Nonadherence to treatment protocol in published randomised controlled trials: a review

Table 2 Reporting of key points in 100 trial reports

	Report for all trials	Report according to treatment duration
		Single (one-off) intervention (n = 12)	Short-term intervention (n = 22)	Long-term intervention (n = 66)
1. Randomisation	100 (100%)	-	-	-
2. Adherence to treatment protocol
a. Initiation	-	10 (83%)	15 (68%)	33 (50%)
b. Completion/persistence	-	-	12 (55%) [17 (77%)]^a	31 (47%) [51 (77%)]^a
c. Adherence over implementation period
i) Method	-	-	1 (50%)^b	21 (47%)^b
ii) Justification for definition	-	-	0 (0%)^c	0 (0%)^c
iii) Results	-	-	5 (23%)^d	22 (33%) [28 (42%)]^d
3. Analysis
a. Number analysed	100 (100%)^e	-	-	-
b. Analysis set composition	43 (43%) [91 (91%)]^f	-	-	-

Cells left blank are not required. ^aNumber (%) of trials that fully reported [partially reported] on persistence or completion of randomised treatment. ^bNumber of short-term (n = 2) or long-term (n = 45) intervention trials with patient-administered treatment (2) is used as denominator for %. ^cNumber of short-term (n = 1) or long-term (n = 18) intervention trials with patient-administered treatment that reported adherence definition is used as denominator for %. ^dNumber (%) of trials that fully reported [partially reported] some measure of nonadherence on the part of patient or treatment provider. ^eNumber (%) of trials that reported the number included in analysis of primary outcome. ^fNumber (%) of trials that fully reported [partially reported] analysis set composition; partially reporting trials stated that analysis was by intention to treat (n = 47) or per protocol (n = 1) but did not explicitly explain composition of analysis sets.

ISSN: 1745-6215