Table 2.

Experiential training of the research assistants with the Confusion Assessment Method

Site	RA	CAM experiential training procedures
		One-to-one sessions (N)	Pilot interviews with patients (N)			Inter-rater reliability assessments (N)
			N	No delirium	Delirium	N	No delirium	Delirium
1	1	6	4	1	3	10	5	5
	2	6	4	1	3	10	5	5
2	3	0	4	4	0	6	6	0
	4	0	–	–	–	7	7	0
	5	0	4	3	1	5	5	0
	6	0	0	–	–	9	9	0
3	7	2	4	4	0	19	18	1
	8	4	4	4	0	20	19	1
	9	4	4	4	0	9	9	0
	10	4	4	4	0	12	12	0
	11	3	4	4	0	10	10	0
4	12	0	3	3	0	13	12	1
	13	3	4	3	1	10	9	1
	14	0	2	2	0	10	10	0
	15	–	–	–	–	–	–	–
	16	1	2	2	0	5	5	0
	17	6	3	3	0	6	6	0
	18	1	4	4	0	2	2	0
5	19	–	–	–	–	–	–	–
	20	2	3	3	0	6	6	0
	21	2	3	2	1	6	6	0
	22	–	–	–	–	–	–	–
	23	–	–	–	–	–	–	–
6	24	2	0	–	–	25	22	3
	25	2	0	–	–	13	11	2
	26	1	0	–	–	5	5	0
	27	2	0	–	–	4	4	0
	28	1	0	–	–	4	4	0
	29	1	0	–	–	6	5	1
	30	1	0	–	–	3	3	0
	31	1	0	–	–	12	10	2
7	32	2	1	1	0	7	4	3
	33	2	1	1	0	7	4	3
	34	2	1	1	0	7	5	2
	35	2	1	1	0	7	5	2
8	36	3	4	4	0	10	10	0
	37	3	4	4	0	10	10	0
Total		69	72	63	9	295a	263	32
Mean (SD)		2.1 (1.72)	2.3 (1.74)			8.9 (4.94)
Median (IQR)		2 (1–3)	3 (0–4)			7 (6–10)
Range		0–6	0–4			2–25

RA research assistant, CAM confusion assessment method, SD standard deviation, IQR interquartile range

^a139 patient assessments were 1:1 (interviewer:observer); 3 were 1:2; 2 were 1:3