Tuesday, August 10, 2010

Doc....do I need to be admitted to ICU for this??

We had a a 48-year-old woman presented to the ER with a 48-hour history of anorexia, fever, chills, and cough and left pleuritic chest pain. On examination she looked acutely ill and seemed to be confused. Her temp was 102.2°F; HR110; RR 30; and BP 90/60 mm Hg. Sats was 89% with an infiltrate on chest xray. So the diagnosis is Community acquired pneumonia.

Can she go home? or needs admission to the floor? or needs ICU monitoring?

2 commonly used prognostic scoring systems for Community acquired pneumonia (CAP) are the Pneumonia severity Index(PSI) and the CURB 65.These were initially developed to predict mortality, but later have been used to triage patients with CAP. while PSI identifies patients with low risk of death, CURB identifies high mortality risk pts.

PSI - includes 20 variables. was developed from a cohort of nearly 140000 pts.(PORT cohort)has been validated in variety of pt. population. It divides pts into 5 classes, with class 1-3 having low mortality( 0.1 to 2.8%). class 4 & 5 had higher mortality. This does not translate into site of care decision making as nearly 27% of class 1-3 pts may end up in ICU(Eur Resp 2001).Also it doesn't take immunosupressed pts, COPD, pts with SOB etc.(e.g -a 30 yo HIV with SOB and smoking history would not be a high risk in PSI) So in triaging pts, PSI can't replace a physician's judgement, but can help to make that decision.

CURB 65 - developed and used widely in Europe,this is easy to use in ER, with scores >3 having a >20% 30-day mortality. Again,this has been translated into triage decision making, with scores 0-1 being managed as outpatients. Eventhough CURB predicted hospitalisation well in some studies(Capelastegui et al.), it can't replace clinical judgement for site of care.

So both these scores are not designed for making triage decisions,but can help clinician by providing an idea about mortality. They also are not very reliable in the elderly population.

Once decided to admit....then the need for ICU monitoring has to be determined, and again these 2 scores are not designed to do that. So ATS & IDSA have come up with a criteria ,which has sensitivity & specificity of 71% & 88% resp. for predicting ICU admission.
www.idsociety.org/WorkArea/DownloadAsset.aspx?id=9974. This is a pocket guide.

Used appropriately, these scores will help care for the needy, and at the same time save unnecessary healthcare spendng!!


  1. thanks for sharing!- Your intern (NG)