Plasma exchange for Guillain-Barré syndrome.

The impact of plasma exchange (PE) on the treatment of severe Guillain-Barré syndrome (GBS) was studied by comparing all the 16 patients treated with PE in a London teaching hospital between January 1985 and August 1987, with 64 historical controls drawn from a series of patients observed during a prospective study in South East England in 1983 and 1984. There were no GBS-related deaths in the PE treated group but seven in the historical controls (2P = 0.39). The median duration of ventilation was only 20 days (range 7-64) in the PE group compared with 36 days (range 14-365) in the surviving patients in the control group (2P = 0.06, 95% confidence interval of difference in medians -36 to 0 days). The PE group walked earlier without aid (median 55.5 compared with 86 days, 2P = 0.04, 95% confidence interval of difference in medians -88 to -2 days). Three months after the onset of neuropathy all PE treated patients were able to walk with support or better, whereas 19 of the surviving historical controls were unable to walk even with support (2P = 0.009). The costs of PE were offset by savings in intensive care unit and hospital expenditure.