engleski

Alternative Data Processing in the Culig Adherence Scale

Background: Perseverance in taking therapy is an important determinant of its success, but the evaluation of patient adherence is extremely complex. The Culig scale with 16 questions of perseverance is used in Croatia. Objectives: The determination of psychometric properties of the Culig questionnaire of perseverance, in which the individual results are formed in an alternative way. Methods: Raw values of each indicator, 68 of them, are processed, instead of the original 16 composite variables of the Culig scale. At the same time, reactions on certain indicators are normalized by the inverse integral of the standard normal distribution, after which the data are standardized. After the data are processed, results are formed for each subject as factor scores on the first main component of the inter-correlation matrix of indicators. Such a procedure allows calculation of the contribution of individual indicators to the overall results on the scale, and with the method of the factor analysis it is possible to accurately show the interconnection of individual indicators, as well as their possible groupings. Results: Scale reliability (Cronbach alpha = 0.94) becomes significantly higher than in classical procedures. The first factor is defined by the highest number of individual indicators and represents the main subject of the whole scale measurement ; indicators with the highest saturation are connected to the perseverance in taking the drug and to the indicators of the subjects' mood, with an emphasis on the sad and depressed mood. Interestingly, the indicators of the relationship of the patient with the doctor or the pharmacist were not significantly associated with indicators of perseverance in taking the drug, but are on an independent dimension. Conclusions: Alternative methods of processing the individual results and psychometric properties determined by that process are presented in the Culig questionnaire of perseverance. These kind of properties are better than those that are usual of certain conventional ways of processing. Such a process opens the possibility for a multivariate analysis of the results in the questionnaire, which is important for further research.

adherence; scale; reliability

nije evidentirano