Assessment of data quality in a multi-centre cross-sectional study of participation and quality of life of children with cerebral palsy
1 School of Population and Health Sciences, Newcastle University, 21 Claremont Place, Newcastle, NE2 4AA, UK
2 School of Clinical Medical Sciences, Newcastle University, 4th Floor William Leech Building, Medical School, Framlington Place, Newcastle, NE2 4HH, UK
3 Enable Ireland, Lavanagh Centre, Ballintemple, Cork, Ireland
4 Institut National de la Santé et de la Recherche Medicale, U558 Département de Santé Publique, Faculté de Médecine, 37 allées J Guesde, 31 073 Toulouse, France
5 Göteborg University, The Queen Silvia Children's Hospital, S-41685 Göteborg, Sweden
6 Université Joseph Fournier, SIIM-Pole Exploitation, CHU de Grenoble BP 217, 38043 Grenoble cedex 9, France
7 National Institute of Public Health, Oster Farimagsgade 5, 1399 Copenhagen, Denmark
8 School of Nursing & Midwifery, Queen's University Belfast, 21 Stranmillis Road, Belfast, BT9 5AF, Northern Ireland
9 Azienda Sanitaria Locale Viterbo, Viale Trento 18 H, 01100 Viterbo, Italy
10 Klinik fur Kinder, Universitatsklinikum Schleswig-Holstein, Ratzeburger Allee 160, 23538 Lubeck, Germany
11 Sir James Spence Institute, Newcastle University, Royal Victoria Infirmary, Newcastle NE1 4LP, UK
BMC Public Health 2006, 6:273 doi:10.1186/1471-2458-6-273Published: 6 November 2006
SPARCLE is a cross-sectional survey in nine European regions, examining the relationship of the environment of children with cerebral palsy to their participation and quality of life. The objective of this report is to assess data quality, in particular heterogeneity between regions, family and item non-response and potential for bias.
1,174 children aged 8–12 years were selected from eight population-based registers of children with cerebral palsy; one further centre recruited 75 children from multiple sources. Families were visited by trained researchers who administered psychometric questionnaires. Logistic regression was used to assess factors related to family non-response and self-completion of questionnaires by children.
431/1,174 (37%) families identified from registers did not respond: 146 (12%) were not traced; of the 1,028 traced families, 250 (24%) declined to participate and 35 (3%) were not approached. Families whose disabled children could walk unaided were more likely to decline to participate. 818 children entered the study of which 500 (61%) self-reported their quality of life; children with low IQ, seizures or inability to walk were less likely to self-report. There was substantial heterogeneity between regions in response rates and socio-demographic characteristics of families but not in age or gender of children. Item non-response was 2% for children and ranged from 0.4% to 5% for questionnaires completed by parents.
While the proportion of untraced families was higher than in similar surveys, the refusal rate was comparable. To reduce bias, all analyses should allow for region, walking ability, age and socio-demographic characteristics. The 75 children in the region without a population based register are unlikely to introduce bias.