Clinical and Experimental Pharmacology and Physiology (1 992) 19,294-296
SHORT COMMUNICATION SAMPLE SIZE REQUIREMENTS FOR CLINICAL TRIALS OF ISOLATED SYSTOLIC HYPERTENSION C. A. Silagy, J. J. MeNeil and B. P. McGrath* Department of Social & Preventive Medicine and *Department of Medicine, Monash Medical Centre, Monash University, Melbourne, Victoria, Australia (Received 13 December 1991)
SUMMARY 1. This study investigated components of blood pressure variability in elderly subjects with isolated systolic hypertension (ISH) using both ambulatory blood pressure monitoring (ABPM) and casual clinic blood pressure measurement. These were then used to determine sample size requirements for clinical trials of different designs. 2. Eleven elderly subjects not receiving antihypertensive medication were seen on four occasions at weekly intervals. On each occasion blood pressure was measured in the clinic and then for 24 h using a non-invasive ABPM device. Nested analysis of variance was used to calculate the 'between subject' and 'between subject within occasion' components of blood pressure variability. 3. Increasing the number of readings or occasions where measurement was performed in a parallel group trial only reduced the variability substantially when the number of subjects involved was less than 50. Use of a cross-over design substantially reduced the sample size required. 4. ABPM appears most useful as a strategy for reducing sample size in parallel group trials in ISH involving small numbers of subjects measured on one occasion.
Key words: ambulatory blood pressure monitoring, isolated systolic hypertension, clinical trial design.
INTRODUCTION Blood pressure (BP) varies substantially with repeated measurement in an individual because of factors such as imprecise measurement, diurnal fluctuation, differing levels of physical activity and sympathetic stimulation. The variation in a group of individuals is even greater because differences between individuals are added to those occurring within individuals. In clinical trials of new antihypertensive therapies it is necessary to use statistical tests, such as analysis of
variance (ANOVA), to separate the treatment effect from other sources of variation referred to above. A practical method now exists for obtaining multiple measurements of BP using 24 h non-invasive ambulatory monitoring (ABPM). Intuitively it might be expected that ABPM would reduce the standard error of mean individual BP and therefore the number of subjects required to identify a given treatment effect. However some of this advantage may be lost by the
Correspondence: Dr C. A. Silagy, Department of Social & Preventive Medicine, Alfred Hospital, Commercial Rd, Prahran, Vic. 3181, Australia. Presented at the High Blood Pressure Research Council of Australia meeting on 12-13 December 1991, Adelaide, Australia.
Sample size requirementsfor ISH lack of standardization of the conditions under which the readings are made. This might particularly affect BP levels in the elderly, in whom readings are known to be more variable than in younger age groups (Rowlands et al. 1984). The aim of the this study was to determine whether the use of ABPM would reduce BP variability, and hence sample size requirements, in clinical trials involving elderly subjects with isolated systolic hypertension (ISH).
METHODS Subjects Eleven subjects (45% males) with a mean of 75 years (range 65-84 years) who had ISH (defined as a systolic BPL160 mmHg and diastolic BP
SHORT COMMUNICATION SAMPLE SIZE REQUIREMENTS FOR CLINICAL TRIALS OF ISOLATED SYSTOLIC HYPERTENSION C. A. Silagy, J. J. MeNeil and B. P. McGrath* Department of Social & Preventive Medicine and *Department of Medicine, Monash Medical Centre, Monash University, Melbourne, Victoria, Australia (Received 13 December 1991)
SUMMARY 1. This study investigated components of blood pressure variability in elderly subjects with isolated systolic hypertension (ISH) using both ambulatory blood pressure monitoring (ABPM) and casual clinic blood pressure measurement. These were then used to determine sample size requirements for clinical trials of different designs. 2. Eleven elderly subjects not receiving antihypertensive medication were seen on four occasions at weekly intervals. On each occasion blood pressure was measured in the clinic and then for 24 h using a non-invasive ABPM device. Nested analysis of variance was used to calculate the 'between subject' and 'between subject within occasion' components of blood pressure variability. 3. Increasing the number of readings or occasions where measurement was performed in a parallel group trial only reduced the variability substantially when the number of subjects involved was less than 50. Use of a cross-over design substantially reduced the sample size required. 4. ABPM appears most useful as a strategy for reducing sample size in parallel group trials in ISH involving small numbers of subjects measured on one occasion.
Key words: ambulatory blood pressure monitoring, isolated systolic hypertension, clinical trial design.
INTRODUCTION Blood pressure (BP) varies substantially with repeated measurement in an individual because of factors such as imprecise measurement, diurnal fluctuation, differing levels of physical activity and sympathetic stimulation. The variation in a group of individuals is even greater because differences between individuals are added to those occurring within individuals. In clinical trials of new antihypertensive therapies it is necessary to use statistical tests, such as analysis of
variance (ANOVA), to separate the treatment effect from other sources of variation referred to above. A practical method now exists for obtaining multiple measurements of BP using 24 h non-invasive ambulatory monitoring (ABPM). Intuitively it might be expected that ABPM would reduce the standard error of mean individual BP and therefore the number of subjects required to identify a given treatment effect. However some of this advantage may be lost by the
Correspondence: Dr C. A. Silagy, Department of Social & Preventive Medicine, Alfred Hospital, Commercial Rd, Prahran, Vic. 3181, Australia. Presented at the High Blood Pressure Research Council of Australia meeting on 12-13 December 1991, Adelaide, Australia.
Sample size requirementsfor ISH lack of standardization of the conditions under which the readings are made. This might particularly affect BP levels in the elderly, in whom readings are known to be more variable than in younger age groups (Rowlands et al. 1984). The aim of the this study was to determine whether the use of ABPM would reduce BP variability, and hence sample size requirements, in clinical trials involving elderly subjects with isolated systolic hypertension (ISH).
METHODS Subjects Eleven subjects (45% males) with a mean of 75 years (range 65-84 years) who had ISH (defined as a systolic BPL160 mmHg and diastolic BP
