Objective: Conventional QOL measurement imposes a predetermined value system that is expressed in the domains that are taken into account in the instrument as well as in the weight by which they contribute to an overall score. Individual (ized?) measurements such as SEIQOL (-DW) are based on the premise that QOL is what the individual determines it to be. SEIQOL (- DW) enables respondents to nominate the five aspects (domains) that are most important to their individual QOL (‘‘elicited cues’’). Each aspect is rated (VAS) and the relative importance of each domain to overall QOL is weighted. Following a suggestion by Schwartz & Sprangers (1999) we used SEIQOL-DW for measuring response shift in QOL. In this presentation we discuss problems that we encountered when using SEIQOL-DW for this purpose.
Methods: A qualitative exploratory multiple case study (n=31 patients with small cell lung cancer); SEIQOL-DW; audio recordings and transcripts of administration of SEIQOL-DW.
Results: An often arduous probing process was necessary to elicit cues. Elicited cues are a heterogeneous mixture of attitudes, traits, threats, values, domains, roles etc. (such as ‘‘being positive’’; ‘‘pain’’; ‘‘time to live’’; ‘‘being a grandmother’’). Usually the meaning of such cues for respondents could only be established by further probing. To elicit five cues was sometimes difficult. After completing the SEIQOL-DW, several patients noticed that they had forgotten to mention a cue important to them, such as ‘‘health’’.
Conclusion: As SEIQOL-DW measures are the outcome of an often arduous probing process, it is vulnerable for interviewer bias. As SEIQOLDW is primarily intended as an individual (ized?) measure, the requirement of eliciting a number of five cues seems to be arbitrary. Transcripts are crucial in the analysis of the meaning of elicited cues as well as of the interpretation of their weighting.