Objective

The objective of this study is to examine factors accounting for the incidence of traumatic lumbar puncture (TLP) in infants younger than 1 year old.

Study Design

Retrospective analysis of cerebrospinal fluid (CSF) data from 1,240 neonatal (≤28 days) and 399 infant lumbar puncture (LP) procedures was conducted. Data from two successive LP procedures were obtained from 108 patients. Logistic regression analysis was used to assess factors accounting for the incidence of TLP in the second LP procedure. The following categorical variables were entered into the model: whether the first procedure was traumatic according to criteria of ≥500 and ≥10,000 erythrocytes/µL, whether the LP procedures were performed within a week, and whether the patient was neonatal at the first procedure.

Results

The incidences of TLP were 42.9% in neonates and 22.5% in infants for the criterion of ≥500 erythrocytes/µL, and 16.6 and 10.3% for the criterion of ≥10,000 erythrocytes/µL. Compared with a nontraumatic first LP procedure, if the first procedure was traumatic according to the criterion of ≥10,000 erythrocytes/µL, the odds ratio (OR) of TLP in the second procedure was 5.86 (p = 0.006). Compared with a longer time, if the successive procedures were performed within a week, the OR of TLP was 9.06 (p < 0.0001) according to the criteria of ≥500 erythrocytes/µL and 3.34 (p = 0.045) according to the criteria of ≥10,000 erythrocytes/µL. If the patient was neonatal at the first procedure, the OR of TLP at the second puncture was 0.32 (p = 0.031) according to the criterion of ≥500 erythrocytes/µL.

Conclusions

The incidence of TLP in neonates is twice as high as that in infants. Successive LP procedures performed within a week and a highly blood-contaminated CSF sample in the first procedure each multiplied the odds of TLP in the second procedure, whereas being a neonate at the time of the first procedure reduced the odds of TLP.