How to integrate proxy data from two informants in life event assessment in psychological autopsy

BMC Psychiatry

Table 4 Validity of proxy data by six psychological autopsy information reconstruction methods compared with data obtained from targets

Type	Life event	Positive rate (%)		Sensitivity (%)	Specificity (%)	Youden index (%)	Kappa value
Type	Life event	Proxy data	Target	Sensitivity (%)	Specificity (%)	Youden index (%)	Kappa value
1st	Cat1	16.35	20.91	55.17	93.92	49.09	0.53^**
	Cat2	27.64	28.37	62.71	86.24	48.95	0.49^**
	Cat3	12.74	23.80	26.26	91.48	17.74	0.21^**
	Cat4	21.88	27.16	48.67	88.12	36.79	0.39^**
	Cat5	4.57	12.02	14.00	96.72	10.72	0.15^*
2nd	Cat1	14.42	20.91	45.98	93.92	39.90	0.45^**
	Cat2	20.67	28.37	54.24	92.62	46.86	0.51^**
	Cat3	12.02	23.80	30.30	93.69	23.99	0.29^**
	Cat4	16.59	27.16	44.25	93.73	37.98	0.43^**
	Cat5	6.97	12.02	16.00	94.26	10.26	0.13^*
C	Cat1	21.63	20.91	62.07	89.06	51.13	0.51^**
	Cat2	34.86	28.37	72.88	80.20	53.08	0.50^**
	Cat3	21.39	23.80	45.45	86.12	31.57	0.33^**
	Cat4	28.85	27.16	61.95	83.50	45.45	0.45^**
	Cat5	10.34	12.02	22.00	91.26	13.26	0.14^*
D	Cat1	8.65	20.91	39.08	99.39	38.47	0.49^**
	Cat2	11.78	28.37	39.83	99.33	39.16	0.48^**
	Cat3	3.13	23.80	11.11	99.37	10.48	0.15^**
	Cat4	8.41	27.16	27.43	98.68	26.11	0.33^**
	Cat5	0.96	12.02	8.00	100.00	8.00	0.13^**

Note: Type A and B gained the same results with type 1st and 2nd respectively in each category of life event, and their results were not repeatedly demonstrated. ^*P < 0.01, ^**P < 0.001

ISSN: 1471-244X