Measuring morbidity of children in the community: a comparison of interview and diary data

BACKGROUND: Little is known about the validity of estimates of morbidity experienced at home. METHODS: In the Dutch National Survey of Morbidity and Interventions in General Practice mothers of 1630 children answered a health interview and kept a health diary for 3 weeks (only the first 2 weeks were used). Children's symptoms were recorded during the interview using a check list and monitored in the health diary through open-ended questions. RESULTS: In the interview parents reported symptoms for 65% of their children and in the diary for 54% of children. Ear problems, colds, fever and weakness and anxiety were reported more often in the interview. Mother's mental health was assessed by the General Health Questionnaire; those scoring >4 were assessed as having impaired mental health and these parents reported symptoms for more children in the interview (81%) than in the diary (65%). For similar reference periods, the least educated mothers reported fewer children with symptoms in the diary (45%) than in the interview (66%). More highly educated mothers reported similarly in the diary (67%) and the interview (70%). CONCLUSION: Both data collection methods yield different estimates of community morbidity. Explanations such as telescoping, the seriousness of the symptoms, the amount of psychological distress of the respondent, forgetfulness and literacy limitations are discussed. We recommend that diaries should not be used in less educated populations.