Repeat Lumbar Puncture at 24 Versus 48 Hours After Traumatic Lumbar Puncture in Neonates

Repeating a Lumbar Puncture at 24 Hours Versus 48 Hours After a Macro-traumatic Lumbar Puncture in Newborn Infants: a Randomized Controlled Trial

Visibly traumatic as well as microtraumatic lumbar punctures (LP) are very common in the neonatal period. The presence of blood makes it difficult to interpret cerebro-spinal fluid (CSF) findings. Clinicians often perform a repeat LP in the hope that some of the red blood cells would have cleared by then, allowing a better interpretation of the CSF findings. There is no published information whether a repeat LP provides any added information to the original traumatic LP, and if so what is the best time to repeat an LP after a traumatic LP. In this randomised controlled trial (RCT), we plan to randomly allocate neonates following a visibly traumatic LP to either undergo a repeat LP at 24 hours or 48 hours later to determine which LP gives more accurate results.

STUDY DESIGN An open-label Randomised Control Trial, with stratification and blocking. STUDY SETTING The Newborn unit and Neonatal Unit of Pediatric Emergency of Post graduate Institute of medical education and research (PGIMER). The former is a level 3 Neonatal Intensive Care Unit (NICU). STUDY PERIOD Mentioned elsewhere ELIGIBILITY CRITERIA Mentioned elsewhere CONSENT A parent information sheet was provided. Informed written consent was obtained from the parents of the patient following which subjects were included in the study. The consent included information about the procedure and study related to it. It also included the risk factors associated with and after the procedure. All subjects were included after obtaining wilful consent from respective parents. PROCEDURE Post enrolment uncorrected White blood cell (WBC), corrected WBC count, red blood cell (RBC) count, Glucose concentration, Protein concentration, CSF Procalcitonin concentration, Gram stain and culture and sensitivity of microorganism was performed. Post-LP the subject was randomized and repeat LP was performed either after 24 hours or 48 hours, and uncorrected WBC, corrected WBC count, RBC count, Glucose concentration, Protein concentration, Gram stain, and Culture and sensitivity of microorganism was studied. BASELINE PARAMETERS Gestation, Birth weight, Sex, Postnatal age, Current weight, Indication for initial lumbar puncture in the neonate, Seniority level of the person doing the initial lumbar puncture, Platelet count when available before initial lumbar puncture, The coagulation study report, if available, before initial lumbar puncture and repeat lumbar puncture, Whole blood glucose measured by glucometer before the initial LP. Parameters measured in the initial lumbar puncture o Uncorrected and corrected WBC count: Cells were counted using two methods. One was in Neubauer chamber within 30 minutes of performing an LP. 60 μL of CSF was split in 2 Eppendorf tubes, labeled A and B. 27 μL of CSF in the first Eppendorf tube i.e labeled A was diluted with Methylene Blue using lab standard micropipettes and was used to charge the chamber. Cell count of both WBC and RBC were performed as standard procedure. Another 27 μL CSF from Eppendorf tube labeled B was diluted using 3 μL Turk's fluid with composed of Gentian Violet and Glacial Acetic. The Turk's fluid essentially lyses the RBCs. Following which one side of the chamber is charged and WBCs are calculated in the light grey squares (n). The WBC count per microlitre = n x 25/9. For RBC count - 3 μL of Methylene Blue was added to 27 μL of CSF to make 30 μL. One side of the chamber was charged and RBCs were counted on the dark grey squares (n). The total RBCs counted per microliter = n x 500/9. As mentioned WBCs and RBCs were also counted in automated machines available in the Main Hematology lab as well as the emergency lab. The fully automated machine analyzer is the Sysmex XN-1000(Sysmex-Kobe, Japan). The XN series analyzer has research mode which uses fluorescence flow cytometry and impedance measurements to count the number of WBCs and RBCs in CSF. To measure RBCs, both the optical(for lower concentration ranges) and impedance (for higher concentration) techniques were used. In the WBC Differential Fluorescence channel, the total WBC and differential WBC counts are based on flow cytometry measurements (e.g side scatter, forward scatter, and DNA/RNA fluorescence). Before analysis using the XN hsA research mode, the sample didn't require any preparation. Glucose concentration- Glucose concentration in CSF was estimated using a colorimetry-based method after enzymatic oxidation in the presence of glucose oxidase but didn't break down the protein. The hydrogen peroxide produced during the reaction reacts with phenol and 4-aminophenazone to form a red-violet quiononimine dye which acts as an indicator. The intensity of the color that is produced is directly proportional to the glucose concentration and is measured at a wavelength of 505 nm. The machine model is Adiva 1800 from Siemens. Protein concentration- This is also a quantitative determination by a photometric color test performed in a Beckman Coulter analyzer. Pyrogallol red is combined with molybdate to form a red complex with maximum absorption at a wavelength of 470 nm. The assay is based on the shift absorption that occurs when the pyrogallol red-molybdate complex binds basic amino acid groups of protein molecules. A blue purple complex is formed with a maximum absorbance at a wavelength of 600 nm. The absorbance of this complex is directly proportional to the protein concentration in the sample. Procalcitonin concentration - CSF procalcitonin was determined using a two-step sandwich electrochemiluminescence immunoassay using biotin-labeled procalcitonin specific antibody and a monoclonal procalcitonin specific antibody labeled with a Ruthenium complex (Roche Diagnostics, Germany). 