OBJECTIVE: To examine the effect of questionnaire context on self-reported illness cognition. DESIGN: A single-item measure of the perceived impact of lower urinary tract symptoms (LUTS) was embedded twice in a questionnaire battery completed by community-dwelling middle-aged males (N = 1,790). The impact measure was placed in two distinct questionnaire contexts; at the end of a general somatic symptoms questionnaire, and following an illness-specific symptoms questionnaire. The order of the two questionnaire contexts was counterbalanced in a random subsample. MAIN OUTCOME MEASURES: An established single-item measure of the perceived impact of LUTS. RESULTS: Concordance between the two single-item measures was moderate. Scores on a single-item measure of impact were significantly lower when assessed immediately following the completion of a LUTS-specific questionnaire than when assessed following the completion of a general symptoms questionnaire. There was no evidence of order effects. The observed effect was moderated by the severity of LUTS such that the difference in perceived impact scores between contexts (where general symptoms context > illness-specific context) increased as urinary symptoms increased. CONCLUSION: Questionnaire context systematically influenced responses on self-report measures of illness impact. The magnitude of the context effect was largest in the highest quintile of LUTS severity, a difference of 0.5 on a scale with a range of 3. These findings may have implications for situations where patient reported outcome measures are used to evaluate health care interventions or inform treatment decisions.
OBJECTIVE: To examine the effect of questionnaire context on self-reported illness cognition. DESIGN: A single-item measure of the perceived impact of lower urinary tract symptoms (LUTS) was embedded twice in a questionnaire battery completed by community-dwelling middle-aged males (N = 1,790). The impact measure was placed in two distinct questionnaire contexts; at the end of a general somatic symptoms questionnaire, and following an illness-specific symptoms questionnaire. The order of the two questionnaire contexts was counterbalanced in a random subsample. MAIN OUTCOME MEASURES: An established single-item measure of the perceived impact of LUTS. RESULTS: Concordance between the two single-item measures was moderate. Scores on a single-item measure of impact were significantly lower when assessed immediately following the completion of a LUTS-specific questionnaire than when assessed following the completion of a general symptoms questionnaire. There was no evidence of order effects. The observed effect was moderated by the severity of LUTS such that the difference in perceived impact scores between contexts (where general symptoms context > illness-specific context) increased as urinary symptoms increased. CONCLUSION: Questionnaire context systematically influenced responses on self-report measures of illness impact. The magnitude of the context effect was largest in the highest quintile of LUTS severity, a difference of 0.5 on a scale with a range of 3. These findings may have implications for situations where patient reported outcome measures are used to evaluate health care interventions or inform treatment decisions.