### Description

Counsell et al developed prognostic scores to predict survival and outcome after a stroke. This can help identify patients who may require more aggressive management. The authors are from Western General Hospital (Edinburgh) and the University of Aberdeen in Scotland.

The patients were examined by a neurologist as soon as possible after the stroke, with median delay of 4 days).

Models:

(1) Cox proportion hazard model for survival at 30 days

(2) logistic regression model for alive and independent at 6 months

Parameters:

(1) age of stroke onset to 2 decimal points (I will use date of stroke minus birth date, divided by 365.25, to take into account leap years)

(2) living alone (no one permanently living with the person before the stroke)

(3) independent before stroke (Oxford Handicap Score <= 2, indicating no or minor handicap, before stroke)

(4) verbal component to Glasgow Coma Score (GCS verbal score = 5, or oriented)

(5) ability to lift both arms (able to lift both arms to horizontal, corresponding to MRC score >= 3)

(6) ability to walk (walks without someone's help; can use a walking stick or frame)

 Parameter Finding Points age in years age to 2 decimals points (age) living alone no 2 yes 1 independent before no 2 yes 1 oriented no 2 yes 1 able to lift both arms no 2 yes 1 able to walk without personal assistance no 2 yes 1

from Table 3, page 1045

The reference person for the Cox proportional hazard model is a 70 year old with all findings "No" (points = 2).

Y for Cox model=

= (0.034 * ((age in years) – 70)) – (0.406 * ((points for living alone) – 2)) + (0.501 * ((points for independent) – 2)) + (0.776 * ((points for oriented) – 2)) + (0.851 * ((points for arm lifting) – 2)) + (0.489 * ((points for ability to walk) – 2))

probability of patient survival at 30 days =

= ((survival for reference person) ^ EXP(Y))

where:

• Baseline survival for reference person = 0.631

X for logistic regression model =

= 12.340 – (0.051 * (age in years)) + (0.661 * (points for living alone)) - (2.744 * (points for independence before)) - (2.160 * (points for verbal GCS)) - (2.106 * (points for lifting arms)) - (1.311 * (points for walking))

probability alive and independent at 6 months =

= 1 / (1 + EXP((-1) * X))