T-Space at The University of Toronto Libraries >
Journal of Medical Internet Research >
Volume 9 (2007) >
Please use this identifier to cite or link to this item:
|Title: ||A Survey of Quality Assurance Practices in Biomedical Open Source Software Projects|
|Authors: ||Koru, Günes|
El Emam, Khaled
|Keywords: ||Original Paper|
Open source software
software quality assurance
code inspections and walkthroughs
|Issue Date: ||7-May-2007|
|Publisher: ||Gunther Eysenbach; Centre for Global eHealth Innovation, Toronto, Canada|
|Citation: ||Günes Koru, Khaled El Emam, Angelica Neisa, Medha Umarji. A Survey of Quality Assurance Practices in Biomedical Open Source Software Projects. J Med Internet Res 2007;9(2):e8 <URL: http://www.jmir.org/2007/2/e8/>|
|Abstract: ||[This item is a preserved copy and is not necessarily the most recent version. To view the current item, visit http://www.jmir.org/2007/2/e8/ ]
Open source (OS) software is continuously gaining recognition and use in the biomedical domain, for example, in health informatics and bioinformatics.
Given the mission critical nature of applications in this domain and their potential impact on patient safety, it is important to understand to what degree and how effectively biomedical OS developers perform standard quality assurance (QA) activities such as peer reviews and testing. This would allow the users of biomedical OS software to better understand the quality risks, if any, and the developers to identify process improvement opportunities to produce higher quality software.
A survey of developers working on biomedical OS projects was conducted to examine the QA activities that are performed. We took a descriptive approach to summarize the implementation of QA activities and then examined some of the factors that may be related to the implementation of such practices.
Our descriptive results show that 63% (95% CI, 54-72) of projects did not include peer reviews in their development process, while 82% (95% CI, 75-89) did include testing. Approximately 74% (95% CI, 67-81) of developers did not have a background in computing, 80% (95% CI, 74-87) were paid for their contributions to the project, and 52% (95% CI, 43-60) had PhDs. A multivariate logistic regression model to predict the implementation of peer reviews was not significant (likelihood ratio test = 16.86, 9 df, P = .051) and neither was a model to predict the implementation of testing (likelihood ratio test = 3.34, 9 df, P = .95).
Less attention is paid to peer review than testing. However, the former is a complementary, and necessary, QA practice rather than an alternative. Therefore, one can argue that there are quality risks, at least at this point in time, in transitioning biomedical OS software into any critical settings that may have operational, financial, or safety implications. Developers of biomedical OS applications should invest more effort in implementing systemic peer review practices throughout the development and maintenance processes.|
|Description: ||Reviewer: Murray, Peter|
Reviewer: Pietrobon, R
Reviewer: McDonald, C
|Rights: ||© Günes Koru, Khaled El Emam, Angelica Neisa, Medha Umarji. Originally published in the Journal of Medical Internet Research (http://www.jmir.org, 07.05.2007). Except where otherwise noted, articles published in the Journal of Medical Internet Research are distributed under the terms of the Creative Commons Attribution License (http://www.creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited, including full bibliographic details and the URL (see "please cite as" above), and this statement is included.|
|Appears in Collections:||Volume 9 (2007)|
Items in T-Space are protected by copyright, with all rights reserved, unless otherwise indicated.