The performance of an online osteoporosis detection system a sensitivity and specificity analysis

Shu Fang Chang, Chin-Ming Hong, Rong Sen Yang

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Aims and objectives: To develop an online system for the detection of osteoporosis risk and to test its accuracy. Background: Osteoporosis is a silent killer; usually, there are no symptoms, such as pain, until bone erosion and fracture occur. The risks of osteoporosis have been underestimated and neglected; as a result, osteoporosis can be as dangerous as heart diseases and cancers that lead to a healthcare crisis. Design: Cross-sectional study. Methods: The study participants were individuals presenting for routine health examinations at a medical centre in Taiwan from 2006-2007. Women over 30 years of age who underwent dual-energy X-ray absorptiometry scanning for measurement of bone mineral density were eligible for this study. The system for osteoporosis detection and health risk, which was developed in this study, was analysed. Results: The findings indicated a high sensitivity of 75%, specificity of 75%, positive predictive value of 75% and negative predictive value of 75%. In addition, the online osteoporosis detective system had a higher predictive power (24·2% vs. 11%) and a similar cut-off point (33% vs. 27%) compared with the tool designed by the International Osteoporosis Foundation. Conclusion: The online system for detection of osteoporosis risk had excellent reliability and validity. It performed well in predicting osteoporosis and the cut-off point used for identifying the risk among women at risk of developing osteoporosis. Therefore, it is suitable for the Asian women and can help women achieve the goals of early detection and health promotion. Relevance to clinical practice: Early detection is the only way to prevent osteoporosis. Professional nurses should apply effective technology to promote health care in community-dwelling people.

Original languageEnglish
Pages (from-to)1803-1809
Number of pages7
JournalJournal of Clinical Nursing
Volume23
Issue number13-14
DOIs
Publication statusPublished - 2014 Jan 1

    Fingerprint

Keywords

  • Artificial intelligence
  • Health risk
  • Nursing
  • Osteoporosis
  • Sensitivity
  • Specificity

ASJC Scopus subject areas

  • Nursing(all)

Cite this