Lumbar puncture in infants and children — formative SAQs

SAQ 1 (10 marks)

A 4-month-old previously well infant presents with a 12-hour history of fever, irritability, and poor feeding. On examination the temperature is 38.9°C, heart rate 170, capillary refill 2 seconds, the anterior fontanelle is full but not bulging, and there are no focal neurological signs. The team is planning a septic workup including a lumbar puncture. [1] [5]

1. Outline the structured pre-lumbar-puncture assessment you will perform at the bedside before draping. (4) [1]
2. Describe the positioning and needle selection that maximise the chance of a successful diagnostic lumbar puncture and a reliable opening pressure in this infant. (3) [2] [3]
3. List the CSF tests you will request on the four collection tubes, and state the normal CSF values that apply to this age group. (3) [1]

Model answer — SAQ 1

(1) Pre-LP assessment (4). I run a structured pre-LP check asking four questions. First, raised pressure: I assess the conscious state using a paediatric GCS, examine the pupils and look for any focal neurology, and examine the fontanelle and (where feasible) the fundi for papilloedema; the full fontanelle here without focal signs does not mandate CT, but I would image first if any of these were abnormal. Second, stability: I confirm the observations show no shock, no Cushing's triad, and no respiratory compromise; the capillary refill of 2 seconds and the perfusion are acceptable. Third, bleeding: I ask about a bleeding history, check a platelet count and INR where relevant, and confirm the practical thresholds of platelets above 50 × 10⁹ per litre and INR under 1.5. Fourth, skin: I inspect the puncture site for cellulitis or a spinal dysraphism. I also obtain and document consent, plan the analgesia (topical anaesthesia plus oral sucrose and a parent holding the infant), and confirm the right child and right procedure. [1] [4]

(2) Positioning and needle selection (3). I position the infant in the left lateral decubitus position curled into a fetal position with hips and knees flexed, taking care not to over-flex the neck because an infant can obstruct their airway; I keep the shoulders and hips perpendicular to the bed because rotation narrows the interspinous space and is a common cause of a failed LP. I select an atraumatic pencil-point spinal needle (Whitacre or Sprotte) of an appropriate gauge (a 22-gauge paediatric needle in this age group), which reduces the rate of post-dural puncture headache compared with a cutting Quincke needle without reducing success. The lateral decubitus position is the one that allows a reliable opening pressure, read with the child calm and the legs extended at the moment the manometer is read. [2] [3]

(3) CSF tests and normal values for this age (3). I collect four numbered tubes and send them for: tube 1, cell count and differential; tube 2, protein and glucose with a paired plasma glucose taken at the same time; tube 3, Gram stain, microscopy and culture, with bacterial and viral multiplex polymerase chain reaction if clinically indicated; tube 4, any additional studies including viral PCR, cytology, or mycobacterial studies as indicated. For this 4-month-old the accepted normal CSF values are a white cell count under 5 per microlitre in the older infant (with a higher tolerance in the neonate, up to 20 to 30), protein around 0.15 to 0.45 g/L (higher accepted in the neonate), and a CSF glucose above 60% of a paired serum glucose (typically over 2.5 mmol/L). [1]

SAQ 2 (10 marks)

A 3-year-old presents with fever, irritability, and a decreased conscious state. Five minutes after arrival the team decides to perform a lumbar puncture for suspected meningitis. The child then becomes harder to rouse, develops an irregular breathing pattern, and the right pupil is larger than the left. [4] [5]

1. What has happened, and what is the correct immediate management? (4) [4]
2. Explain why a lumbar puncture should not be performed in this child, and the principles that protect the child from this complication. (3) [4]
3. Outline the implications of having given a dose of empirical antibiotics for the diagnostic value of a later CSF sample. (3) [5]

Model answer — SAQ 2

(1) Diagnosis and immediate management (4). This child has developed clinical signs of cerebral herniation — a falling conscious state, irregular respirations, and an asymmetric pupil indicate impending or established brainstem compression. I abandon the planned LP immediately. I call for senior paediatric and intensive-care help, secure the airway and breathing with elective intubation if the conscious state is sufficiently depressed, keep the head midline and the bed head-up to 30 degrees, arrange an urgent CT head, and treat the raised intracranial pressure with hypertonic saline (3%) or mannitol per the local protocol while arranging neurosurgical and intensive-care input. The LP is now contraindicated until imaging and stabilisation are complete. [4]

(2) Why LP should not be performed, and the protective principles (3). A lumbar puncture in a child with raised intracranial pressure from any cause — here, presumed cerebral oedema or a space-occupying lesion — risks precipitating cerebral herniation by creating a pressure gradient: draining CSF from the lumbar theca lowers the pressure below the foramen magnum, drawing the cerebellar tonsils downward. The protective principles are the structured pre-LP check (raised pressure, neurology, stability, bleeding, skin) and the imaging-first rule: a CT head before LP whenever there is focal neurology, papilloedema, a decreased or fluctuating conscious state, a prolonged postictal state, immunocompromise, or a known CNS lesion or shunt. The pre-LP check is designed to make this complication preventable. [4]

(3) Effect of prior antibiotics on later CSF (3). The dominant myth is that one dose of antibiotic renders the CSF uninterpretable, and this is incorrect. CSF cell counts, glucose, protein and Gram stain remain informative for 24 to 48 hours after a dose of antibiotic, and bacterial and viral PCR are barely affected. The test most affected is bacterial culture, which may be negative after effective antibiotic exposure; the implication is that a deferred LP is still diagnostically useful and empirical treatment should never be delayed to obtain the LP. In suspected bacterial meningitis the rule is antibiotics first, LP when the child is stable. [5]