Maintenance Haemodialysis-Daly and Hassall BMEDICSAL JORNAL ferences between patients treated at home or in lhospital, by day or at night. Furthermore, the accumulation of metabolites between dialyses has no apparent effect on this measure.The findings concerning home dialysis are of particular importance, since the spouse provides one form of control. It is seen that on dialysis nights both patients and spouses lose sleep, but the patients regain their "sleep debt" more than their spouses on non-treatment nights. This suggests a psychological rather than a metabolic cause of insomnia on dialysis. It would add to the evidence showing stress on families of patients being treated with dialysis, particularly where the treatment is conducted in the home. This stress on the spouse would seem understandable in terms of their feelings about the patient's illness, and having responsibility for operating a complex life-support system. But, obviously, other criteria must be considered in deciding on the location for treatment-for example, the additional time the home dialysis patient spends with his family.Regarding the reliability of these data, Lewis (1969) investigated the accuracy of subjective estimates of sleep with objective (E.E.G.) recordings. He showed that, though subjects are not accurate in their estimates of the quantity of sleep, their estimates do shift in the direction of the objective measures. Thus they are reliable indices of change of sleep. Furthermore, in this study the data from patients in two geographically separate treatment centres showed no significant differences whereas there were significant differenices between spouses and patients.