Rasch Analysis of the 9-Item Shared Decision Making Questionnaire in Women With Breast Cancer

Cancer Nurs. 2019 May/Jun;42(3):E34-E42. doi: 10.1097/NCC.0000000000000607.

Abstract

Background: Shared decision making (SDM) is a best practice to help patients make optimal decisions by a process of healthcare, especially for women diagnosed with breast cancer and having heavy burden in long-term treatments. To promote successful SDM, it is crucial to assess the level of perceived involvement in SDM in women with breast cancer.

Objective: The aims of this study were to apply Rasch analysis to examine the construct validity and person reliability of the 9-item Shared Decision Making Questionnaire (SDM-Q-9) in women with breast cancer.

Methods: The construct validity of SDM-Q-9 was confirmed when the items fit the Rasch model's assumptions of unidimensionality: (1) infit and outfit mean square ranged from 0.6 to 1.4; (2) the unexplained variance of the first dimension of the principal component analysis was less than 20%. Person reliability was calculated.

Results: A total of 212 participants were recruited in this study. Item 1 did not fit the model's assumptions and was deleted. The unidimensionality of the remaining 8 items (SDM-Q-8) was supported with good item fit (infit and outfit mean square ranging from 0.6 to 1.3) and very low unexplained variance of the first dimension (5.3%) of the principal component analysis. The person reliability of the SDM-Q-8 was 0.90.

Conclusions: The SDM-Q-8 was unidimensional and had good person reliability in women with breast cancer.

Implications for practice: The SDM-Q-8 has shown its potential for assessing the level of perceived involvement in SDM in women with breast cancer for both research and clinical purposes.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Breast Neoplasms / therapy*
  • Decision Making, Shared*
  • Female
  • Humans
  • Middle Aged
  • Reproducibility of Results
  • Surveys and Questionnaires*