Description

Garcia-Ordonez et al developed a discriminant rule for predicting mortality in an elderly patient admitted to the hospital with community acquired pneumonia. This can help identify patients who may benefit from more aggressive management and closer monitoring. The authors are from the Complejo Hospitalario Universitario "Carlos Haya" in Malaga, Spain.


 

Patient selection: >= 65 years of age with community-acquired pneumonia (presence of new pulmonary infiltrate, signs and symptoms for a lower respiratory tract infection, acquisition outside of a hospital or a health care facility).

 

Exclusion criteria:

(1) HIV-positive

(2) hospitalized in the 2 weeks prior to admission

(3) admitted from a nursing home or chronic care facility

(4) had another disease that could explain the radiographic findings

 

Prognostic factors associated with mortality identified by multivariate analysis:

(1) bilateral radiographic infiltrate

(2) tachypnea

(3) elevated urea (serum urea > 7 mmol/L)

(4) absence of fever (temperature < 37.8°C)

(5) confusion

(6) shock (systolic blood pressure < 90 mm Hg, urinary excretion < 20 mL per hour for 4 hours or the need for vasopressors)

 

total number of prognostic factors =

= SUM(number of factors present)

 

Interpretation:

• minimum number of prognostic factors: 0

• maximum number of prognostic factors: 6

• The prognostic rule identified patients as high risk for mortality with a total score >= 3.

 

Performance of rule:

• Sensitivity 61%, specificity 91%.

• Positive predictive value 54%, negative predictive value 93%, overall accuracy 87%.

• 39% of the patients who died had a score < 3.

• Some additional risk factors that might be worth considering are hypothermia and previous immunizations.

 


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