| Autor: | Alain Leplège, Emmanuel Ecosse, E. Dohin, R. Ceinos, Jacques Pouchot, N. Rude |
|---|---|
| Rok vydání: | 1997 |
| Předmět: |
Rasch model
Psychometrics business.industry Public Health Environmental and Occupational Health medicine.disease Developmental psychology Acquired immunodeficiency syndrome (AIDS) Quality of life Cronbach's alpha Item response theory medicine Set (psychology) Construct (philosophy) business Clinical psychology |
| Zdroj: | Quality of Life Research. 6:585-594 |
| ISSN: | 0962-9343 |
| Popis: | Assessment of the quality of life (QoL) of human immunodeficiency virus (HIV)-infected subjects is often based on questionnaires in which the items or questions are not seen to be relevant by patients, nor by the users of the data obtained. It therefore seemed appropriate to return to the issue. The methodological and bibliographical research as well as the consultations we conducted convinced us that the elaboration of a new questionnaire was both necessary and possible. In order to do so, we adopted methodological principles based on the Sickness Impact Profile development methodology. First a bibliographical research was conducted in order to study instruments already used for HIV infection. Then, experts concerned with HIV infection and members of patients' associations were interviewed to assess how opportune the development of a new instrument could be. Following this, a methodology was established for the design and construction of the new instrument. One hundred and eighteen candidate questions were generated from an analysis of the content of 20 patients' interviews, which were subsequently submitted to 102 patients, to obtain finally a set of 31 questions from the interpretation of the results obtained from classic psychometric analysis and also from non-classic methods (item response theory and Rasch model). The concept being measured is the impact of illness being experienced by HIV-infected subjects from their own perspective. The range of health states covered by this questionnaire relates to fairly mild conditions. Rasch analysis of this set of 31 questions (HIV-QL31) shows that it corresponds to one unidimensional construct. A single score can be calculated by simple summation of dichotomous response options. This score is highly reliable (Cronbach's α coefficient = 0.93) and is also discriminant regarding the severity of clinical status. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|