Table 3.
Mixed multinomial logit model estimates
| Attributes and levels | Mixed multinomial logit | Mixed multinomial logit (with systematic preference heterogeneity) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | |||||||||
| COEFF | SE | 95% CI | COEFF | SE | 95% CI | COEFF | SE | 95% CI | COEFF | SE | 95% CI | |
| ASC1 | − 1.06*** | 0.15 | − 1.36 to − 0.76 | – | – | – | − 1.07*** | 0.15 | − 1.37 to − 0.77 | – | – | – |
| ASC2 | − 1.13*** | 0.15 | − 1.42 to − 0.84 | – | – | – | − 1.14*** | 0.15 | − 1.44 to − 0.85 | – | – | – |
| ASC3 | 0.40*** | 0.13 | 0.14 to 0.66 | – | – | – | 0.40*** | 0.13 | 0.14 to 0.67 | – | – | – |
| Modality | ||||||||||||
| Postal letter | Reference | |||||||||||
| 1.61*** | 0.14 | 1.33 to 1.90 | 1.09*** | 0.15 | 0.79 to 1.39 | 2.45*** | 0.40 | 1.66 to 3.24 | 1.10*** | 0.16 | 0.80 to 1.41 | |
| Robot phone call | 0.67*** | 0.12 | 0.43 to 0.91 | 1.40*** | 0.14 | 1.13 to 1.68 | 1.06*** | 0.40 | 0.28 to 1.85 | 1.40*** | 0.14 | 1.12 to 1.68 |
| Phone call with a person | 1.18*** | 0.13 | 0.93 to 1.43 | 1.50*** | 0.13 | 1.23 to 1.76 | 0.90** | 0.40 | 0.11 to 1.69 | 1.45*** | 0.13 | 1.19 to 1.71 |
| Automated text message | 2.42*** | 0.18 | 2.07 to 2.78 | 1.53*** | 0.15 | 1.24 to 1.83 | 2.72*** | 0.48 | 1.78 to 3.66 | 1.52*** | 0.15 | 1.23 to 1.81 |
| Timing | ||||||||||||
| 1 month before the appointment | Reference | |||||||||||
| 1 week before the appointment | 0.87*** | 0.09 | 0.69 to 1.05 | 0.76*** | 0.09 | 0.58 to 0.93 | 1.03*** | 0.28 | 0.48 to 1.57 | 0.75*** | 0.09 | 0.57 to 0.93 |
| Less than 3 days before the appointment | 0.99*** | 0.11 | 0.77 to 1.20 | 1.04*** | 0.10 | 0.85 to 1.24 | 0.81** | 0.32 | 0.19 to 1.43 | 1.05*** | 0.10 | 0.85 to 1.25 |
| Content | ||||||||||||
| Basic details | Reference | |||||||||||
| Basic details + clinician name | 0.12 | 0.08 | − 0.03 to 0.27 | 0.34** | 0.14 | 0.07 to 0.60 | − 0.03 | 0.26 | − 0.55 to 0.48 | 0.34** | 0.14 | 0.07 to 0.61 |
| Basic details + appointment cost | 0.13* | 0.07 | − 0.02 to 0.28 | 0.60*** | 0.11 | 0.39 to 0.82 | 0.52** | 0.26 | 0.01 to 1.04 | 0.60*** | 0.11 | 0.39 to 0.82 |
| Interactivity | ||||||||||||
| No ability to cancel or modify the appointment | Reference | |||||||||||
| Ability to cancel or modify the appointment | 1.36*** | 0.11 | 1.14 to 1.57 | 1.48*** | 0.09 | 1.30 to 1.66 | 1.28*** | 0.34 | 0.62 to 1.95 | 1.50*** | 0.09 | 1.31 to 1.68 |
| Systematic preference heterogeneity | ||||||||||||
| Age × phone call with a person | – | – | – | – | – | – | − 0.015* | 0.008 | − 0.03 to 0.0002 | – | – | – |
| Age × automated/robot phone call | – | – | – | – | – | – | − 0.011 | 0.008 | − 0.027 to 0.005 | – | – | – |
| Age × e-mail | – | – | – | – | – | – | 0.001 | 0.008 | − 0.015 to 0.016 | – | – | – |
| Age × automated text message | – | – | – | – | – | – | − 0.010 | 0.009 | − 0.028 to 0.009 | – | – | – |
| Age × 1 week before appointment | – | – | – | – | – | – | 0.877D–04 | 0.006 | − 0.107D–01 to 0.109D–01 | – | – | – |
