نشریه مدیریت اطلاعات سلامت
سال دهم شماره 1 (پیاپی 29، فروردین و اردیبهشت 1392)

Reliability of diagnoses coding is essential for the use of data in national and international levels. The present study compared the reliability of diagnoses coding with 10th version of the International Classification of Diseases (ICD-10) between the two groups of coders.

Two hundred and forty five priorly coded medical records from five public hospitals affiliated to Zahedan University of Medical Sciences were re-coded by hospital coders and researcher in the first half of 2011. Data collection was done through a checklist. Validity of this instrument was confirmed by the experts. Intra-rater reliability (reliability of coder) and inter-rater reliability (reliability of prior codes with researcher codes or current codes of coder with researcher codes) were assessed in different level of diagnoses using Cohen’s Kappa.

In the most cases، inter-rate reliability was almost perfect. Only in level of first three character of principle diagnoses value of Kappa was moderate (K = 0. 52). Furthermore، external reliability in most cases was upper than moderate. Only reliability of codes dedicated by researcher and prior codes on medical records in the fifth character of principle diagnosis was low (K = 0. 18)، the fourth character of the first other diagnosis was moderate (K= 0. 60)، and also in chapter level of the third، the other diagnosis were moderate (K= 0. 54). Furthermore، Kappa value (between coder and researcher) for principle diagnosis in the first three character was moderate (K = 0. 47)، and in the fifth character level was low (K = 0. 18).

Reliability of diagnoses coding was appropriate in level of chapter and the fourth character and was not appropriate in three levels of the first and fourth characters. This could be resulted from error selection of principle diagnosis in multiple-coding and negligence of coder to level of specificity in coding. Thus، implementation of in-service educational programs for coders seems essential.

Library software is a computer-based program that is used in libraries and information centers for information storage، processing and retrieval. Interface environment is a part of library software that establishes connection between computer and user and also shifts information between user and the system. The present study aimed to determine the attitudes of librarians in Isfahan University of Medical Sciences towards the specifications of Pars Azarakhsh library software interface in 2009.

A questionnaire was used for data collection، consisting of 51 closed questions. The study population included the librarians in Isfahan University of Medical Sciences. Librarians were all those who were working in libraries of Isfahan University of Medical Sciences (n = 50). Descriptive statistics methods and correlation test were used for data analysis.

The results of this research showed that librarians believed some factors like pageant، vocabulary and messages of software as well as learning how to use software were in medium and upper than medium. Moreover، librarian''s attitude toward help and other capabilities was lower than medium. All the librarians believed that Pars Azarakhsh software user interface environment was medium (3. 20). In addition، there was a non-significant difference between Pars Azarakhsh software user interface and computer literacy of librarians. There was a significant difference between letters and messages of software and library literacy of librarians and there was no significant difference between other variables of Pars Azarakhsh software user interface and library literacy of librarians.

Findings showed that librarians'' view about Pars Azarakhsh variables was at medium level. Among these variables، some such as how to work with software، pageant، vocabulary، and messages of software were in the best situation. However، help features and other capabilities need to be developed more.