An evaluation of the discriminant and predictive validity of relative social disadvantage as screening criteria for priority access to public general dental care, in Australia

Table 4 Unconditional regression model for those requiring treatment < 6 months using 1 + RSD and logistic regression model for those requiring treatment < 6 months using self-reported oral condition screening criteria and 1 + RSD

Criteria	Beta coefficient	SE	Sig z	OR
Model 1
1 + RSD	1.35	0.19	0.000	3.87 (2.65, 5.65)
I have sore gums	−0.113	0.529	0.848	0.89 (0.27, 2.85)
I need a scale and clean	−1.156	0.086	0.000	0.31 (0.18, 0.53)
I have mouth ulcers	−0.013	1.067	0.990	0.98 (0.11,8.22)
I need gum treatment	0.412	1.182	0.598	1.51 (0.32,7.00)
I have a broken denture	1.001	1.540	0.077	2.72 (0.89,8.25)
My denture needs to be fixed	−0.121	0.257	0.676	0.88 (0.50,1.56)
I have Halitosis (bad breath)	−0.119	0.944	0.911	0.88 (0.11,7.14)
Model 2
1 + RSD	1.302	0.827	0.000	3.67 (2.36,5.71)
I have a broken filling	−0.234	0.217	0.393	0.79 (0.46,1.35)
I need a filling	0.141	0.276	0.555	1.15 (0.72,1.84)
I have bleeding gums	−0.495	0.546	0.581	0.60 (0.10,3.53)
I have a loose tooth	0.291	0.989	0.693	1.33 (0.31,5.69)
I have a broken tooth	−0.049	0.270	0.862	0.95 (0.54, 1.66)
I need an extraction	1.370	1.854	0.004	3.93 (1.56, 9.91)
I have a chipped tooth	−0.263	0.379	0.593	0.76 (0.29,2.02)

Sensitivity 81.5%, Specificity 63.4%, Positive Predictive Value 87.2%.
Negative Predictive Value 52.8% (95% CI,) Cut off set at 0.7.
Criteria of having a lost denture, a clicking jaw and a perceived need for a crown and bridge have been dropped from the model due to colllinearity.

ISSN: 1472-6963