精神病学中的决策支持——使用计算机化决策支持系统与人工诊断的诊断结果比较。

Decision support in psychiatry - a comparison between the diagnostic outcomes using a computerized decision support system versus manual diagnosis.

作者信息

Bergman Lars G, Fors Uno G H

机构信息

Dept, of Learning, Informatics, Management and Ethics (LIME), Karolinska Institutet, Berzelius way 3, S- 171 77 Stockholm, Sweden.

出版信息

BMC Med Inform Decis Mak. 2008 Feb 8;8:9. doi: 10.1186/1472-6947-8-9.

DOI:10.1186/1472-6947-8-9

PMID:18261222

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC2258291/

Abstract

BACKGROUND

Correct diagnosis in psychiatry may be improved by novel diagnostic procedures. Computerized Decision Support Systems (CDSS) are suggested to be able to improve diagnostic procedures, but some studies indicate possible problems. Therefore, it could be important to investigate CDSS systems with regard to their feasibility to improve diagnostic procedures as well as to save time.

METHODS

This study was undertaken to compare the traditional 'paper and pencil' diagnostic method SCID1 with the computer-aided diagnostic system CB-SCID1 to ascertain processing time and accuracy of diagnoses suggested. 63 clinicians volunteered to participate in the study and to solve two paper-based cases using either a CDSS or manually.

RESULTS

No major difference between paper and pencil and computer-supported diagnosis was found. Where a difference was found it was in favour of paper and pencil. For example, a significantly shorter time was found for paper and pencil for the difficult case, as compared to computer support. A significantly higher number of correct diagnoses were found in the diffilt case for the diagnosis 'Depression' using the paper and pencil method. Although a majority of the clinicians found the computer method supportive and easy to use, it took a longer time and yielded fewer correct diagnoses than with paper and pencil.

CONCLUSION

This study could not detect any major difference in diagnostic outcome between traditional paper and pencil methods and computer support for psychiatric diagnosis. Where there were significant differences, traditional paper and pencil methods were better than the tested CDSS and thus we conclude that CDSS for diagnostic procedures may interfere with diagnosis accuracy. A limitation was that most clinicians had not previously used the CDSS system under study. The results of this study, however, confirm that CDSS development for diagnostic purposes in psychiatry has much to deal with before it can be used for routine clinical purposes.

摘要

背景

新型诊断程序可能会改善精神病学的正确诊断。计算机化决策支持系统（CDSS）被认为能够改善诊断程序，但一些研究指出了可能存在的问题。因此，研究CDSS系统在改善诊断程序以及节省时间方面的可行性可能很重要。

方法

本研究旨在将传统的“纸笔”诊断方法SCID1与计算机辅助诊断系统CB-SCID1进行比较，以确定诊断建议的处理时间和准确性。63名临床医生自愿参与该研究，并使用CDSS或手动方式解决两个纸质病例。

结果

未发现纸笔诊断和计算机辅助诊断之间存在重大差异。若发现有差异，则有利于纸笔诊断。例如，与计算机辅助相比，纸笔诊断在处理困难病例时所用时间明显更短。使用纸笔方法诊断“抑郁症”时，在困难病例中发现正确诊断的数量明显更多。尽管大多数临床医生认为计算机方法有帮助且易于使用，但与纸笔诊断相比，它花费的时间更长且正确诊断更少。

结论

本研究未发现传统纸笔方法与计算机辅助精神病诊断在诊断结果上有任何重大差异。若存在显著差异，传统纸笔方法优于测试的CDSS，因此我们得出结论，用于诊断程序的CDSS可能会干扰诊断准确性。一个局限性是大多数临床医生此前未使用过所研究的CDSS系统。然而，本研究结果证实，用于精神病学诊断目的的CDSS在可用于常规临床目的之前还有很多工作要做。