Protocol for a randomized controlled trial of early prophylactic feeding via gastrostomy versus standard care in high risk patients with head and neck cancer

Table 1 Examples of sample sizes required with varying assumptions and a Type I Error of 0.05

	Number of patients required in each arm
Power	SD = 9.4[25]		SD = 5.34[35]
	5% difference*	7% difference**	5% difference*	7% difference**
80%	56	29	19	10
90%	75	39	25	13

*Difference to reach clinically significant target weight loss of <5% (as studies have shown weight loss of 9.3-9.75%) [25, 35].
**Difference seen with pre-treatment dietary counselling intervention [35].

ISSN: 1472-6955