Reliability and validity of the Northwick Park Dependency Score (NPDS) Swedish version 6.0

Abstract

Objective: To translate, to test inter-rater and intra-rater reliability and concurrent validity of the Basic Care Needs (BCN) section of the Northwick Park Dependency Score (NPDS). Design: Test–retest reliability and validity testing. Observed data were collected by the staff (nursing staff, occupational therapists). Setting: Three rehabilitation units. Subjects: Forty inpatients between 16 and 65 years of age with brain injury were included. Main measures: Inter-rater and intra-rater reliability was calculated by percentage agreement (PA) and unweighted kappa measure. Concurrent validity was examined by computing Goodman–Kruskal's gamma, a nonparametric statistic for degree of association between the BCN score and the total score of the Functional Independence Measure (FIM). Results: Inter-rater reliability showed a good percentage agreement between nursing staff. Between nursing staff and occupational therapists the percentage agreement was lower especially in one item. Intra-rater reliability showed a good percentage agreement for all assessors. Concordance was good with a gamma 0.83 and an asymptotic error (ase) of 0.04, for nursing staff and for occupational therapists — 0.87, ase 0.04. Conclusion: The BCN section of NPDS was found to be inter-rater and intra-rater reliable, to have concurrent validity. Further studies are needed on clinical utility. The instrument can be used for assessment of dependency for individual patients with brain injury, and information when transferring between different caregivers. Further studies need to investigate the sensitivity of the instrument.