Relatives' view on collaboration with nurses in acute wards: development and testing of a new measure.

ABSTRACT

BACKGROUND: Collaboration between relatives and nurses in acute care settings is sparsely investigated, and that mostly from nurses' point of view. Feasible and valid instruments are needed for assessing collaboration, its prerequisites and outcome. OBJECTIVES: To develop and test an instrument to assess, from the relatives' perspective, collaboration between relatives of frail elderly patients and nurses in acute hospital wards, as well as prerequisites for, and outcome of, collaboration. DESIGN: Instrument development and psychometric testing. SETTING: Acute medical and geriatric wards. PARTICIPANTS: One hundred fifty-six relatives. Women constituted 74.8%, offspring 63.9% and spouses 20%, respectively. METHODS: A model for collaboration was developed and underpinned the development and construction of the instrument. Face and content validity was examined by relatives and an expert panel, before testing it among 156 relatives. Construct validity was assessed by principal component analysis and test for correlation between factors. Predictive validity was assessed by comparing factor scores with scores in outcome measures. Internal consistency was assessed by Cronbach's alpha for factors, item-to-total correlation and item-to-item correlation. Systematic internal dropout was investigated. RESULTS: A five-factor solution labelled "influence on decisions", "quality of contact with nurses", "trust and its prerequisites", "achieved information level" and "influence on discharge" showed Cronbach's alpha values between .83 and .94. Correlation between factors showed coefficients between .16 and .60. Item-to-total correlation values ranged between .34 and .83 and mean inter-item-correlation coefficients between .40 and .56. Predictive validity was indicated. Systematic internal dropout was related to higher age and lower educational level. CONCLUSION: The instrument was mainly valid and reliable. The instrument is, to our knowledge, the first of its kind and should be tested on larger samples in various cultural contexts. The feasibility of the instrument may benefit from a reduction of number of items.