TY - JOUR
T1 - Diagnostic Accuracy of Clinical Signs and Biochemical Parameters for External Ventricular CSF Catheter-Associated Infection
AU - Dorresteijn, Kirsten R.I.S.
AU - Verheul, Rolf J.
AU - Ponjee, Gabriëlle A.E.
AU - Tewarie, Rishi Nandoe
AU - Müller, Marcella C.A.
AU - Van De Beek, Diederik
AU - Brouwer, Matthijs C.
AU - Jellema, Korné
N1 - Funding Information: The Article Processing Charge was funded by the authors. Publisher Copyright: © American Academy of Neurology.
PY - 2022/8/1
Y1 - 2022/8/1
N2 - Background and ObjectivesFew prospective well-designed diagnostic accuracy studies have been performed to study the parameters of infection in patients suspected for external ventricular catheter-associated infection. Our objective was to analyze the diagnostic accuracy of clinical characteristics and biochemical and microbiological parameters in diagnosing external ventricular CSF catheter-associated infection.MethodsFrom 2014 to 2017, we performed a single-center cohort study in consecutive patients at the intensive care unit who required an external ventricular CSF catheter in the Hague, the Netherlands. CSF was sampled and analyzed daily. Ventricular catheter-associated infection was defined according to the 2017 Infectious Diseases Society of America's Clinical Practice Guidelines. We compared clinical characteristics and biochemical parameters between patients with and without infection from 3 days before to 3 days after the day the CSF sample was collected that grew bacteria.ResultsA total of 103 patients were included of whom 15 developed a catheter-associated infection (15%). The median day cultures were positive was 3 days after CSF collection (interquartile range [IQR] +2 to +4). On day 0, none of the tests could differentiate between patients with and without infection. The CSF leukocyte count was increased in patients with ventricular catheter-associated infection as compared with patients without on days +2 and +3. The difference was most prominent on day +2 (1,703 × 106/L [IQR 480-6,296] vs 80 × 106/L [IQR 27-251]; p < 0.001; area under the curve [AUC] 0.87 [95% confidence interval (CI) 0.71-1.00]). Sensitivity for the CSF leukocyte count at a cutoff level >1,000 × 106/L was 67% (95% CI 30-93), and specificity was 100% (95% CI 90-100); the positive predictive value was 100%, and the negative predictive value was 92% (95% CI 83-97). The percentage of polymorphonuclear cells (PMNs) was higher in patients with infection on days +1 and +2 (day +2 89% [IQR 78-94] vs 59% [IQR 39-75]; p < 0.01; AUC 0.91 [95% CI 0.81-1.0]).DiscussionAn elevated CSF leukocyte count and increased percentage of PMNs are the strongest indicators for external catheter-associated infections on the days before culture positivity. New CSF markers of drain-associated infection should be studied to enable earlier diagnosis and treatment in patients with an infection and reduce antibiotic treatment in those with no infection.Classification of EvidenceThis study provides Class I evidence that in individuals requiring an external ventricular CSF catheter, an elevated CSF leukocyte count and an increased percentage of PMNs are the strongest indicators of catheter-associated infections in the days before CSF culture positivity.
AB - Background and ObjectivesFew prospective well-designed diagnostic accuracy studies have been performed to study the parameters of infection in patients suspected for external ventricular catheter-associated infection. Our objective was to analyze the diagnostic accuracy of clinical characteristics and biochemical and microbiological parameters in diagnosing external ventricular CSF catheter-associated infection.MethodsFrom 2014 to 2017, we performed a single-center cohort study in consecutive patients at the intensive care unit who required an external ventricular CSF catheter in the Hague, the Netherlands. CSF was sampled and analyzed daily. Ventricular catheter-associated infection was defined according to the 2017 Infectious Diseases Society of America's Clinical Practice Guidelines. We compared clinical characteristics and biochemical parameters between patients with and without infection from 3 days before to 3 days after the day the CSF sample was collected that grew bacteria.ResultsA total of 103 patients were included of whom 15 developed a catheter-associated infection (15%). The median day cultures were positive was 3 days after CSF collection (interquartile range [IQR] +2 to +4). On day 0, none of the tests could differentiate between patients with and without infection. The CSF leukocyte count was increased in patients with ventricular catheter-associated infection as compared with patients without on days +2 and +3. The difference was most prominent on day +2 (1,703 × 106/L [IQR 480-6,296] vs 80 × 106/L [IQR 27-251]; p < 0.001; area under the curve [AUC] 0.87 [95% confidence interval (CI) 0.71-1.00]). Sensitivity for the CSF leukocyte count at a cutoff level >1,000 × 106/L was 67% (95% CI 30-93), and specificity was 100% (95% CI 90-100); the positive predictive value was 100%, and the negative predictive value was 92% (95% CI 83-97). The percentage of polymorphonuclear cells (PMNs) was higher in patients with infection on days +1 and +2 (day +2 89% [IQR 78-94] vs 59% [IQR 39-75]; p < 0.01; AUC 0.91 [95% CI 0.81-1.0]).DiscussionAn elevated CSF leukocyte count and increased percentage of PMNs are the strongest indicators for external catheter-associated infections on the days before culture positivity. New CSF markers of drain-associated infection should be studied to enable earlier diagnosis and treatment in patients with an infection and reduce antibiotic treatment in those with no infection.Classification of EvidenceThis study provides Class I evidence that in individuals requiring an external ventricular CSF catheter, an elevated CSF leukocyte count and an increased percentage of PMNs are the strongest indicators of catheter-associated infections in the days before CSF culture positivity.
UR - http://www.scopus.com/inward/record.url?scp=85140574323&partnerID=8YFLogxK
U2 - https://doi.org/10.1212/CPJ.0000000000200059
DO - https://doi.org/10.1212/CPJ.0000000000200059
M3 - Article
C2 - 36382125
SN - 2163-0402
VL - 12
SP - 298
EP - 306
JO - Neurology. Clinical practice
JF - Neurology. Clinical practice
IS - 4
ER -