OBJECTIVE: To identify the characteristics associated with better structural quality of protocols.Design. Analytic and transversal.Setting. Health centres in Spain with physiotherapy protocols. PARTICIPANTS: Guidelines for physiotherapy practice or protocols worked out between 1990 and 1996, inclusive. MAIN MEASUREMENTS: The design quality of the documents was evaluated by using percentages of compliance with eight explicit validated criteria and the proportion of criteria complied with in each protocol out of the total possible. To identify the characteristics linked to structural quality, multiple and logistic regression multivariate analyses were performed. RESULTS: There was greater structural quality in documents worked out by teams of over five authors and after 1992. Variables that did not affect structural quality were the source of the protocol (Insalud or autonomous community with devolved powers) and the kind of professional (only physiotherapists or multi-disciplinary) who worked it out. CONCLUSIONS: The quality of the documents varies according to the date and the number of authors who work them out. There was significantly greater quality when this was after 1992 and there were over five authors. Therefore, under the opposite circumstances (few authors), there is greater risk of low-quality documents.
OBJECTIVE: To identify the characteristics associated with better structural quality of protocols.Design. Analytic and transversal.Setting. Health centres in Spain with physiotherapy protocols. PARTICIPANTS: Guidelines for physiotherapy practice or protocols worked out between 1990 and 1996, inclusive. MAIN MEASUREMENTS: The design quality of the documents was evaluated by using percentages of compliance with eight explicit validated criteria and the proportion of criteria complied with in each protocol out of the total possible. To identify the characteristics linked to structural quality, multiple and logistic regression multivariate analyses were performed. RESULTS: There was greater structural quality in documents worked out by teams of over five authors and after 1992. Variables that did not affect structural quality were the source of the protocol (Insalud or autonomous community with devolved powers) and the kind of professional (only physiotherapists or multi-disciplinary) who worked it out. CONCLUSIONS: The quality of the documents varies according to the date and the number of authors who work them out. There was significantly greater quality when this was after 1992 and there were over five authors. Therefore, under the opposite circumstances (few authors), there is greater risk of low-quality documents.