Table 3.

Summary of the judgements made by the 17 evaluators for Example 1 of the pilot study. Should depressed patients in primary care be treated with SSRIs rather than tricyclics?

Rater	Quality of outcome			Relative importance of outcome			Overall quality	Balance benefits vs harm	Recommendation
	Depression severity	Transient side effects	Poisining fatalities	Depression severity	Transient side effects	Poisining fatalities
1	H	H	Vl	7	9	9	H or Vl	Uncertain net benefit	Don't do it
2	M	M	Vl	9	6	7	Vl	Net benefit	Probably do it
3	H	H	M	8	7	9	H	Uncertain net benefit	Toss up
4	H	H	L	6	5	6
5	M	M	M	9	6	8	M	Net benefit	Do it
6	M	M	Vl	9	6	9	Vl	Net benefit	Do it
7	M	M	L	8	7	8	L	Net benefit	Do it
8	H	H	Vl	9	5	3	H	Net benefit	Probably do it
9	M	M	L	9	6	8	L	Net benefit	Probably do it
10	M	M	L	9	7	8	L	Net benefit	Probably do it
11	H	H	L	8	5	7	L	Trade offs	Probably do it
12	H	H	L	8	5	7	L	Trade offs	Probably do it
13	M	H	M	9	7	9	M	Net benefit	Do it
14	M	M	L	9	9	5 OR 9	M	Net benefit	Probably do it
15	M	M	Vl	9	6	8	Vl	Uncertain net benefit	Toss up
16	M	M	Vl	9	5	9	Vl	Not net benefit	Don't do it
17	M	M	M	9	9	9	M	Net benefit	Toss up