Repeated lumbar puncture in patients with pneumococcal meningitis: practical or anxiolytic relevance?

Abstract number: 1733_703

Alvarez A., Mourvillier B., Bout H., Bruel C., Ferreira L., Lakhal K., Pajot O., Bouadma L., Regnier B., Wolff M.

Introduction: Repeated lumbar puncture (RLP) is recommanded in patients with pneumococcal meningitis who have not responded clinically after 48 h of appropiate antimicrobial therapy or in pneumococcal meningitis caused by penicillin- or cephalosporin-resistant strains (PNSP). Nevertheless the clinical impact of this practice is uncertain.

Patients and Methods: Retrospective analysis of charts of patients admitted in our 25-bed university medical ICU between January 2000 and December 2005, for pneumococcal meninigitis. All patients were appropiately treated according to international guidelines. The following data were collected at admission: clinical features, CSF characteristics: leukocyte count, protein and glucose concentrations, Gram stain and MIC of penicillin G. RLP data and changes in antibiotics dosage after results, were also noted.

Results: Among 51 included patients, 32 (63%) underwent RLP (RLP+) and were compared to 19 (37%) without RLP (RLP-). The mean delay between the first lumbar puncture and RLP was 4±2.5 days. RLP+ patients were older (59±14 vs 48±17 years), more severe at admission (SAPS II: 43±13 vs 29±4); had a more severely altered mental status (GCS 8±3 vs 12±3); and ICU length of stay was longer (23±15 d vs 4±3 d). The number of PNSP strains was 7 and 2 respectively (p = 0.52). Characteristics of the first lumbar puncture were not different in the two groups, except for leucocyte count (1,766±2,190 vs 4,064±4,080, p < 0.05). There was no difference in mortality (9% in RLP+ group vs 5% in RLP- group, p = 1) but a tendency toward a larger number of sequelae in the RLP+ group (34.4% vs 9.5%, p = 0.08). Gram stain was positive in 47% of RLP but cultures were always sterile. The comparison between admission and RLP is shown in the table.

Cefotaxime concentrations in CSF were measured in 14 patients and showed a level >5MIC in 4 and ≥10MIC in 9 patients, respectively. In one case cefotaxime concentration was <5MIC but the outcome was favourable. In none of the patients, the results of the RLP led to modifications of the antibiotic treatment.

 Day 1Repeated lumbar puncturep
Temperature (°C)39.2±1.237.7±1.1<0.05
Protein (g/L)6.7±4.33.3±2.6<0.05
Glucose level (mmol/L)1.1±22.7±2.5<0.05

Conclusion: When internationals treatment guidelines for pneumococcal meningitis are applied, based on pneumococcal resistance epidemiology, bacteriological failure seems improbable. Therefore, RLP may be unnecessary in this context. This hypothesis needs to be confirmed by a larger prospective study.

Session Details

Date: 31/03/2007
Time: 00:00-00:00
Session name: European Society of Clinical Microbiology and Infectious Diseases
Location: ICC, Munich, Germany
Presentation type:
Back to top