Appendix 4

Multilevel analysis with the square root of consultations during the post-intervention year as the dependent variable, without imputed data (model 1), and with imputed data and intention-to-treat (model 2).

Multilevel modelModel 1 (n = 166)Model 2 (n =209)
VariablesBetaSE (beta)P-valueBetaSE (beta)P-value
 Timea−1.06250.1015<0.001−0.98140.1394<0.001
 Group Ab−0.45110.17460.010−0.13770.14590.345
 Group Bc−0.11850.25420.641−0.00720.18020.968
 Interaction between time and group A0.97880.1421<0.0010.52100.17760.003
 Interaction between time and group B1.34210.1464<0.0010.87080.1916<0.001
 Age0.03080.0047<0.0010.03750.0037<0.001
 Trait anxiety (STAI-T)0.00590.00680.3860.00780.00550.158
 Satisfaction with GP (group A)−0.23530.18170.195−0.12600.14580.388
 Satisfaction with GP (group B)0.00040.17880.9980.02770.14740.851
 Logarithm of number of chronic diseases0.46150.15430.0030.28900.12500.021
 Hypochondria (GH+WI)0.04390.02070.0340.02050.01560.190
Random-effects parameters
 Doctor0.20600.09990.10550.0874
 Patient0.72350.05310.51500.0644
 Residual0.52750.02950.68750.0422
  • a Time = observation time (baseline and 1 year post-intervention).

  • b Group A = contrast between intervention group and control group 1.

  • c Group B = contrast between intervention group and control group 2. SE: standard error. STAI = State–Trait Anxiety Inventory.23 GH+WI = General Hypochondria and Witheley Index from the Illness Behaviour Questionnaire.26