Geeson et al reported the MOAT model (Medicines Optimisation Assessment Model) for predicting medication-related problems for a patient. This can alert the pharmacist or other clinician to a patient at increased risk. The authors are from Luton and Dunstable NHS Foundation Trust, UCL School of Pharmacy and Imperial College Healthcare NHS Trust.
Parameters:
(1) number of comorbidities
(2) eGFR in mL per min per 1.73 square meter
(3) white blood cell count in 10^9/L
(4) number of medicines
(5) history of previous allergy
(6) nervous system and/or mental disorder
(7) respiratory system disorder
(8) gastrointestinal problem
(9) therapy with aminoglycosides and glycopeptides
(10) therapy with other systemic antimicrobials
(11) epilepsy medicines
Parameter
|
Finding
|
Points
|
number of comorbidities
|
|
0.125 * (number)
|
eGFR
|
|
-0.0308 * (eGFR)
|
WBC
|
|
0.0234 * (WBC)
|
number of medicines
|
|
0.0347 * (number)
|
history of allergy
|
no
|
0
|
|
yes
|
0.272
|
nervous or mental disorder
|
no
|
0
|
|
yes
|
0.354
|
respiratory disorder
|
no
|
0
|
|
yes
|
-0.234
|
GI disorder
|
no
|
0
|
|
yes
|
-0.533
|
aminoglycoside or glycopeptide
|
no
|
0
|
|
yes
|
0.331
|
other systemic antimicrobial
|
no
|
0
|
|
yes
|
0.311
|
epilepsy medicines
|
no
|
0
|
|
yes
|
0.385
|
value of X =
= SUM(points for all of the parameters) - 1.674
probability of a medicine related problem =
= 1 / (1 + EXP((-1) * X))
Performance:
• The are under the ROC curve is 0.657.