Pia Liljamo1, Ulla-Mari Kinnunen2, Pasi Ohtonen3, Kaija Saranto2. 1. Division of Operative Care, Oulu University Hospital, Finland. 2. Department of Health and Social Management, University of Eastern Finland, Kuopio, Finland. 3. Division of Operative Care, Medical Research Center Oulu, Oulu University Hospital and University of Oulu, Finland.
Abstract
AIMS: The aim of this study was to measure the inter-rater reliability of the Oulu Patient Classification and to discuss existing methods of reliability testing. BACKGROUND: The Oulu Patient Classification, part of the RAFAELA® System, has been developed to assist nursing managers with the proper allocation of nursing resources. Due to the increased intensity of inpatient care during recent years, there is a need for the reliability testing of the classification, which has been in clinical use for 20 years. DESIGN: Retrospective statistical study. METHODS: To test inter-rater reliability, a pair of nurses classified the same patients, without knowledge of each other's ratings, as a part of annually conducted standardization. Data on the parallel classifications (n = 19,997) was obtained from inpatient units (n = 32) with different specialties at a university hospital in Finland during 2010-2015. Parallel classification practices were also analysed. The reliability of the overall classification and its subareas were calculated using suitable statistical coefficients. RESULTS: Inter-rater reliability coefficients were a reliable or almost perfect means of considering the nursing intensity category and various practices, but there were detectable differences between subareas. The lowest agreement levels occurred in the subareas 'Planning and Coordination of Nursing Care' and 'Guiding of Care/Continued Care and Emotional Support'. CONCLUSIONS: There is a need to develop the descriptions of subareas and to clarify the related concepts. Precise nursing documentation can promote a high level of agreement and reliable results. The traditional overall proportion of agreement does not provide an adequate picture of reliability - weighted kappa coefficients should be used instead.
AIMS: The aim of this study was to measure the inter-rater reliability of the Oulu Patient Classification and to discuss existing methods of reliability testing. BACKGROUND: The Oulu Patient Classification, part of the RAFAELA® System, has been developed to assist nursing managers with the proper allocation of nursing resources. Due to the increased intensity of inpatient care during recent years, there is a need for the reliability testing of the classification, which has been in clinical use for 20 years. DESIGN: Retrospective statistical study. METHODS: To test inter-rater reliability, a pair of nurses classified the same patients, without knowledge of each other's ratings, as a part of annually conducted standardization. Data on the parallel classifications (n = 19,997) was obtained from inpatient units (n = 32) with different specialties at a university hospital in Finland during 2010-2015. Parallel classification practices were also analysed. The reliability of the overall classification and its subareas were calculated using suitable statistical coefficients. RESULTS: Inter-rater reliability coefficients were a reliable or almost perfect means of considering the nursing intensity category and various practices, but there were detectable differences between subareas. The lowest agreement levels occurred in the subareas 'Planning and Coordination of Nursing Care' and 'Guiding of Care/Continued Care and Emotional Support'. CONCLUSIONS: There is a need to develop the descriptions of subareas and to clarify the related concepts. Precise nursing documentation can promote a high level of agreement and reliable results. The traditional overall proportion of agreement does not provide an adequate picture of reliability - weighted kappa coefficients should be used instead.
Authors: Peter Griffiths; Christina Saville; Jane Ball; Jeremy Jones; Natalie Pattison; Thomas Monks Journal: Int J Nurs Stud Date: 2019-11-29 Impact factor: 5.837