Gavin Andrews1. 1. World Health Organization Collaborating Centre in Evidence for Mental Health Policy and School of Psychiatry, University of New South Wales at St Vincents Hospital, Darlinghurst. gavina@unsw.edu.au
Abstract
OBJECTIVE: The SF-36 and the shorter SF-12 are widely used in clinical research and increasingly so in practice. Scoring is complex and the validity of a brief scorer for the SF-12 in a community and a clinic sample was examined. METHOD: The sample comprised respondents with complete data on the SF-12 from the community sample in the 1997 Australian National Survey of Mental Health and Well-being (n = 10,641) and from a clinical sample from the Clinical Research Unit for Anxiety and Depression (n = 1,725). Each SF-12 was scored by the standard method (using weights to five decimal places) or by the brief method (using weights that are integers only). RESULTS: Scores generated by the standard scorer correlated almost perfectly with the brief scorer in both the community and clinic samples. Means and standard deviations were similar and no individual scores deviated by more than 2.89 in the community sample or by 3.06 in the clinical sample. IMPLICATIONS: The brief, rounded integer scorer for the SF-12 is an appropriate substitute for the standard scorer when hand scoring is an advantage.
OBJECTIVE: The SF-36 and the shorter SF-12 are widely used in clinical research and increasingly so in practice. Scoring is complex and the validity of a brief scorer for the SF-12 in a community and a clinic sample was examined. METHOD: The sample comprised respondents with complete data on the SF-12 from the community sample in the 1997 Australian National Survey of Mental Health and Well-being (n = 10,641) and from a clinical sample from the Clinical Research Unit for Anxiety and Depression (n = 1,725). Each SF-12 was scored by the standard method (using weights to five decimal places) or by the brief method (using weights that are integers only). RESULTS: Scores generated by the standard scorer correlated almost perfectly with the brief scorer in both the community and clinic samples. Means and standard deviations were similar and no individual scores deviated by more than 2.89 in the community sample or by 3.06 in the clinical sample. IMPLICATIONS: The brief, rounded integer scorer for the SF-12 is an appropriate substitute for the standard scorer when hand scoring is an advantage.
Authors: Sina Babazadeh; Michelle M Dowsey; James D Stoney; Peter F M Choong Journal: Knee Surg Sports Traumatol Arthrosc Date: 2012-10-21 Impact factor: 4.342