The aim of this study was to test the equivalence of once- and thrice-daily dosing with tobramycin by comparing efficacy and safety in adult patients with cystic fibrosis. Sixty adult patients with an acute respiratory exacerbation were randomized to receive either 10 mg·kg−1 tobramycin once-daily or 3.3 mg·kg−1 tobramycin thrice-daily. Primary efficacy and safety endpoints were defined as changes in respiratory function and changes in renal function and hearing. Both groups showed a significant increase in respiratory function without a clinically significant change in renal function. For changes in forced vital capacity % predicted and serum potassium and magnesium levels, equivalence was demonstrated. For the variables forced expiratory volume in one second and forced mid-expiratory flow % pred and serum creatinine levels, there was insufficient power to demonstrate equivalence. One patient in each group showed bilateral impairment in pure tone audiogram after treatment. This study demonstrated significant clinical improvement with both once- and thrice-daily tobramycin dosing. Equivalence between the two regimens was shown for some, but not all primary endpoints. Once-daily dosing should be used with careful monitoring of safety and efficacy until large multicentre studies confirm these encouraging results.

1Royal Dundee tiff Hospital, Dundee, UK 2Aysgarth Statistical Consultancy, 220 Smith House Lane, Lightcliffe, Halifax, West Yorkshire, UK 3Lilly Industries, Dextra Court, Chapel Hill, Basingstoke, Hampshire, UK

In a study in seven large renal units in England, the morbidity and mortality of all patients starting continuous ambulatory peritoneal dialysis (CAPD) and haemodialysis during 1983-85 were monitored prospectively over a 4-year period and related to reasons for choice of therapy and potential risk factors. 610 new patients (median age 52 years, range 3-80 years) started CAPD; 16% had diabetes mellitus and 21% cerebrovascular or cardiovascular disease. 329 patients (median age 48 years, range 5-77 years) started haemodialysis; 7% had diabetes mellitus and 17% cerebrovascular or cardiovascular disease. The Kaplan-Meier patient survival estimates at 4 years were 74% for haemodialysis and 62% for CAPD; technique survival figures for the same period were 91% for haemodialysis and 61% for CAPD. Cox's proportional hazards regression analysis showed that cerebrovascular/cardiovascular disease, age over 60 years, and diabetes mellitus were important predictors for survival in CAPD patients; there were no risk factors associated with permanent change to haemodialysis. In the haemodialysis group early change to CAPD was associated with presence of cerebrovascular or cardiovascular disease. The major cause of drop-out in both groups was transplantation. The mean length of hospital admission was 14.8 days per patient-year for CAPD and 12.4 days per patient-year for haemodialysis.

231 babies, born at less than or equal to 32 weeks' gestation were enrolled in a randomised, controlled trial to assess the efficacy of vitamin E (dl-alpha-tocopherol acetate) in the prevention of periventricular haemorrhage. Daily supplementation with 20 mg/kg vitamin E intramuscularly during the first 3 days of life was associated with a rise in plasma vitamin E concentration and a reduction in hydrogen peroxide haemolysis of red blood cells in vitro. Among babies without haemorrhage on entry to the trial (n = 210), supplemented babies had a lower frequency of intraventricular haemorrhage than controls (8.8% v 34.3%; p less than 0.005) and a lower combined frequency of intraventricular and parenchymal haemorrhage (10.8% v 40.7%; p less than 0.0001) on the final ultrasound brain scan. This protective effect was observed in both inborn and referred babies but was stronger in the former. Supplementation had no effect on mortality, but among survivors fewer supplemented babies than controls had intraventricular or parenchymal haemorrhage (10.7% v 32.6%; p less than 0.001). Possibly, vitamin E scavenges free radicals generated during ischaemic injury of the subependymal region and thereby limits tissue damage and the extent of periventricular haemorrhage on reperfusion.

In a multi-centre study by seven large renal units in the United Kingdom, the morbidity and mortality of all patients starting CAPD and haemodialysis during a 2-year period (1983-1984) has been monitored prospectively and related to reasons for choice of therapy and potential risk factors. Over this period 338 new patients (mean age 48; range 3-77 years) started CAPD; 17% had diabetes mellitus and 25% had cerebro/cardiovascular disease. One hundred and seventy-five patients (mean age 45; range 5-73 years) started haemodialysis; 6% had diabetes mellitus and 14% had cerebro/cardiovascular disease. The Kaplan-Meier actuarial patient survival estimates at 2 years were haemodialysis 84% and CAPD 83%, whilst technique survival figures for the same period were haemodialysis 92% and CAPD 80%. Cox's regression analysis showed that cerebro/cardiovascular disease and age greater than 60 years were most important predictors for survival in CAPD patients, in whom smoking appeared to be a significant risk factor, for permanent change of therapy to haemodialysis. The major cause of 'drop out' in both groups was transplantation, whilst hospitalisation was 14.9 days per patient year for CAPD and 12.8 for haemodialysis patients. Within the CAPD group a temporary change to haemodialysis (less than 2 months) occurred on 106 occasions (each of mean of 19 days duration), amounting to 10 days per patient year of therapy. CAPD was used twice as often as haemodialysis for managing new patients. After 2 years hospitalisation technique and patient survival were comparable in the two groups, with cerebro/cardiovascular disease, age, and smoking being significant predictors of outcome.

Daubenspeck and Ogden in a recent paper recommended the use of directional statistics in the analysis of response slopes, and their advice has been followed by other workers. Their method is not valid, since it does not follow directly from their model. An efficient estimator of the slope (i.e., an estimator with minimum variance) is well known and is given here with a confidence interval for the true slope. They were also concerned with the two-sample problem to compare the slopes from two different samples. The method for this is more complicated but is summarized here. The likelihood ratio test and point and interval estimates are given. We discuss Daubenspeck and Ogden's example and the reason why, despite its invalidity, their method gave good results using their own data. Their data are also used to illustrate the methods described in this paper, and examples are given to highlight the practical differences between the two methods. Step-by-step procedures are included in the appendixes to enable readers to use these methods.