The Cambridge Diabetes Risk Score estimates the probability (P) that an adult without known diabetes currently has undiagnosed type 2 diabetes using a logistic model.

First calculate the linear predictor logit(P):

logit(P) = -6.322 − 0.879×(female) + 0.063×age (years) + 1.222×(on antihypertensive medication) + 2.191×(on systemic steroids) + BMI term + family history term + smoking term.

Predictor	Category	Coefficient added to logit(P)
Sex	Male	0
Sex	Female	−0.879
BMI (kg/m²)	<25.0	0
BMI (kg/m²)	25.0–27.49	0.699
BMI (kg/m²)	27.5–29.99	1.970
BMI (kg/m²)	≥30.0	2.518
Family history	No diabetic parent or sibling	0
Family history	Parent or sibling with diabetes	0.728
Family history	Parent and sibling with diabetes	0.753
Smoking	Non-smoker	0
Smoking	Ex-smoker	−0.218
Smoking	Current smoker	0.855
Medication	Currently prescribed antihypertensive	+1.222
Medication	Currently prescribed systemic steroids	+2.191

The predicted probability is then P = 1 / (1 + e^−logit(P)). The Fellowed implementation also reports P as a percentage.