A randomized comparison of three data collection models for the measurement of parent experiences with diabetes
Oyvind Bjertnaes1, Hilde Hestad Iversen2, Torild Skrivarhaug3
1Norwegian Institute of Public Health, Nydalen, PO Box 4404, 0403, Oslo, Norway. oyvindandresen.bjertnaes@fhi.no.
Insights
Pen-and-paper or mixed-mode surveys for child diabetes care yield higher response rates than electronic-only options. However, electronic surveys are more cost-effective, with similar parent experience data collected across all methods.
Area of Science:
- Health Services Research
- Pediatric Endocrinology
- Survey Methodology
Background:
- Measuring parent experiences in pediatric outpatient care is crucial for quality improvement.
- Previous research has not extensively compared mixed-mode (pen-and-paper and electronic) versus single-mode data collection for this population.
- Understanding optimal data collection methods can enhance patient engagement and reduce survey costs.
Purpose of the Study:
- To compare the effectiveness of three distinct data collection methods for measuring parent experiences in pediatric diabetes outpatient care.
- To evaluate response rates, respondent characteristics, reported experiences, and survey costs across different data collection modalities.
- To inform best practices for data collection in studies involving parents of children with diabetes.
Main Methods:
- A randomized national trial in Norway involving 2606 parents of children with diabetes.
- Three groups were established: pen-and-paper only, mixed-mode (pen-and-paper and electronic), and electronic-only response options.
- Statistical analyses included logistic and multilevel linear regression to compare response probabilities and parent experiences.
Main Results:
- Response rates were significantly higher for pen-and-paper (61.8%) and mixed-mode (62.4%) groups compared to electronic-only (41.6%).
- Parent-reported experiences showed minimal significant differences between groups, with only the organization scale differing significantly between mixed-mode and electronic-only groups.
- The electronic-only model was the most cost-effective, with lower costs per response compared to pen-and-paper and mixed-mode approaches.
Conclusions:
- While pen-and-paper or mixed-mode surveys achieve higher response rates, electronic-only methods offer significant cost-effectiveness.
- Parent experiences and respondent demographics are comparable across the tested data collection methods.
- The choice of data collection method should balance response rate considerations with cost-efficiency and specific study objectives.
Background:
The objective of this study was to compare three data collection methods for the measurement of parent experiences with hospital outpatient care for child and adolescent diabetes, based on a randomised national trial in Norway involving both pen-and-paper and electronic response options.
Methods:
The sample frame was patients registered in the Norwegian Childhood Diabetes Registry. Parents of patients were randomised into the following groups (n = 2606): group A, who were posted questionnaires with only a pen-and-paper response option (n = 859); group B, who were posted questionnaires with both an electronic and a pen-and-paper response option (n = 886); and group C, who were posted questionnaires with only an electronic response option (n = 861). The three groups were compared on response rate, background variables about respondents, main study results and survey costs. Statistical analysis included logistic regression to test group differences in response probabilities and multilevel linear regression to test differences in parent experiences.
Results:
The response rate was 61.8% for group A, 62.4% for group B and 41.6% for group C. The probability of answering was significantly higher for group A (OR = 2.3, p < 0.001) and B (OR = 2.3, p < 0.001) compared to group C. Respondent age, gender, education, living with the child and the degree of participation in consultations did not differ significantly between the three groups. Group differences in parent-reported experiences were small, varying from 1.0 (equipment and doctor contact) to 2.4 (outcome), on a scale from 0 to 100. Only one of 18 group differences was significant: the mixed group had significantly higher score than the electronic group on the organization scale (p < 0.05). The total cost of the electronic model was less than half the cost of the other models, and cost per response was 5.1 euros for the electronic model compared to 8.2 euros for group A and 7.6 euros for group B.
Conclusions:
The models with pen-and paper questionnaire included had more than 20% higher response rate than the model with an electronic-only response option. Background variables and parent-reported experiences were similar between the three groups, and the electronic model was the more cost-effective model.
Related Concept Videos
01:13Data Collection by Experiments
An example of the experimental method is a public...
00:55Parental Care
01:13Randomized Experiments
Simple randomization
Simple...
01:29Data Collection II
01:30Data Collection I
01:07Data Collection by Survey