Investigating increased admissions to neonatal intensive care in England between 1995 and 2006: data linkage study using Hospital Episode Statistics

Table 4 Number of pairs matched using guestimate probabilities (1995)

Cut off weight	N pairs	N EPICure	% EPICure	N HES	% HES
15	2093	537	80.39	1846	0.38
16	1939	528	79.04	1726	0.35
17	792	365	54.64	692	0.14
18	467	302	45.21	401	0.08
19	435	285	42.66	380	0.08
20	335	256	38.32	294	0.06
21	270	216	32.34	237	0.05
22	229	200	29.94	208	0.04
23	202	182	27.25	193	0.04
24	175	166	24.85	167	0.03
25	158	150	22.46	152	0.03
26	145	138	20.66	142	0.03
27	140	133	19.91	137	0.03
28	112	110	16.47	109	0.02
29	97	96	14.37	96	0.02
30	86	86	12.87	86	0.02
31	67	67	10.03	67	0.01
32	50	50	7.49	50	0.01
34	47	47	7.04	47	0.01
35	41	41	6.14	41	0.01
37	31	31	4.64	31	0.01
38	26	26	3.89	26	0.01
39	9	9	1.35	9	0.00
40	4	4	0.60	4	0.00
42	2	2	0.30	2	0.00
43	0	0	0.00	0	0.00

Table of the number of pairs in 1995 matched from each data set for differing cutoffs in the value of the weight calculated by the Fellegi-Sunter (guestimate) method of data linkage

ISSN: 1471-2288