Reliability of the VCM1 Questionnaire when administered by post and by telephone

To assess the reliability of different methods of administration of the VCM1 vision-related quality-of-life questionnaire by: a) comparing responses obtained by post to responses obtained in a research clinic and b) comparing responses obtained by telephone to responses obtained in a research clinic.

Questionnaire responses given in advance by post (96 subjects) or by telephone (92 subjects) were compared to those subsequently given at a visit to a research clinic. The questionnaire included the VCM1 and two other questions commonly used in surveys of visual impairment (reading small print and recognising a face across the street).

Similar levels of vision-related quality-of-life (VR-QOL) impairment were reported by post and in the research clinic. However, the participants in the telephone test group reported less VR-QOL impairment by telephone than they subsequently reported in the clinic (P = 0.0001). The mean score difference between telephone and clinic administration was 3.2% of the VCM1 questionnaire scale. Lower social class (P = 0.002) and increasing duration of interview (P = 0.003) were associated with a tendency to under-report VR-QOL impairment by telephone. Interference with reading small print (P = 0.0001) and recognising a face across the street (P = 0.0001) were also under-reported by telephone.

Telephone interviewing caused a general bias towards under-reporting of visual problems which was not confined to the VCM1. Care is required when planning outcome studies and questionnaire surveys to ensure that different methods of questionnaire administration produce comparable results.