Recommendations for the analysis of individually randomised controlled trials with clustering in one arm – a case of continuous outcomes

Table 5 Summary of results for the Acupuncture case study (n=215)

Model	Treatment estimate	Standard error	Residual variance	Random variance	Control variance	ICC
Ignore clustering	5.69	3.61	604.67
Random effects
Individual clusters of size 1	5.56	3.75	598.68	6.68		0.0110^a
One large cluster	5.49	5.02	598.59	10.61		0.0174^a
Artificial-clusters	5.59	3.95	599.29	6.63		0.0109^a
Partially clustered
Individual clusters of size 1	5.49	3.82	598.59	10.61		0.0174^b
One large cluster	5.49	3.82	598.59	10.61		0.0174^b
Artificial-clusters	5.49	3.82	598.59	10.61		0.0174^b
Heteroskedastic
Individual clusters of size 1	5.54	3.61	649.46	8.56	491.52	0.00744^b
One large cluster	5.54	3.61	649.46	8.56	491.52	0.00744^b
Artificial-clusters	5.54	3.61	649.46	8.56	491.52	0.00744^b

ISSN: 1471-2288