Risk stratification for CNS infection: A potential tool to avoid unwarranted lumbar punctures – An observational study

Abstract

Objective

Central nervous system (CNS) infection poses a diagnostic challenge especially in elderly patients who frequently exhibit atypical symptoms. Our study aimed to identify patients with a low risk of CNS infection, in whom lumbar puncture (LP) could be avoided.

Methods

Observational study of consecutive adult patients who underwent a LP in the emergency room (ER) of Hospital Universitari Vall d’Hebron between January 2017 and December 2021. We performed a univariate and multivariate analysis to identify factors associated with non-CNS infection. These factors were used to create a combined variable, and its diagnostic positive predictive value and specificity to detect patients without CNS infections were calculated.

Results

We included 489 patients of which 77 (15.7%) were diagnosed with CNS infection. Median age was 62 years (IQR 41–78) and 240 (49.1%) were male. In the multivariate analysis, variables associated with non-CNS infection were female sex (OR 1.89; 95% CI 1.12–3.20), age older than 80 years (OR 3.14; 95% CI 1.20–8.19), previous cognitive impairment (OR 3.91; 95% CI 1.18–13.01), and clinical presentation without meningitis triad (fever, headache and neck stiffness) (OR 4.12; 95% CI 1.72–9.85). A composite variable encompassing age older than 80, cognitive impairment, and the absence of the meningitis triad was used as a diagnostic tool to identify patients with non-CNS infection, exhibiting a 98% positive predictive value and 99% specificity.

Conclusions

This study identifies factors associated with a low risk of CNS infection. Thus, a more precise clinical approach could help clinicians to detect patients who would not benefit from a LP.