Comparison of randomization techniques for clinical trials with data from the HOMERUS‐trial

Figures & data

Figure 1 Flow diagram of the run‐in phase in HOMERUS.

Table I. Baseline characteristics of the randomized patients for the total group, the office pressure (OP) group and the self‐pressure (SP) group.

	Total, 459	OP, 230	SP, 229
Males, n (%)	249 (54)	126 (55)	123 (54)
Age, years	55±11	55±11	55±11
BMI, kg/m^2	28±5	28±4	28±5
Obesity, BMI>30 kg/m^2, n (%)	112 (24)	55 (24)	57 (24)
Waist‐to‐hip ratio	0.9±0.1	0.9±0.1	0.9±0.1
Smoking, n (%)	82 (18)	40 (17)	42 (18)
Drugs,^an (%)	321 (70)	158 (69)	163 (71)
Beta‐blockers, n (%)	133 (29)	72 (31)	61 (27)
Diuretics, n (%)	131 (29)	66 (29)	65 (28)
ACE, n (%)	112 (24)	55 (24)	57 (25)
ARB, n (%)	68 (15)	32 (14)	36 (16)
Calcium‐channel blockers, n (%)	56 (12)	27 (12)	29 (13)
Alpha‐blockers, n (%)	8 (2)	2 (1)	6 (3)
Lipid‐lowering drugs, n (%)	62 (14)	33 (14)	29 (13)
TAI, n (%)	37 (8)	19 (8)	18 (8)

n, number of patients; BMI, body mass index; ACE, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blockers; TAI, thrombocyte aggregation inhibitors.

Data are depicted as number of patients and percentage or as mean±standard deviation.

^aNumber of patients that used antihypertensive drugs at study enrolment.

Table II. Laboratory results at baseline for the total group, the office pressure (OP) group and the self‐pressure (SP) group.

	Total, 430	OP, 214	SP, 216
Glucose, mmol/l (SD)	5.5 (1)	5.4 (1)	5.6 (1)
Diabetes mellitus, n (%)	25 (6)	17 (8)	8 (4)
Cholesterol, mmol/l (SD)	5.7 (1)	5.7 (1)	5.7 (1)
Hypercholesterolaemia, n (%)	80 (19)	45 (21)	35 (16)
Creatinine, μmol/l (SD)	82 (16)	82 (14)	82 (17)
Creatinine level elevated, n (%)	12 (3)	6 (3)	6 (3)
Microalbuminuria, n (%)	58 (13)	30 (14)	28 (13)
Proteinuria, n (%)	4 (1)	2 (1)	2 (1)
Left ventricular hypertrophy, n (%)	61 (14)	32 (15)	29 (13)

n, number of patients. Data are depicted as number of patients and percentage or as mean±standard deviation (SD).

Table IIIa. Result of random allocation using the minimization method.

Treatment group	Region				Location		Age (years)			Gender		Clinical symptoms		Run‐in
	1	2	3	4	Clinic	GP	<40	40–60	>60	M	F	Y	N	Y	N
OP (n)	101	38	46	45	157	73	23	129	78	126	104	9	221	221	9
SP (n)	99	47	43	40	156	73	22	128	79	123	106	8	221	219	10
	200	85	89	85	313	146	45	257	157	249	210	17	442	440	19

Table IIIb. Result of random allocation using the randomization with four permuted blocks and without strata.

Treatment group	Region				Location		Age (years)			Gender		Clinical symptoms		Run‐in
	1	2	3	4	Clinic	GP	<40	40–60	>60	M	F	Y	N	Y	N
OP (n)	93	46	51	40	164	66	28	128	74	125	105	9	221	221	9
SP (n)	107	39	38	45	149	80	17	129	83	124	105	8	221	219	10
	200	85	89	85	313	146	45	257	157	249	210	17	442	440	19

Table IIIc. Result of random allocation using the randomization with four permuted blocks and with 16 strata.

Treatment group	Region				Location		Age (years)			Gender		Clinical symptoms		Run‐in
	1	2	3	4	Clinic	GP	<40	40–60	>60	M	F	Y	N	Y	N
OP (n)	91	44	48	46	157	72	19	137	73	126	103	7	222	220	9
SP (n)	109	41	41	39	156	74	26	120	84	123	107	10	220	220	10
	200	85	89	85	313	146	45	257	157	249	210	17	442	440	19

Random allocation for six outcome factors accomplished by three different methods of randomization. A number of 459 patients were separated into two treatment groups (SP and OP) with maintenance of distribution balance for the six outcome factors. 1, Maastricht; 2, Nijmegen; 3, Amsterdam; 4, Groningen; GP, general practitioner; M, male; F, female; Y, yes; N, no.

Figure 2 Twenty‐four‐hour ambulatory blood pressure measurement (ABPM) illustrated for each outcome factor that was used for the minimization method. ***p<0.001; **p<0.03. Data are depicted as mean±standard deviation.

Figure 3 Twenty‐four‐hour ambulatory blood pressure measurements for the office pressure (OP) and self‐pressure (SP) group as randomized with different randomization procedures. *p<0.05. Data are depicted as mean±standard deviation.

Figure 4 Twenty‐four‐hour ambulatory blood pressure measurements compared for patient characteristics that could influence prognosis. No LVH, patients without left ventricular hypertrophy; LVH, patients with left ventricular hypertrophy; Malb, microalbuminuria; Hchol, hypercholesterolaemia; DM, diabetes mellitus; Hcreat, elevated creatinine level. ***p<0.001;**p<0.03. Data are depicted as mean±standard deviation.

Table IV. Office blood pressure measurement (OBPM), ambulatory blood pressure measurement (ABPM) and self‐blood pressure measurement (SBPM) at enrolment.

	Total				OP group				SP group
	Subjects	Systolic (mmHg)	Diastolic (mmHg)	HR (beats/min)	Subjects	Systolic (mmHg)	Diastolic (mmHg)	HR (beats/min)	Subjects	Systolic (mmHg)	Diastolic (mmHg)	HR (beats/min)
OBPM
1st visit	459	161±19	95±10	74±12	230	161±21	95±10	73±12	229	161±18	96±10	74±12
2nd visit	358	164±19	97±10	76±13	171	165±20	97±10	75±13	187	163±18	97±9	77±12
3rd visit	430	166±20	97±10	77±13	214	165±21	98±11	76±13	216	166±19	97±10	77±12
ABPM
Mean	412	143±14	88±9	74±9	205	142±14	87±9	74±10	207	143±14	88±10	73±9
Day	412	148±15	93±10	78±11	205	148±15	93±10	79±12	207	149±15	92±11	78±10
Night	406	127±15	76±10	65±9	203	126±16	75±11	66±9	203	128±15	76±10	65±8
SBPM		Not performed				Not performed			205	155±15	92±9	72±9

HR, heart rate.

Data are depicted as mean±standard deviation.