Utility of biomarkers to predict bacterial meningitis in elderly patients

Dear Editor:

It was with great interest that we read the study by Morales Casado et al.¹ recently published in Neurología and addressing the utility of biomarkers of infection and inflammation (BII; procalcitonin [PCT] and C-reactive protein [CRP]) for predicting bacterial meningitis in the emergency department. Although we agree that both BII have a good predictive ability for detecting bacterial meningitis and differentiating bacterial from viral meningitis,² we would like to add further comments regarding the subgroup of patients older than 75. The utility of BII, particularly CRP, in elderly patients is still controversial due to immunosenescence, a process that alters immune response and especially inflammatory response. Elderly patients show higher levels of proinflammatory cytokines (for example, IL-6 and IL-8) against bacterial infection. Likewise, CRP determinations in this population yield a higher rate of false-positive results and have lower sensitivity and specificity for differentiating bacterial from viral infections.^2,3

As the authors pointed out, although acute meningitis is not ranked among the most frequent infections in the emergency department,⁴ it is one of the most severe, as well as one of the types of infection most frequently leading to severe sepsis and septic shock.⁵ In the past few years, the incidence of infections in emergency departments has increased significantly (P < .001) among the elderly (31.7% vs 24.8%).⁴ Clinical severity and mortality in this subgroup have also increased.⁶ In this population, suspecting and confirming bacterial meningitis is especially important given the severity of this process and the largely non-specific, highly variable symptoms associated with infectious processes.³ This results in delayed diagnosis in these patients and leads to administering unnecessary antimicrobial drugs in more than 50% of the cases.⁷ In the study by Morales Casado et al.,¹ PCT achieved an excellent diagnostic power for detecting bacterial meningitis, with an area under the ROC curve (AUC) of 0.99, 94% sensitivity, and 100% specificity for a cut-off point of 0.74 ng/mL. CRP was also found to have a good diagnostic power, with an AUC of 0.91; however the cut-off point was higher (90 mg/L) and both sensitivity (67%) and specificity (86%) were lower. We hypothesise that CRP results may be overestimated in this study and reflect the mean age in their sample (44 years). Published evidence shows that: (1) 50-65% of all patients older than 75 have high CRP levels in the absence of underlying infections (false positives)^2,8 upon arriving at the emergency department, and (2) the diagnostic power of CRP decreases in older populations, whereas that of PCT remains the same, as in cases of bacteraemia (including cases of associated bacterial meningitis).⁹ We therefore feel that CRP determinations must be interpreted with caution in elderly patients, especially in severe cases such as suspected bacterial meningitis. Over 50% of the patients with fever in emergency departments undergo CRP tests (but not PCT tests)¹⁰ to differentiate between viral and bacterial meningitis, and more than 40% of all infectious processes in emergency departments affect elderly patients. The above suggests that isolated CRP measurements have a very limited diagnostic power in the emergency department for confirming or ruling out a bacterial aetiology of acute meningitis (unlike PCT). We need specific studies analysing the diagnostic power of BII by age group to confirm that the diagnostic ability of CRP is similar to that of PCT in children, decreases in adults, and is limited in elderly patients.²

Conflicts of interest

The authors have no conflicts of interest to declare.