Article: Burstein B, Sabhaney V, Bone JN, Doan Q, Mansouri FF, Meckler GD. Prevalence of Bacterial Meningitis Among Febrile Infants Aged 29-60 Days With Positive Urinalysis Results: A Systematic Review and Meta-analysis. JAMA Netw Open. 2021;4(5):e214544. doi:10.1001/jamanetworkopen.2021.4544
Fellow: Jordan S. Lim, MD; Baylor College of Medicine, Texas Children’s Hospital
Summary: Fever in the first months of life is one of the most common pediatric problems in the United States. Urinary tract infections (UTIs) are the most frequent bacterial infections in this population with approximately 10% of febrile infants less than 60 days old having underlying UTIs. All previously published clinical guidelines and risk assessment tools, such as the PECARN rule or Step-by-Step algorithm, include positive urinalysis (UA) results as a risk factor for invasive bacterial infections (bacteremia and bacterial meningitis). Using these tools, lumbar puncture is recommended for febrile infants with positive UAs. Several studies have evaluated the prevalence of concomitant bacterial meningitis in infants with positive UAs but due to the rarity of this outcome, any single study is unable to provide precise estimates. This systematic review and meta-analysis aims to synthesize the evidence for the prevalence of bacterial meningitis among well-appearing febrile infants 29 to 60 days of age with positive urinalysis to help inform whether lumbar puncture is routinely required in this population.
The authors performed a comprehensive search of the MEDLINE and Ovid Embase databases for publications from 2000 to 2018. Included studies reported on well-appearing, term infants aged 29 – 60 days being evaluated for fever and with both UA data as well as meningitis status available. Meningitis status was defined as positive CSF culture results, bacteremia with CSF pleocytosis, or suggestive history at clinical follow-up if CSF was not obtained. A positive UA was defined as a WBC of 10/μL (0.01 × 109 /L) or higher on an uncentrifuged specimen (or ≥ 5/μL [0.01 × 109 /L] per high-power field on a centrifuged specimen), a positive Gram stain result, or any finding of leukocyte esterase or nitrites. Overall, 48 individual studies, 17 distinct data sets, and 25,374 infants from 2000 -2018 were included for analysis.
The prevalence of bacterial meningitis in well-appearing febrile infants 29 to 60 days of age with positive UA results was 0.25% (95% CI, 0.21%-0.36%) for clinical meningitis to 0.44% (95% CI, 0.25%-0.78%) for culture proven meningitis. Importantly, the prevalence of bacterial meningitis for these infants with a positive UA was not higher than that in infants with negative UA results. The estimates for infants with a negative UA ranged from 0.28% (95% CI, 0.21%-0.36%) to 0.50% (95% CI, 0.33% - 0.76%). Overall, these results suggest that well-appearing, febrile infants aged 29 – 60 days old with positive UA results are not at increased risk of bacterial meningitis compared to infants with a negative UA. Furthermore, for well-appearing febrile infants aged 29 – 60 days, the decision to perform lumbar puncture and invasive CSF testing should not be guided by urinalysis results alone.
Strength: This systematic review and meta-analysis is the largest and most comprehensive study to date (48 studies, 17 distinct data sets, and 25,374 infants from 2000 -2018) to evaluate this clinical question. The large number of infants included allows for pooled prevalence of bacterial meningitis to be ascertained, which has been difficult to assess accurately given its overall low prevalence in this population. The authors focused on the idea population to fit the clinical question (well appearing, positive UA) and only included studies with clinical follow up data available for infants without CSF. Additionally, the authors confirmed the integrity of the data in included studies by acquiring the datasets from authors.
Limitations: There is the possibility for “double counting” of infants given the use of national, multicenter databases; however, steps were taken to account for this through review of individual datasets. Also, there is a possibility for false elevation of the pooled prevalence of bacterial meningitis among infants with negative urinalysis results as infants with a normal urinalysis who did not undergo lumbar puncture would not be included. However, this would mean the true prevalence is lower than what was found. The threshold for CSF pleocytosis varied from study to study, which could lead to some variability in meningitis classification. Lastly, the study did not analyze whether abnormal inflammatory markers (i.e., CRP or procalcitonin) in conjunction with a positive UA would increase the risk for bacterial meningitis. Further research in this area is needed.
Major Takeaway: The pooled prevalence of bacterial meningitis among well-appearing, febrile infants aged 29 – 60 days with positive UA results was not higher compared with similar infants who had negative UA results. These results suggest the rate of concomitant bacterial meningitis in this population is low and that well-appearing febrile infants aged 29 – 60 days with positive UA results are not at increased risk of bacterial meningitis compared with similar infants with negative UA results. Therefore, invasive CSF testing (lumbar puncture) in well-appearing, febrile infants in the second month of life based on a positive UA result alone is not supported by differential risk.
How this article should impact our practice: Previously, all published clinical guidelines and risk assessment tools have utilized positive UA results as a risk factor. The evidence from this study suggests that this is no longer supported by the evidence and that the decision to perform invasive CSF testing and lumbar puncture in well-appearing, febrile infants in the second month of life should not be based solely upon the results of urinalysis. Of note, the recently published 2021 AAP Clinical Practice Guidelines for Febrile Infants no longer utilizes UA results as a risk factor for determining whether to perform lumbar puncture and CSF analysis in these infants.
