Abstract
PURPOSE: Extremely premature neonates diagnosed with post-haemorrhagic hydrocephalus (PHH) are recognised to have particularly poor outcomes. This study assessed the impact of a number of variables on outcomes in this cohort, in particular the choice of shunt valve mechanism. METHODS: Electronic case notes were retrospectively reviewed of all premature neonates admitted to our centre for management of hydrocephalus between 2012 and 2021. Data included (i) gestational age, (ii) birth weight, (iii) hydrocephalus aetiology, (iv) surgical intervention, (v) shunt system, (vi) 'surgical burden' and (vii) wound failure and infection rate. Data was handled in Microsoft Excel and statistical analysis performed in SPSS v27.0 RESULTS: N = 53 premature hydrocephalic patients were identified (n = 28 (52.8%) female). Median gestational age at birth was 27 weeks (range: 23-36 + 6 weeks), with n = 35 extremely preterm patients and median birth weight of 1.9 kg (range: 0.8-3.6 kg). Total n = 99 programmable valves were implanted (n = 28 (28.3%) de novo, n = 71 (71.2%) revisions); n = 28 (28.3%) underwent n ≥ 1 pressure alterations, after which n = 21 (75%) patients had symptoms improve. In n = 8 patients exchanged from fixed to programmable valves, a mean reduction of 1.9 revisions per patient after exchange was observed (95%CI: 0.36-3.39, p = 0.02). Mean overall shunt survival was 39.5 weeks (95%CI: 30.6-48.5); 33.2 weeks (95%CI: 25.2-41.1) in programmable valves and 35.1 weeks (95%CI: 19.5-50.6) in fixed pressure (p = 0.22) with 12-month survival rates of 25.7% and 24.7%, respectively (p = 0.22). Shorter de novo shunt survival was associated with higher operation count overall (Pearson's R: - 0.54, 95%CI: - 0.72 to - 0.29, p < 0.01). Wound failure, gestational age and birth weight were significantly associated with shorter de novo shunt survival in a Cox regression proportional hazards model; gestational age had the greatest impact on shunt survival (Exp(B): 0.71, 95%CI: 0.63-0.81, p < 0.01). CONCLUSION: Hydrocephalus is especially challenging in extreme prematurity, with a shorter de novo shunt survival associated with higher number of future revisions. Programmable valves provide flexibility with regard to pressure setting, with the potential for fewer shunt revisions in this complex cohort.