Relationship between intracranial pressure (ICP) and intracranial pulse amplitude (ICPa)–Derived compliance in severe traumatic brain injury: A retrospective analysis of continuous ICP-ICPa monitoring

Background and Objective: Intracranial pressure (ICP) monitoring remains a keystone of severe traumatic brain injury (TBI) management. However, ICP alone may not reflect intracranial compensatory reserve. Intracranial pulse amplitude (ICPa) derived from ICP waveform analysis has emerged as a potential surrogate marker of intracranial compliance. This study evaluated the relationship between intracranial pressure (ICP) and intracranial pulse amplitude (ICPa) in severe traumatic brain injury (TBI) during continuous invasive monitoring. Exploratory analysis assessed data-derived ICPa values associated with concurrent intracranial hypertension. Methodology: This retrospective analytical study was conducted in the Department of Neurosurgery and Neurocritical Care Unit, Jinnah Hospital, Lahore, from January 2020 to December 2025. Twenty-one severe TBI patients underwent RAUMEDIC Neurovent-P monitoring. After artifact removal, 1,520 paired hourly observations were analyzed. Mixed-effects regression was the primary method. GEE, patient-cluster bootstrap ROC analysis, leave-one-patient-out validation, and exploratory segmented regression were also used. Results: Mean age was 35.62±14.13 years; 71.4% (15) were male, and 28.6% (6) were female. Pearson correlation was r=0.495 (p<0.001), and within-patient correlation was r=0.712 (p<0.001). Mixed-effects analysis showed a 1.393 mmHg rise in ICP per unit ICPa (95% CI 1.323-1.462; p<0.001). GEE showed higher odds of concurrent ICP ≥20 mmHg (OR 1.423; 95% CI 1.282-1.579; p<0.001). Patient-cluster bootstrap median AUC was 0.720 (95% interval 0.627-0.836). The cutoff of 4.3 was unstable on internal validation. Conclusion: ICPa remained positively associated with ICP after repeated measurements were accounted for. It may provide complementary physiological information. The data-derived cutoff is exploratory and requires external validation.