Description

Halm et al studied the risk of an adverse event occurring after discharge from the hospital following admission for a community-acquired pneumonia. The number of unstable factors present in the 24 hours before discharge correlated with disability, readmission and death. The authors are from 5 university hospitals in the US and Canada.


 

Patient evaluation: during the 24 hours prior to discharge

 

Inclusion and exclusion criteria:

(1) >= 18 years of age

(2) clinical and radiologic evidence of acute pneumonia

(3) not HIV-positive

(4) not previously admitted within the past 10 days

 

Parameters associated with clinical stability:

(1) temperature

(2) heart rate

(3) respiratory rate

(4) systolic blood pressure

(5) oxygen saturation

(6) oral intake

(7) mental status

Parameter

Finding

Points

temperature

<= 37.8°

0

 

> 37.8°C

1

heart rate

<= 100 beats per minute

0

 

> 100 breaths per minute

1

respiratory rate

<= 24 breaths per minute

0

 

> 24 breaths per minute

1

systolic blood pressure

>= 90 mm Hg

0

 

< 90 mm Hg

1

oxygen saturation

>= 90%

0

 

< 90%

1

ability to maintain oral intake

able

0

 

unable

1

mental status

normal

0

 

abnormal

1

 

number of unstable factors =

= SUM(points for all 7 parameters)

 

Interpretation:

• minimum number of unstable factors: 0

• maximum number of unstable factors: 7

• The greater the number of unstable factors, the greater risk of an adverse event following discharge from the hospital.

• The presence of 1 unstable factor raised the risk of an adverse event slightly (adjusted odds ratio 1.1 for death to 1.5 for disability at 30 days).

 

Adverse Event

Adjusted Odds Ratio when >= 2 Unstable Findings Present

p value

death

14.1

< 0.001

readmission

3.5

0.02

unable to return to usual activities within 30 days of discharge

1.6

0.007

from Table 3, page 1282

 

Performance in predicting adverse events:

• The negative predictive value was 89%, while the positive predictive value was only 17% for >= 1 factor and 46% for >= 2 factors.

• Specificity was 82% for >= 1 unstable factor and 99% for >= 2 factors.

• Sensitivity was only 27.5% for >= 1 unstable factor and 7.5% for >= 2 factors.

 

Limitations:

• The poor sensitivity and positive predictive value indicates that factors other than patient stability are important influences for an adverse outcome.

 


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