Naslov
Textual Analysis of EuroRec Repository Statements: Comparison of FGS and GPR Sets

Autori
Hercigonja-Szekeres, Mira ; Ilakovac, Vesna ; Devlies, Jos ; De Moor, Georges

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, neobjavljeni rad, stručni

Skup
Seamless Care – Safe Care The Challenges of Interoperability and Patient Safety in Health Care ; EFMI STC 2010

Mjesto i datum
Reykjavík, Island, 02.06.2010. - 04.06.2010

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
textual analysis ; consistency of translations ; multilingual dictionary ; FGS statements ; GPR statements ; EuroRec

Sažetak
European Institute for Health Record (EuroRec) composed a generic and comprehensive repository of statements describing use functions, structuring and data elements of EHR systems. Two sets of statements are especially pointed out: Fine Grained Statements (FGS) derived from source statements which were originally extracted from original EHR system specifications and test plans, and Good Practice Requirements (GPR) statements recomposed from FGS into the more common useful building blocks representing homogenous sets of FGSs. The aim of this study was to compare vocabulary of FGS and GPR sets and to assess the possibility of creating unique coherent multilingual dictionary for FGS and GPR by identifying most frequent words and segments within repository statements using statistical textual analysis. We performed lexicometric analysis, analysis of repeated segments and text concordance analysis. The Dtm-Vic software was used for the analysis. The FGS set comprised 1688 statements with 21470 words in total. There were 179 statements in GPR set with 5079 words in total. The GPR set had richer vocabulary then FGS set (difference in proportions of distinct words 10.8%, 95% CI from 9.7% to 12.0%). A number of most frequent repeated segments in the FGS set were not found or had very low frequency in the GPR set. Although the GPR set is derived from FGS, multiple words segments in these sets were composed differently. Thus both sets should be first investigated separately, and after this procedure it would be possible to use common expressions (words, repeated segments, longer text units) and translate them equally in both sets. Statistical textual analysis might be a useful tool to bridge a gap in the multilingual environment of the process of unified EHR system quality assessment.

Izvorni jezik
Engleski

Znanstvena područja
Javno zdravstvo i zdravstvena zaštita, Informacijske i komunikacijske znanosti

POVEZANOST RADA