Table 5.
Coefficients for preferences and class membership for 4 groups from latent class logit modelling (data pooled from patients in DCE1 and DCE2)
| Latent Class | Class 1: situation/living well | Class 2: information | Class 3: cost dominant | Class 4: inconsistent | ||||
| Coefficient | 95% CI | Coefficient | 95% CI | Coefficient | 95% CI | Coefficient | 95% CI | |
| Preferences* | ||||||||
| Information | 0.80 | (0.72 to 0.88) | 1.95 | (1.78 to 2.13) | 0.24 | (0.14 to 0.35) | 0.01 | (−0.07 to 0.10) |
| Situation | 2.08 | (1.95 to 2.21) | 0.98 | (0.84 to 1.12) | 0.31 | (0.18 to 0.44) | 0.18 | (0.08 to 0.28) |
| Living well | 1.82 | (1.70 to 1.94) | 0.78 | (0.65 to 0.90) | 0.44 | (0.32 to 0.55) | 0.25 | (0.16 to 0.33) |
| Communication | 0.45 | (0.37 to 0.52) | 0.23 | (0.10 to 0.36) | 0.11 | (0.01 to 0.21) | 0.11 | (0.03 to 0.20) |
| Cost | −0.06 | (−0.07 to −0.06) | −0.05 | (−0.06 to −0.04) | −0.25 | (−0.27 to −0.23) | 0.02 | (0.01 to 0.02) |
| Class membership† | ||||||||
| Constant | – | – | −0.77 | (−1.71 to 0.17) | 1.00 | (0.15 to 1.85) | 0.45 | (−0.38 to 1.27) |
| DCE2‡ | – | – | −0.21 | (−0.75 to 0.32) | −0.02 | (−0.53 to 0.49) | 0.07 | (−0.40 to 0.54) |
| Relationship: Single§ | – | – | −0.16 | (−0.69 to 0.36) | −0.37 | (−0.89 to 0.15) | −0.10 | (−0.58 to 0.39) |
| Education: University¶ | – | – | 0.20 | (−0.29 to 0.69) | −0.05 | (−0.54 to 0.43) | 0.01 | (−0.44 to 0.46) |
| Job status: Not working** | – | – | −0.18 | (−0.97 to 0.62) | −0.08 | (−0.82 to 0.67) | −0.30 | (−1.02 to 0.41) |
| Job status: Retired** | – | – | −0.15 | (−0.72 to 0.43) | −0.59 | (−1.17 to −0.01) | −0.81 | (−1.36 to −0.27) |
| Job status: Disabled** | – | – | 0.31 | (−0.67 to 1.29) | 0.08 | (−0.82 to 0.99) | −0.15 | (−1.00 to 0.70) |
| Perceived health: Fair†† | – | – | −0.05 | (−0.59 to 0.49) | −0.44 | (−0.96 to 0.08) | −0.09 | (−0.59 to 0.41) |
| Perceived health: Poor†† | – | – | −0.58 | (−1.33 to 0.16) | −1.06 | (−1.78 to −0.34) | −0.45 | (−1.11 to 0.21) |
| Gender: Female‡‡ | – | – | −0.10 | (−0.61 to 0.42) | −0.82 | (−1.31 to −0.33) | −0.62 | (−1.08 to −0.16) |
| Income: £15 600-£31 1999 | – | – | 0.24 | (−0.46 to 0.95) | −0.61 | (−1.22 to 0.01) | −0.10 | (−0.70 to 0.50) |
| Income: £31200+§§ | – | – | 0.40 | (−0.34 to 1.15) | −1.15 | (−1.88 to −0.43) | −0.12 | (−0.78 to 0.53) |
| Income: Not prepared to say§§ | 0.93 | (0.08 to 1.78) | −0.19 | (−1.02 to 0.64) | −0.13 | (−1.03 to 0.78) | ||
| Class share | N | % | N | % | N | % | N | % |
| Predicted membership | 286 | 39.9% | 137 | 19.1% | 126 | 17.6% | 168 | 23.4% |
| Data quality* | ||||||||
| Serial Non-participation (n=12) | 0 | – | 0 | – | 0 | – | 12 | 100% |
| Dominance (n=84) | 11 | 13.1% | 6 | 7.1% | 2 | 2.4% | 65 | 77.4% |
| Response time (n=107) | 4 | 3.7% | 9 | 8.4% | 46 | 43.0% | 48 | 44.9% |
*In the Preferences and Data quality sections, only the dominant values are highlighted in bold.
† In the Class membership section, statistically significant coefficients are highlighted in bold.
‡Compared with DCE1.
§Compared with married/co-habiting.
¶Compared with no university education.
**Compared with working.
††‘Bad’ or ‘very bad’ self-rated health compared with ‘good’ or ‘very good’.
‡‡Compared with male.
§§Compared with <£15 600 per annum.