Table 2.

Words associated with a pancreatitis health state are more frequently found in the short gap bins, suggesting that the separation is grouping notes written during visits that are more concentrated on the pancreatitis diagnosis into the 0–3 day measurement gap. The raw, normalized, and coverage frequencies of words in each gap sorted by the difference in coverage between the two bins is shown in this table. The words with a larger than 1% difference are shown, each of the words shown is highly associated with the 0–3 gap.

Frequency of Words in Each Bin
	0–3 day measurement gap			3+ day measurement gap
Words	Raw Frequency	% Total Words	% Notes containing that Word	Raw Frequency	% Total Words	% Notes containing that Word	Difference in Coverage
pancreatitis	10732	0.14	6.94	13303	0.01	1.41	5.52
lipase	4908	.06	4.28	9728	0.01	1.50	2.79
amylase	3855	.05	3.48	98246	0.01	1.30	2.19
withdrawal	4139	.05	2.80	12562	0.01	1.28	1.52
librium	3303	.04	2.00	6064	0.00	0.59	1.42
pancreatic	4393	.06	2.76	15992	0.01	1.38	1.38
epigastric	3668	.05	2.92	15767	0.01	1.89	1.04