P. Icard et al., UTILITY OF C-REACTIVE PROTEIN MEASUREMENTS FOR EMPYEMA DIAGNOSIS AFTER PNEUMONECTOMY, The Annals of thoracic surgery, 57(4), 1994, pp. 933-936
Serum C-reactive protein (CRP) levels were studied serially during the
postoperative period in 151 consecutive patients who underwent pneumo
nectomy. Virtually all patients who had a simple postoperative course
(115 of 120), as well as 9 patients who had a bronchial infection of t
he remaining lung, 3 with a pulmonary embolus, and 2 who suffered post
operative bleeding requiring reoperation, demonstrated a similar posto
perative evolution in their CRP values: a rapid postoperative rise unt
il a peak or a plateau (mean peak value, 132 +/- 25 mg/L) was reached
within 3 to 6 days, followed by a progressive decline to a value of le
ss than 75 mg/L on day 9, and less than 50 mg/L on day 12. Conversely,
all 12 patients who suffered empyema postoperatively, as well as 3 pa
tients with bacterial pneumonia, 1 patient with chylothorax, and 1 pat
ient with inflammatory pericarditis, demonstrated either a markedly pe
rsistent elevation in their CRP values or a secondary rise in the leve
ls which exceeded 100 mg/L. Because of the high sensitivity (100%) and
specificity (91.4%) of the CRP levels in detecting postpneumonectomy
empyema, we recommend the routine use of this measure. Furthermore, a
low CRP value after pneumonectomy (less than 50 mg/L) may help in deci
ding whether to confidently discharge a patient from the hospital in t
he absence of empyema. The negative predictive value of this method wa
s found to be 100%.