Study protocol: NeoCLEAR: Neonatal Champagne Lumbar punctures Every time - An RCT: a multicentre, randomised controlled 2 x 2 factorial trial to investigate techniques to increase lumbar puncture success

Background The neonatal period carries the highest risk of bacterial meningitis (similar to 1 in 5000 births), bearing high mortality (similar to 10%) and morbidity (20-50%) rates. Lumbar puncture (LP) remains essential to the diagnosis of meningitis. Though LP is a common procedure in neonates, success rates are lower (50-60%) than in other patient populations. None of the currently-practised neonatal LP techniques are supported by evidence from adequately-powered, randomised controlled trials (RCTs). NeoCLEAR aims to compare two modifications to the traditional technique which are free, accessible, and commonly practised: sitting (as opposed to lying) position, and 'early' (as opposed to 'late') stylet removal. Methods/design Written parental informed consent permitting, infants in neonatal/maternity wards, of 27(+ 0) to 44(+ 0) weeks corrected gestational age and weighing >= 1000 g, who require an LP, will be randomly allocated to sitting or lying position, and to early or late stylet removal. The co-primary objectives are to compare success rates (the proportion of infants with cerebrospinal fluid red cell count < 10,000/mm(3) on first LP procedure) in 1020 infants between the two positions, and between the two methods of stylet removal. Secondary outcomes relate to LP procedures, complications, diagnoses of meningitis, duration of antibiotics and hospital stay. A modified intention-to-treat analysis will be conducted. Discussion Two modifications to the traditional LP technique (sitting vs lying position; and early vs late stylet removal) will be simultaneously investigated in an efficient and appropriately-powered 2 x 2 factorial RCT design. Analysis will identify the optimal techniques (in terms of obtaining easily-interpretable cerebrospinal fluid), as well as the impact on infants, parents and healthcare systems whilst providing robust safety data. Using a pragmatic RCT design, all practitioners will be trained in all LP techniques, but there will inevitably be variation between unit practice guidelines and other aspects of individual care. An improved LP technique would result in: center dot Fewer uninterpretable samples, repeated attempts and procedures center dot Reduced distress for infants and families center dot Decreased antibiotic use and risk of antibiotic resistance center dot Reduced healthcare costs due to fewer procedures, reduced length of stay, shorter antibiotic courses, and minimised antibiotic-associated complications