurate physician judgment is in diagnosis SBIs in children. In this elegant study, the authors explored the clinical decision-making process of physicians in Australia who were evaluating children under 5 years of age with a febrile illness. This is the largest prospective study to date (almost 16 000 children, of whom 1120 had a SBI) of children with fever. Of note in this study, only 65-80% of children who were ultimately diagnosed with a SBI were given antibiotics initially. The authors felt that this problem could be due to two factors: (1) physician difficulty with incorporating multiple pieces of information, resulting in an underestimation of the likelihood of SBI; and (2) errors in the interpretation of common tests, such as urinalysis and chest x-ray. The systematic solution developed by the authors is a decision-support tool which incorporates multiple variables and provides the clinician with an estimate of the risk of SBI. Their final model, which includes 26 key variables, was validated in a large subset of their population. In this study, the decision-support tool was more accurate than the physicians' clinical judgment. The results from this paper, if replicated in other populations, represent a more effective way to sort out the children with serious bacterial infections from those with more benign causes of fever.