Demet Unalan1, Ferhan Soyuer, Ahmet Ozturk. 1. Department of Vocational College Health Services, School of Medicine, Erciyes University, Kayseri, Turkey. Tel. +90 (352) 4375279. Fax. +90 (352) 4375936. E-mail: dunalan@erciyes.edu.tr.
Abstract
OBJECTIVE: To compare the usefulness of the Nottingham Health Profile (NHP) and the Short Form 36 (SF-36) as general outcome measures for chronic stroke patients with respect to the response rate, internal consistency reliability, validity analyses, and agreement in similar domains of the 2 instruments. METHODS: In this prospective study, 90 consecutive stroke patients attending the Neurology outpatient clinic at Erciyes University, Kayseri, Turkey from March 2004 to March 2005 were evaluated for the study. Seventy outpatients who had a stroke 6 months previously were included in the study. As a datacollecting device, SF-36 and NHP scales were used. RESULTS: The prevalence of patients with highest quality of life score (ceiling effect) was higher for the NHP scale (10-35.7%) than for the SF-36 scale (1.4-37.1%). The prevalence of patients with lowest quality of life score (floor effect) was also higher for the NHP scale (1.4-45.7%) than for the domains of SF-36 (1.4-30%). The internal consistencies of the subscales of both the SF-36 and the NHP showed satisfactory values. Regarding convergent validity, correlations were found between comparable subscales of the 2 instruments. Limits of agreement in similar domains of the 2 instruments were very large. In all 5 demonstrated Bland-Altman plots, there was agreement of the scales in the measurement of the similar fields of quality of life. CONCLUSION: Both the SF-36 and the NHP have acceptable degrees of reliability, convergent validity, and response rate. Limits of agreement in similar domains of the 2 instruments were very large.
OBJECTIVE: To compare the usefulness of the Nottingham Health Profile (NHP) and the Short Form 36 (SF-36) as general outcome measures for chronic strokepatients with respect to the response rate, internal consistency reliability, validity analyses, and agreement in similar domains of the 2 instruments. METHODS: In this prospective study, 90 consecutive strokepatients attending the Neurology outpatient clinic at Erciyes University, Kayseri, Turkey from March 2004 to March 2005 were evaluated for the study. Seventy outpatients who had a stroke 6 months previously were included in the study. As a datacollecting device, SF-36 and NHP scales were used. RESULTS: The prevalence of patients with highest quality of life score (ceiling effect) was higher for the NHP scale (10-35.7%) than for the SF-36 scale (1.4-37.1%). The prevalence of patients with lowest quality of life score (floor effect) was also higher for the NHP scale (1.4-45.7%) than for the domains of SF-36 (1.4-30%). The internal consistencies of the subscales of both the SF-36 and the NHP showed satisfactory values. Regarding convergent validity, correlations were found between comparable subscales of the 2 instruments. Limits of agreement in similar domains of the 2 instruments were very large. In all 5 demonstrated Bland-Altman plots, there was agreement of the scales in the measurement of the similar fields of quality of life. CONCLUSION: Both the SF-36 and the NHP have acceptable degrees of reliability, convergent validity, and response rate. Limits of agreement in similar domains of the 2 instruments were very large.