Introduction: Assessment of the pressure-volume index (PVI) during lumbar infusion study (LIS) has been proposed to evaluate the overall compliance of the cranio-spinal system. It is calculated from the measurement of CSF pressure changes, ΔP from Pb to Pp, in response to repeated bolus injections of a volume (ΔV) within the lumbar subarachnoid space.
Material and methods: We retrospectively analyzed 18 patients who underwent LIS for suspicion of normal pressure hydrocephalus, including a series of three fast bolus injections of 3 mL of saline at different levels of CSF pressure. We compared two methods for PVI calculation: (a) PVIslope using the slope α of a linear fit ΔP = α(Pb – P 0), PVI = ΔV/log10(α + 1); (b) PVImean using the PVI calculated independently for each bolus injection assuming P 0 = 0, PVI = mean(ΔV/log10(Ppi/Pbi))i=1.3.
Results: We found a significant discrepancy between the two methods: the average difference (PVIslope – PVImean) was -3.93 mL (95% confidence interval [8.77; -16.64]). In the PVIslope, method, the mean P 0 was 2.12 mmHg (±3.41 mmHg).
Discussion: The clinical reliability of PVImean (assuming P 0 = 0) depends on the value of P 0. PVIslope provides results, independent of P 0. Future studies should focus on determining pathological PVI range rather than fixed cut-off values.