Fellow: Jordan S. Lim, MD; Baylor College of Medicine, Texas Children’s Hospital
Summary: Fever in the first months of life is one of the most common pediatric problems in the United States. Urinary tract infections (UTIs) are the most frequent bacterial infections in this population with approximately 10% of febrile infants less than 60 days old having underlying UTIs. All previously published clinical guidelines and risk assessment tools, such as the PECARN rule or Step-by-Step algorithm, include positive urinalysis (UA) results as a risk factor for invasive bacterial infections (bacteremia and bacterial meningitis). Using these tools, lumbar puncture is recommended for febrile infants with positive UAs. Several studies have evaluated the prevalence of concomitant bacterial meningitis in infants with positive UAs but due to the rarity of this outcome, any single study is unable to provide precise estimates. This systematic review and meta-analysis aims to synthesize the evidence for the prevalence of bacterial meningitis among well-appearing febrile infants 29 to 60 days of age with positive urinalysis to help inform whether lumbar puncture is routinely required in this population.
The authors performed a comprehensive search of the MEDLINE and Ovid Embase databases for publications from 2000 to 2018. Included studies reported on well-appearing, term infants aged 29 – 60 days being evaluated for fever and with both UA data as well as meningitis status available. Meningitis status was defined as positive CSF culture results, bacteremia with CSF pleocytosis, or suggestive history at clinical follow-up if CSF was not obtained. A positive UA was defined as a WBC of 10/μL (0.01 × 109 /L) or higher on an uncentrifuged specimen (or ≥ 5/μL [0.01 × 109 /L] per high-power field on a centrifuged specimen), a positive Gram stain result, or any finding of leukocyte esterase or nitrites. Overall, 48 individual studies, 17 distinct data sets, and 25,374 infants from 2000 -2018 were included for analysis.
The prevalence of bacterial meningitis in well-appearing febrile infants 29 to 60 days of age with positive UA results was 0.25% (95% CI, 0.21%-0.36%) for clinical meningitis to 0.44% (95% CI, 0.25%-0.78%) for culture proven meningitis. Importantly, the prevalence of bacterial meningitis for these infants with a positive UA was not higher than that in infants with negative UA results. The estimates for infants with a negative UA ranged from 0.28% (95% CI, 0.21%-0.36%) to 0.50% (95% CI, 0.33% - 0.76%). Overall, these results suggest that well-appearing, febrile infants aged 29 – 60 days old with positive UA results are not at increased risk of bacterial meningitis compared to infants with a negative UA. Furthermore, for well-appearing febrile infants aged 29 – 60 days, the decision to perform lumbar puncture and invasive CSF testing should not be guided by urinalysis results alone.
Strength: This systematic review and meta-analysis is the largest and most comprehensive study to date (48 studies, 17 distinct data sets, and 25,374 infants from 2000 -2018) to evaluate this clinical question. The large number of infants included allows for pooled prevalence of bacterial meningitis to be ascertained, which has been difficult to assess accurately given its overall low prevalence in this population. The authors focused on the idea population to fit the clinical question (well appearing, positive UA) and only included studies with clinical follow up data available for infants without CSF. Additionally, the authors confirmed the integrity of the data in included studies by acquiring the datasets from authors.
Limitations: There is the possibility for “double counting” of infants given the use of national, multicenter databases; however, steps were taken to account for this through review of individual datasets. Also, there is a possibility for false elevation of the pooled prevalence of bacterial meningitis among infants with negative urinalysis results as infants with a normal urinalysis who did not undergo lumbar puncture would not be included. However, this would mean the true prevalence is lower than what was found. The threshold for CSF pleocytosis varied from study to study, which could lead to some variability in meningitis classification. Lastly, the study did not analyze whether abnormal inflammatory markers (i.e., CRP or procalcitonin) in conjunction with a positive UA would increase the risk for bacterial meningitis. Further research in this area is needed.
Major Takeaway: The pooled prevalence of bacterial meningitis among well-appearing, febrile infants aged 29 – 60 days with positive UA results was not higher compared with similar infants who had negative UA results. These results suggest the rate of concomitant bacterial meningitis in this population is low and that well-appearing febrile infants aged 29 – 60 days with positive UA results are not at increased risk of bacterial meningitis compared with similar infants with negative UA results. Therefore, invasive CSF testing (lumbar puncture) in well-appearing, febrile infants in the second month of life based on a positive UA result alone is not supported by differential risk.
How this article should impact our practice: Previously, all published clinical guidelines and risk assessment tools have utilized positive UA results as a risk factor. The evidence from this study suggests that this is no longer supported by the evidence and that the decision to perform invasive CSF testing and lumbar puncture in well-appearing, febrile infants in the second month of life should not be based solely upon the results of urinalysis. Of note, the recently published 2021 AAP Clinical Practice Guidelines for Febrile Infants no longer utilizes UA results as a risk factor for determining whether to perform lumbar puncture and CSF analysis in these infants.