Table 2.

Model predicting anticipated consultation benefit (n = 1032)

	Final model
	B	Sig	Odds	95% CI
Gender	0.363	0.019	1.44	1.06 to 1.95
Number of different chronically used medicationsa
0–3	−0.619	0.006	0.54	0.35 to 0.84
4–5	−0.317	0.152	0.73	0.47 to 1.12
6–7	−0.169	0.450	0.84	0.55 to 1.31
Contact with general practice over 2 years
Consultations	0.015	0.060	1.01	1.00 to 1.03
Repeat prescriptionsb
0	−1.833	0.000	0.16	0.11 to 0.23
1–3	−1.970	0.000	0.14	0.09 to 0.22
Home visits, dichotomised	0.735	0.000	2.09	1.45 to 3.00
Diagnosesc
Cancer	0.302	0.053	1.35	1.00 to 1.84
Infectious heart disease	1.470	0.029	4.35	1.16 to 16.28
Myocardial infarction	0.735	0.002	2.09	1.31 to 3.33
Cerebrovascular accident	0.390	0.074	1.48	0.96 to 2.26
Cox arthrosis	−0.510	0.021	0.60	0.39 to 0.93
Psoriasis	−0.565	0.037	0.57	0.33 to 0.97
Z episoded	−0.406	0.020	0.67	0.47 to 0.94
Constant	0.420	0.144	1.52	0.00 to 0.00
Model summary	Apparent value	Cross-validated value		95% CI
C-statistic	0.777	0.724		0.700 to 0.747

^a Reference category was ≥8 medications.

^b Reference category was ≥4 repeated drug prescriptions.

^c See Supplementary Table S1 for all included ICPC codes per cluster.

^d Presence of ICPC code(s) from the Z chapter (social problems, see Supplementary Table S3). B = beta. ICPC = International Classification of Primary Care. Sig = significance, the P-value.