Skip to main content
. 2014 Jan 18;23(12):1189–1200. doi: 10.1007/s00787-013-0515-6

Table 4.

Results of the logit model by adolescent and child groups, in order of importance of attributes (most important to least important)

Attributea Adolescent Child
OR (SE)b 95 % CI OR (SE)b 95 % CI
Degree of symptom control
 Reference group: minimally improved
  Much improved 2.70 (0.16) 2.41, 3.03 3.61 (0.17) 3.3, 3.95 
  Very much improved 4.85 (0.31) 4.28, 5.49 6.37 (0.31) 5.79, 7.01
Duration of symptom control
 Reference group: 4–6 h
  10 h 1.38 (0.08) 1.22, 1.55 1.34 (0.06) 1.23, 1.47
  12 h 1.59 (0.10) 1.41, 1.79 1.60 (0.08) 1.46, 1.76
Loss of appetite
 1 % increase in risk of loss of appetite 0.98 (0.01) 0.97, 0.99 0.95 (0.01) 0.94, 0.96
Potential of treatment abuse
 1 % increase in potential of treatment abuse 0.97 (0.00) 0.97, 0.98
Vomiting
 1 % increase in risk of vomiting 0.97 (0.01) 0.96, 0.98 0.96 (0.01) 0.95, 0.97
Sleep disturbance
 1 % increase in risk of sleep disturbance 0.96 (0.00) 0.95, 0.97 0.96 (0.02) 0.95, 0.96
Frequency of medication
 One additional administration in dosage per day 0.89 (0.03) 0.83, 0.94 0.86 (0.02) 0.82, 0.90

CI confidence interval, OR odds ratio, SE standard error

aIn this model, the attributes are the independent variables and parents’ choice data are the dependent variable. All attributes are significant predicators (p < 0.01)

bStandard error shows whether the ORs are significantly different from 1