How is CIVIQ scored?
The score for each dimension was obtained by adding up the scores of each constituent item and the global index was obtained by adding up the scores of the 20 items. Items were weighted equally. The minimum and maximum values of the scales are dependent on the number of items used in each of the dimensions and on the number of levels or categories for each item. In order to compare the mean scores between dimensions or scales, absolute scores were then converted into an index. The method chosen was the one described by John E. Ware for SF-36.1 For each dimension, we calculated S, the sum of the scores for the patients’ answers to the questions; m, the minimum theoretical value if all of the answers were on the first level of the scale for all of the items belonging to the dimension; and M, the maximum theoretical value if all of the items were scored at the maximum level on the scale for all items belonging to the dimension. The standardized score for each dimension was obtained by applying the following equation: (S-m)/(M-m)X100. For each dimension, we therefore obtained a result ranging from 0 to 100. In order to facilitate interpretation of the results, the scoring system can be reversed: the highest figure can be allocated to the lowest response option and vice versa so as to obtain a score directly proportional to the quality of life. According to this scoring method, improvement in quality of life between two study times is represented by an increase in score.
The strength of such a scoring method is that whatever the number of items a questionnaire or one of its dimension is made of, the scores always range from 0 to 100. This allows consistency of the scale since the number of items in each of dimension is different.
- Ware JE Jr. The SF-36 Health Survey. In: Spilker B, ed. Quality of life and pharmacoeconomics in clinical trials. 2nd ed. Philadelphia, Pa: Lippincott-Raven; 1996:337-345.
How to handle missing data for CIVIQ?
Questionnaires with 3 or more missing scores should be excluded from analysis.
The procedure entitled ‘multiple imputation’ results in correctly estimated standard errors and confidence intervals and is recommended.1
The handling of missing data should be individually debated with statisticians.
- Dondersa ART, van der Heijdenc GJM, Stijnend T, Moonsc KGM. Review: A gentle introduction to imputation of missing values. J Clin Epidemiol. 2006;59:1087-1091.