Development of an assessment tool for the multidisciplinary evaluation of neurological dependency: preliminary findings

• Published in: Scandinavian journal of caring sciences 2014-03, Vol.28 (1), p.193-203
• Main Authors: Pierce, Elaine, McLaren, Susan
• Format: Article
• Language: English
• Subjects: Adult; assessment instrument; construct validity; Humans; mixed methods research; multidisciplinary team; Nervous System Diseases - diagnosis; Nervous System Diseases - physiopathology; neurological dependency; Nursing; Physical Examination; rehabilitation; Reproducibility of Results; severe and complex disability; tool development
• ISSN: 0283-9318; 1471-6712
• DOI: 10.1111/scs.12018

Background Multidisciplinary inputs are of vital importance in the comprehensive care management and rehabilitation of individuals with long‐term neurological conditions. Although many ordinal measures of disability correlate with patient dependency and care, few are based on numbers of different health professionals required, or their skill‐mix, for assistance associated with levels of dependency. Aim and objectives To develop an instrument to assess dependency in adult patients suffering from severe and complex neurological disability for use by multidisciplinary teams. Objectives were to establish content validity, construct validity, inter‐rater reliability and internal consistency, and to demonstrate preliminary clinical utility from the perspective of professional staff. Design and methods A mixed methods design utilised qualitative and quantitative approaches. Stage I involved developmental fieldwork using focus groups (n = 3), questionnaires (n = 70), expert panel (n = 20) and direct observation (n = 12). In stage 2, intensive refinement through direct observation, construct validity and reliability testing (n = 100) of the instrument was completed. The research complied with the research governance and ethical standards set by the Department of Health (UK). Results In the final format, an instrument was developed which comprised two sections. Section 1: an ordinal scaled basic care section, with 15 categories of dependency and level of descriptors based on numbers, grade and time taken by staff providing assistance. Factor analysis identified one dominant and four subsidiary factors which explained 67% of the total variance. Overall inter‐rater agreement was 0.87 (kappa coefficient: range 0.67–1.0), and Cronbach's coefficient alpha was 0.733. Section 2: comprised a nominally measured specialist care section for inputs from nurses and therapists. The mean time taken to complete assessments was 12 minutes. Conclusion The instrument satisfied preliminary selective criteria for validity and reliability. Further research is necessary to satisfy other requirements of psychometric testing and clinical utility.