The DUNDRUM-1 structured professional judgment for triage to appropriate levels of therapeutic security: retrospective-cohort validation study

Table 5 Item to outcome agreement where outcome is matched to appropriate level or higher.

	All assessed over a three month period April to June 2009 (n = 246)		All diverted from prison to hospital over a two year period 2008-2009 (n = 100)
	Observed proportion in Agreement^a	99% CI of observed proportion in agreement	Observed proportion in Agreement^a	99% CI of observed proportion in agreement
TS1: serious violence	.85	.79 - .91	.88	.79 - .97
TS2: serious self harm	.91	.86 - .96	.98	.94 - .99
TS3: immediacy of violence risk	.89	.84 - .94	.72	.60 - .84
TS4: immediacy of self harm risk	.97	.94 - .99	.99	.96 - .99
TS5: specialist forensic need	.96	.92 - .99	.95	.89 - .99
TS6: absconding risk	.89	.85 - .95	.92	.84 - .99
TS7: preventing access	.92	.87 - .96	.90	.82 - .98
TS8: victim sensitivities	.93	.89 - .97	.81	.71 - .91
TS9: complex risks	.81	.75 - .87	.79	.68 - .89
TS10: institutional behaviour	.91	.86 - .96	.99	.96 - .99
TS11: legal procedure	.95	.92 - .99	.90	.82 - .98

First sample all assessed after screening over a three month period, n = 246; second sample all diverted to hospitals over a two year period n = 100. ^aBinomial test for observed proportions differing from test proportion 0.5, based on Z approximation all p < 0.001 (2-tailed).

ISSN: 1471-244X