Description

Nierman et al developed an outcome prediction model for very old (>= 85 years of age) patients admitted to the intensive care unit (ICU). This can help separate patients who may be able to return home from those with a more guarded prognosis. The authors are from Mount Sinai Medical Center in New York City.


 

Patient selection: age >= 85 years of age

 

Parameters:

(1) age

(2) gender

(3) independence and assistance with activities of daily living (ADL)

(4) type of intensive care unit (ICU)

(5) heart rate at admission

(6) number of organ system failures

(7) count of ICU procedures

Parameter

Finding

Points

gender

female

0

 

male

1

independent in ADL

yes

1

 

no

0

requires assistance in ADL

yes

1

 

no

0

type of ICU

CCU (cardiac)

0

 

CSICU (cardiac surgery)

0

 

MICU (medical)

1

 

SICU (surgical)

1

 

NSICU (neurosurgical)

1

 

where:

• I assume that a person is either independent or requires assistance in ADL (a yes for one implies the other is no).

 

Organ Failures

Finding

Points

respiratory

yes (see below)

1

 

no

0

hematologic

WBC count < 2,000 per µL or platelet count < 20,000 per µL

1

 

WBC count >= 2,000 per µL and platelet count >= 20,000 per µL

0

neurologic

Glasgow coma score < 10 when not sedated

1

 

Glasgow coma score >= 10

0

sepsis

yes (see below)

1

 

no

0

 

where:

• Respiratory failure is indicated by 1 or more of the following: PaO2 < 50 mm Hg; PaCO2 > 45 mm Hg with an acute respiratory acidosis (pH < 7.37); FIO2 > 0.40 while on mechanical ventilation.

• Sepsis = systolic blood pressure < 90 mm Hg with one or more of the following: positive blood culture, localized abscess proven by surgery or on radionuclide scan, or rectal temperature > 39°C.

 

number of organ failures =

= SUM(points for organ failures present)

 

ICU procedures

Points

pressors

1

mechanical ventilation

1

pulmonary artery catheter

-1

 

count for ICU procedures =

= SUM(points for ICU procedures)

 

index =

= (0.035 * (age in years)) + (0.037 * (points for gender)) + (1.324 * (points for ADL independent)) + (1.021 * (points for ADL assistance)) - (0.821 * (points for ICU)) - (0.021 * (heart rate in beats per minute)) - (0.596 * (count for ICU procedures)) - (0.696 * (number of organ system failures))

 

Outcome

Constant

death vs (skilled nursing facility or home)

-0.872

(death or skilled nursing facility) vs home

0.805

 

probability of death =

= 1 / (1 + EXP((index + 0.872))

 

probability of discharge home =

= 1 / (1 + EXP((-1) * ((index) – 0.805)))

 

probability of discharge to a skilled nursing facility =

= 1 – (probability of death) – (probability of discharge home)

 

where:

• In the example given on page 1857, the patient's index is given as 0.075. If the data is entered and calculated, the value is –0.075.

 


To read more or access our algorithms and calculators, please log in or register.