Can procalcitonin guide decisions about antibiotic management?
BACTEREMIA
Observational studies from as far as back as 1999 have examined the association of procalcitonin levels with bacteremia. The study designs were generally similar, with procalcitonin levels checked at time of blood culture, mostly in emergency rooms, and the procalcitonin value correlated with blood culture results. The general conclusion has been that procalcitonin has diagnostic value in ruling out bacteremia but should be used in the context of pretest probability rather than in isolation.
Hattori et al8 performed one of the largest studies, in 1,331 patients, using a procalcitonin level cutoff of 0.9 ng/mL. The sensitivity was 72% and specificity was 69%, which are not impressive; however, the negative predictive value was 95%, and even higher at lower cutoff values. Further, procalcitonin was significantly better at predicting bacteremia than either the white blood cell count or C-reactive protein level, with the latter two being hardly better than random chance.
Hoeboer et al9 performed a meta-analysis of various studies with a total of 16,514 patients. Using a cutoff of 0.5 ng/mL, they reported a sensitivity of 76% and a specificity of 69% with a negative predictive value of 97% in emergency rooms, 95% on regular wards, and 98% in ICUs. The high negative predictive value of procalcitonin can allow clinicians to stratify bacteremia risk to determine which patients need blood cultures, which in turn may help clinicians order blood cultures more appropriately and avoid unnecessary costs, delays, and harms associated with false-positive results, such as additional visits, additional testing, and unnecessary use of antibiotics.
MENINGITIS
As with bacteremia, observational studies have reported fairly high negative predictive values for procalcitonin in bacterial meningitis. The correlation is not surprising, given that most cases of bacterial meningitis occur due to hematogenous dissemination.
A 2015 meta-analysis of 9 studies and 725 patients reported a pooled sensitivity of 90%, specificity 90%, positive likelihood ratio 27.3, and negative likelihood ratio 0.13.10 Cutoffs for procalcitonin levels varied, but the most common value was 0.5 ng/mL. The authors also noted that the diagnostic utility of procalcitonin was far superior to C-reactive protein in this scenario, concluding that serum procalcitonin is a highly accurate test to distinguish between bacterial and viral causes in suspected meningitis.10
OTHER CLINICAL APPLICATIONS
Postoperative infection
Small studies have assessed procalcitonin as a marker to rule out postoperative infections,11,12 but the heterogeneity of study designs and populations makes it difficult to combine the studies for meta-analysis. Nevertheless, the general trend is that there may be a role for procalcitonin, and that procalcitonin has better diagnostic yield than the white blood cell count or C-reactive protein level. The optimal cutoff depends on the surgery, since a small elevation in procalcitonin can be expected with the stress of surgery; and since the degree of elevation varies with type of surgery, the result must be interpreted with caution.
Malignancy
In malignancy-associated conditions such as neutropenic fever and tumor fever, the clinical utility of procalcitonin is somewhat diminished, as malignancy can cause elevated procalcitonin levels (especially in metastatic disease), but a low concentration still has a fair negative predictive value (approximately 90%) for bloodstream infections.13
A retrospective study suggested that the ratio of procalcitonin to C-reactive protein could improve diagnostic accuracy in patients with malignancies, presumably because an elevation of procalcitonin out of proportion to elevation in C-reactive protein favored a bacterial infection rather than nonspecific inflammation related to malignancy.14
Cardiac syndromes
In cardiac syndromes, dyspnea and abnormal chest imaging may make it difficult to exclude respiratory infections. Schuetz et al15 reviewed the potential value of procalcitonin testing in a variety of cardiac disorders, especially in acute cardiovascular conditions whose presentation resembles that of sepsis or acute respiratory tract infection. They concluded it may have a role in diagnosis and prognosis in these settings, as well as guiding drug therapy.
Localized infections
Though localized infections such as cystitis, cellulitis, and osteomyelitis often do not affect procalcitonin levels, the test may help assess illness severity and rule out associated bacteremia.
One study found that a low procalcitonin level was insufficient to rule out urinary tract infection, but procalcitonin levels predicted bacteremia better than any other variable or combination of variables; moreover, procalcitonin had a negative predictive value as high as 97% for ruling out bacteremia associated with urinary tract infection.16
