Skip to main content
. 2022 Jan 24;15(4):473–483. doi: 10.1007/s40271-021-00566-4

Table 5.

Mixed logit regression result

Attributesa β(SE) SD (SE) WTP (£) [95% CI]
Constant
 Opt-in for health check 3.126*** (0.267) 4.113*** (0.258) 36.22 [30.69 to 41.76]
 Day of health check
 On a weekday − 0.101** (0.044) REF − 1.17 [− 2.18 to − 0.16]
 At the weekend 0.101** (0.044) 0.681*** (0.050) 1.17 [0.16 to 2.18]
Access to health check
 Walk in and wait − 0.248*** (0.043) REF − 2.87 [− 3.78 to − 1.96]
 By appointment 0.248*** (0.043) 0.424*** (0.051) 2.87 [1.96 to 3.78]
Provider of health check
 Nurse 0.539*** (0.071) REF 6.24 [4.61 to 7.88]
 Pharmacist 0.414*** (0.044) 0.287*** (0.071) 4.80 [3.90 to 5.69]
 Trainee pharmacist − 0.953*** (0.074) 0.942*** (0.072) − 11.04 [− 12.60 to − 9.48]
Duration of health check
 30 min 0.153*** (0.025) REF 1.77 [1.21 to 2.32]
 45 min − 0.153*** (0.025) 0.057 (0.125) − 1.77 [− 2.32 to − 1.21]
Follow-up call
 No follow-up − 0.308*** (0.040) REF − 3.57 [− 4.41 to − 2.72]
 Follow-up call after 3 months 0.308*** (0.040) 0.509*** (0.045) 3.57 [2.72 to 4.41]
 Cost of health check − 0.086*** (0.004)
 Number of respondents 404
 Log-likelihood − 3604
 BIC 7336
 McFadden Adjusted R2 0.326

BIC Bayesian information criterion, CI confidence interval, SD standard deviation, SE robust standard errors, WTP willingness-to-pay

***p < 0.01, **p < 0.05, *p < 0.1

aCategorical attribute levels were effects coded; the coefficient for the reference level (REF) was calculated by taking the negative sum of the estimated coefficients of all other levels