id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-303845-y6ws3u6x	DeLisle, Sylvain	Combining Free Text and Structured Electronic Medical Record Entries to Detect Acute Respiratory Infections	2010-10-14		.txt	text/plain	5608	246	44	Several factors, some of which have already been mentioned, may limit the generalizability of our results: 1) factors related to the performance of our study at the VA health care system: a) the veterans study population is mostly male and excludes the pediatric population, a key target for ARI surveillance [44] ; b) veterans health care utilization may differ from that observed in uninsured or privately insured individuals; c) clinical practices, documentation and coding habits by VA practitioners may differ from those observed in solo or group practices or in health systems subject to different financial or quality-control incentives; 2) factors related to our study period: optimal CDAs could differ outside the respiratory infection season, or during periods of heightened apprehension for an influenza epidemic; 3) factors related to our iterative CDA development process, which may have over adapted CDAs to VA's particular EMR implementation and to our sample dataset in particular, this despite our efforts to maintain a separation between development and validation data subsets; 4) factors related to our text mining approach: a) we did not employ a spell checker prior to applying the NegEx algorithm.	./cache/cord-303845-y6ws3u6x.txt	./txt/cord-303845-y6ws3u6x.txt