| Age × less than 3 days before appointment | – | – | – | – | – | – | 0.003 | 0.006 | − 0.01 to 0.015 | – | – | – |
| Age × basic details + clinician name | – | – | – | – | – | – | –0.0004 | 0.005 | − 0.011 to 0.01 | – | – | – |
| Age × basic details + appointment cost | – | – | – | – | – | – | − 0.009* | 0.005 | − 0.02 to 0.001 | – | – | – |
| Age × ability to cancel or modify the appointment | – | – | – | – | – | – | 0.004 | 0.007 | − 0.009 to 0.018 | – | – | – |
| Male × phone call with a person | – | – | – | – | – | – | − 0.027 | 0.212 | − 0.442 to 0.387 | – | – | – |
| Male × automated robot phone call | – | – | – | – | – | – | − 0.024 | 0.231 | − 0.477 to 0.428 | – | – | – |
| Male × e-mail | – | – | – | – | – | – | 0.329 | 0.223 | − 0.109 to 0.767 | – | – | – |
| Male × automated text message | – | – | – | – | – | – | 0.386 | 0.260 | − 0.124 to 0.896 | – | – | – |
| Male × 1 week before appointment | – | – | – | – | – | – | − 0.267* | 0.153 | − 0.567 to 0.033 | – | – | – |
| Male × less than 3 days before appointment | – | – | – | – | – | – | 0.018 | 0.180 | − 0.334 to 0.37 | – | – | – |
| Male × basic details + clinician name | – | – | – | – | – | – | 0.223 | 0.148 | − 0.066 to 0.512 | – | – | – |
| Male × basic details + appointment cost | – | – | – | – | – | – | 0.198 | 0.149 | − 0.095 to 0.491 | – | – | – |
| Male × ability to cancel or modify the appointment | – | – | – | – | – | – | − 0.215 | 0.193 | − 0.594 to 0.164 | – | – | – |
| Missed × phone call with a person | – | – | – | – | – | – | − 0.570** | 0.247 | − 1.055 to − 0.086 | – | – | – |
| Missed × automated/robot phone call | – | – | – | – | – | – | 0.301 | 0.258 | − 0.205 to 0.807 | – | – | – |
| Missed × e-mail | – | – | – | – | – | – | 0.316 | 0.252 | − 0.178 to 0.809 | – | – | – |
| Missed × automated text message | – | – | – | – | – | – | − 0.204 | 0.293 | − 0.778 to 0.37 | – | – | – |
| Missed × 1 week before appointment | – | – | – | – | – | – | −0.073 | 0.172 | − 0.41 to 0.263 | – | – | – |
| Missed × less than 3 days before appointment | – | – | – | – | – | – | 0.199 | 0.199 | − 0.192 to 0.589 | – | – | – |
| Missed × basic details + clinician name | – | – | – | – | – | – | 0.183 | 0.166 | − 0.143 to 0.509 | – | – | – |
| Missed × basic details + appointment cost | – | – | – | – | – | – | − 0.261 | 0.170 | − 0.594 to 0.072 | – | – | – |
| Missed × ability to cancel or modify the appointment | – | – | – | – | – | – | − 0.020 | 0.209 | − 0.43 to 0.391 | – | – | – |
| Model fit | ||||||||||||
| LL | − 3434.12 | − 3414.88 | ||||||||||
| AIC/N | 1.69 | 1.69 | ||||||||||
AIC Akaike information criterion, ASC alternative specific constant, CI confidence interval, COEFF coefficient, LL log likelihood, N number of respondents, SD standard deviation, SE standard error, *significant at the 10% level; **significant at the 5% level; ***significant at the 1% level
Dummy coding was used for all attributes
For values indicated as nnnnn.D–xx or D+xx => multiply by 10 to − xx or + xx