200 μL of the CSF sample was incubated at 37 degrees C with the Biotin and Ruthenium labeled anti- procalcitonin antibodies. In the second incubation, streptavidin-coated anti microparticles were added which binds the complex via the biotin-streptavidin interaction. The reaction mixture is aspirated into a measuring cell where microparticles are magnetically captured onto the surface of the electrode. Unbound substances are removed by washing and upon application of a voltage to the electrode, chemiluminescence would be induced which is measured by a photomultiplier. The measuring range of the assay is 2 ng/dl - 10000ng/dl with a functional sensitivity of 6ng/dl and analytical sensitivity of 2 ng/dl. Gram stain of traumatic lumbar puncture Culture and sensitivity of traumatic lumbar puncture As per unit policy, CSF cell count was performed within 30 minutes of LP in the microscope in the NICU side lab by personnel trained in CSF cell count. This was done to avoid degeneration of WBCs and RBCs and drop in WBC count due to delay in the analysis. RANDOMISATION SEQUENCE Eligible subjects were enrolled within less than 24 hours after the initial LP. A block randomization sequence was generated from the website www.randomization.com. Permuted, even-numbered, randomly varying block sizes with a 1:1 allocation ratio were used in this study. ETHICS COMMITTEE APPROVAL The study was approved by The Institutional Ethics Committee of PGIMER, Chandigarh (INT/IEC/2019/000557) POST-RANDOMIZATION EXCLUSIONS Mentioned elsewhere CONCEALMENT OF ALLOCATION The random allocation sequence was concealed using serially numbered opaque sealed envelopes which contained slips of a paper, mentioning a group of randomization. After enrolment, the name of the patient was written on the envelope. Only after writing the name, the envelope was torn and the allocated intervention inside the envelope was implemented subsequently. INTERVENTION GROUPS Subjects were randomly allocated to 2 groups. One group consisted of patients where a repeat LP was done 24 hours after the initial traumatic LP. The 2nd group consisted of patients where LP was repeated at 48 hours after the initial traumatic LP. BLINDING It was not possible to blind doctors and nurses working in the NICU. However, Microbiology, Hematology and Biochemistry lab personnel were blinded to the group of allocation. OUTCOMES MEASURED IN REPEAT LP: Blood glucose before doing LP, Whether CSF sample obtained, CSF sample: whether frank blood, visibly blood-tinged, micro-traumatic, or nontraumatic, CSF RBC count, CSF WBC count (uncorrected and corrected), CSF glucose, CSF protein, CSF Gram stain, culture, and sensitivity. Similar to the first LP, CSF cell count was performed within 30 minutes of LP in the microscope in the NICU side lab by personnel trained in CSF cell count. REFERENCE STANDARD Definite meningitis CSF culture and/or gram stain positive Probable meningitis CSF procalcitonin (PCT) positive. To define CSF PCT positive, we took a cut-off of 33 ng/dL or more. This was based upon the study by Reshi et al(18). The study was conducted in the Neonatal Intensive care unit of the Department of Neonatology and Pediatrics at SKIMS, Srinagar, over a period of 2 years (January 2014 to December 2015). The study was a prospective study that included infants 28 days old, who qualified LP as a part of sepsis workup. They defined traumatic lumbar puncture as a CSF sample with ⩾500 red blood cells/mm3, and leukocyte count adjustments were done at an average ratio of 500 red blood cells to 1 white blood cells. The diagnosis of bacterial meningitis was made in neonates with sepsis with either positive CSF and/or blood culture or positive gram staining. They excluded infants who were more than 28 days old or neonates who had received antibiotics prior to hospital admission, neonates who had undergone recent brain surgery if another focus of infection was present in addition to meningitis or Viral meningitis (herpes simplex virus). Post exclusion criteria the study was concluded on 168 neonates. They observed, at a value of 0.33ng/ml, CSF PCT has a sensitivity of 92%, a specificity of 87%, a positive predictive value (PPV) of 85.2%, and a negative predictive value (NPV) of 93%. We in our study didn't include other CSF conventional parameters for probable meningitis, such as CSF WBC, CSF Sugar, and CSF Protein for the definition of probable meningitis because we intended to study these tests as index tests, and hence they could not also be part of the reference standard. COMPARISON OF OUTCOME MEASURES Mentioned elsewhere

