Table 1.

Quality Assessment

Study, Author, Year	Adequate Randomization	Concealed Allocation	Sufficient Sample Size	Similar Groups	Blinded	Validated Measures	Adequate Follow-Up	Intention-to-Treat Analysis	Insignificant COI	Overall Potential Risk of Biasa
ECOG, Tormey, 1996	+b	+	?	+	−	?	+	?	+	Intermediate
NSABP B-14, Fisher, 2001	Partialc	?	?	+d	+	+	+e	−f	+	Intermediate
Scottish Cancer Trials Breast Group, Stewart, 2001	+	?	?g	+ (at second randomization)	-	+h	+	+	+	Intermediate
ATLAS, Davies, 2012	+	+	+	+	−	+	+	+	+	Low
aTTom, Gray, 2013		NR	NR	NR	NR	NR	NR	NR	NR	Insufficient data available for assessment

NOTE. + indicates criterion was met; − indicates criterion was not met; ? indicates insufficient detail, not reported, and/or uncertain risk of bias.

Abbreviations: COI, conflict(s) of interest; NR, not reported.

^aOverall risk is an estimation based on the collective criteria and the extent of potential bias.

^bCalls to central office.

^cBiased-coin randomization.

^dWomen who participated in the second randomization, or who were registered, had more tumors ≤ 2.0 cm in size than did women in the first randomization. “Reported age is that at the first randomization; at the second randomization, the women were older and more likely to be postmenopausal.”

^eTrial terminated after three of four interim analyses.

^fAnalyzed data from 98% of participants.

^gMethod not given in the two papers that are currently available.

^hEleven patients with protocol violations excluded from 1987 analysis; 10 of 11 included in 2001 analysis.