A clinical audit of the Hakim programmable valve in patients with complex hydrocephalus.

The objective of the investigation was to determine the pattern of use of the Hakim (Medos) programmable valve implanted in patients with complex hydrocephalus and their clinical outcome. A prospective audit of patients with complex hydrocephalus undergoing Hakim programmable valve implantation between 1989 and 1994 in the United Kingdom and Ireland, was followed-up for a minimum of 5 years. Surgical practice and complications were audited together with clinical outcome. One-hundred-and-thirty-nine patients (80 male, 59 female; mean age 43.4 years; median 47 years; range 1 month-84 years) with complex hydrocephalus due to a wide range of aetiologies were implanted with the Hakim programmable valve. Eighty-eight (63%) had large or massive ventricles prior to implantation; seven (5%) were slit. Fifty-five (40%) had previously been shunted with a fixed pressure system. One-hundred-and-thirty-one (94%) of the Hakim programmable shunts were ventriculoperitoneal; four (3%) ventriculoatrial; two (1.4%) cystoperitoneal; and two (1.5%) lumboperitoneal. The initial opening pressure selected ranged from 50 to 200 mmH2O (median 120). Valves were reprogrammed on average 1.7 times with 143 reprogrammings in the first year after implantation; 67 in the second; 19 in the third; three in the fourth; two in the fifth. Forty-nine (36%) valves were never reprogrammed after implantation. During the 5 years audit period, there were 70 (50%) shunt revisions, 40 of which were performed within 1 year of implantation. Thirty-six (27%) shunts were removed. There were 24 (18%) shunt infections. Subdural collections were identified in 37(27%) patients after Hakim programmable valve implantation; 10 (27%) required surgical drainage. Five (3.7%) patients developed symptomatic slit ventricles after Hakim programmable valve implantation. Headache was improved following reprogramming in 27(71%) of the 38 patients with refractory headache. After Hakim programmable valve implantation, patients underwent an average of 4.6 CT scans (range 1-25); 0.3 MRI (range 1-5) and 1.8 skull radiographs (range 1-20). The mean hospital stay per patient over 5 years was 26 days (range 1-110 days). Five years after implantation, the Glasgow Outcome scale was favourable in 64% of patients. The Hakim programmable valve is useful in the management of patients with complex hydrocephalus and may reduce the need for shunt revision for headache. Non-haemorrhagic, post-shunting, subdural collections identified on routine postoperative CT may be treated by reprogramming.