Newborn infants in this group will undergo repeat lumbar puncture at 24 hours after a traumatic lumbar puncture

Newborn infants in this group will undergo repeat lumbar puncture at 48 hours after a traumatic lumbar puncture

Area under the the receiver operating characteristic (ROC) curve of uncorrected CSF WBC count in the repeat LP, with "definite or probable meningitis" taken as the reference standard.

Primary outcome- area under the ROC curve of uncorrected CSF WBC count in the repeat LP- will be compared between the 24 hour LP and the 48 hour LP groups. The reference standard will be "definite or probable meningitis" defined as "CSF culture and/or CSF Gram stain and/or CSF Procalcitonin positive" in the baseline traumatic LP.

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with "definite or probable meningitis" taken as a reference standard

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with "definite or probable meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as in the primary outcome.

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with only "definite meningitis" taken as a reference standard

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with only "definite" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as "CSF culture and/or Gram stain positive".

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with only "probable meningitis" taken as a reference standard

Area under the ROC curve of corrected CSF WBC count (using 500:1 formula) in the repeat LP, with only "probable" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as "CSF Procalcitonin positive".

Area under the ROC curve of CSF glucose in the repeat LP, with "definite or probable meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "definite or probable meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as under primary outcome.

Area under the ROC curve of CSF glucose in the repeat LP, with only "definite meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "definite meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as mentioned earlier.

Area under the ROC curve of CSF glucose in the repeat LP, with only "probable meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "probable meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as mentioned earlier.

Area under the ROC curve of CSF protein in the repeat LP, with "definite or probable meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "definite or probable meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as mentioned earlier.

Area under the ROC curve of CSF protein in the repeat LP, with only "definite meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "definite meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as mentioned earlier.

Area under the ROC curve of CSF protein in the repeat LP, with only "probable meningitis" taken as a reference standard

Area under the ROC curve of CSF glucose in the repeat LP, with "probable meningitis" taken as the reference standard will be compared between the 24-hour LP and the 48-hour LP groups. Reference standard will be defined as mentioned earlier.

Change in the CSF RBC count from the baseline traumatic LP to the repeat LP, expressed as a percentage of the baseline value, will be compared between the 24-hour LP group and the 48-hour LP group

Change in the uncorrected CSF WBC count from the baseline traumatic LP to the repeat LP, expressed as a percentage of the baseline value, will be compared between the 24 hour LP group and the 48 hour LP group

Inclusion Criteria: Newborn infants admitted anywhere in the Neonate unit or Neonatal Unit of Paediatric Emergency Underwent lumbar puncture as part of the workup for sepsis Cerebrospinal fluid visibly tinged with blood (macro traumatic). To ensure objectivity, two Pediatricians must have a consensus that the CSF sample is visibly blood-tinged when viewed in bright ambient light. Exclusion Criteria: Clinically too unstable to undergo 2 lumbar punctures within the span of 48 hours Lumbar puncture performed after more than 3 doses of antibiotics administered CSF sample of traumatic lumbar puncture not sent for analysis of WBC, RBC counts, glucose, protein, Gram stain, and culture Sample appears to be pure blood from the initial needle prick. Post-randomization exclusion Clinically too unstable to undergo the 2nd LP Development of fresh contraindications for performing the 2nd LP