Simon et al. (2004; 39:206–17)

In an article in the 15 July 2004 issue of the journal (Simon L, Gauvin F, Amre DK, Saint-Louis P, Lacroix J. Serum procalcitonin and C-reactive protein levels as markers of bacterial infection: a systematic review and meta-analysis. Clin Infect Dis 2004; 39:206–17), several errors were made in the data presented, requiring corrections to the text, the abstract, tables 4 and 5, and figures 1 and 2.

The data in the sixth and seventh paragraphs of the Results section require a number of corrections and adjustments. In the sixth paragraph, the estimations of sensitivity and specificity for bacterial infections compared with noninfective causes of inflammation require the following corrections. Pooled sensitivity for procalcitonin (PCT) was 85% (95% CI, 0.76–0.91), compared with 78% (95% CI, 0.70–0.85) for C-reactive protein (CRP); there was a statistically significant difference between the sensitivities: 7% (95% CI, 0.23–0.11; P < .05). Pooled specificity for PCT was also higher than for CRP: 83% (95% CI, 0.68–0.92) versus 60% (95% CI, 0.38–0.79), respectively, and this difference was statistically significant: 23% (95% CI, 0.18–0.28; P < .05). This was confirmed on calculation of the Q value, which was higher for PCT (Q = 0.82; 95% CI, 0.64–0.95) than that for CRP (Q = 0.75; 95% CI, 0.68–0.82).

In the seventh paragraph of Results, the data for evaluation of these markers for bacterial infections versus viral infections requires the following corrections. Pooled sensitivity for PCT was 82% (95% CI, 0.65–0.92), compared with 73% (95% CI, 0.62–0.82) for CRP, and the difference is statistically significant: 9% (95% CI, 0.02–0.16; P < .05). Pooled specificities were also higher for PCT than for CRP: 88% (95% CI, 0.50–0.98) versus 81% (95% CI, 0.55–0.93), respectively, with a difference of 7% (95% CI, -0.01 to 0.13; P > .05). The Q value calculated from the curves was slightly higher for PCT (Q = 0.85; 95% CI, 0.83–0.87) than for CRP (Q = 0.82; 95% CI, 0.74–0.90), suggesting that, in terms of overall accuracy, PCT is somewhat better than CRP in differentiating between bacterial and viral infections.

In addition, corrections similar to those made in the text are required in the abstract. The corrected abstract should read as follows:

A meta-analysis was performed to evaluate the accuracy of determination of procalcitonin (PCT) and C-reactive protein (CRP) levels for the diagnosis of bacterial infection. The analysis included published studies that evaluated these markers for the diagnosis of bacterial infections in hospitalized patients. PCT level was more sensitive (85% [95% confidence interval {CI}, 0.76–0.91] vs. 78% [95% CI, 0.70–0.85]) and more specific than CRP level (83% [95% CI, 0.68–0.92] vs. 60% [95% CI, 0.38–0.79]) for differentiating bacterial from noninfective causes of inflammation. The Q value for PCT markers was higher (0.82 vs 0.75). The sensitivity for differentiating bacterial from viral infections was also higher for PCT markers (82% [95% CI, 0.65–0.92] vs. 73% [95% CI, 0.62–0.82]), as was the specificity (88% [95% CI, 0.50–0.98] vs. 81% [95% CI, 0.55–0.93]). The Q value was higher for PCT markers (0.85 vs. 0.82). PCT markers also had a higher positive likelihood ratio and lower negative likelihood ratio than did CRP markers in both groups. On the basis of this analysis, the diagnostic accuracy of PCT markers was higher than that of CRP markers among patients hospitalized for suspected bacterial infections.

In tables 4 and 5, the false positive (FP) and the false negative (FN) values were inadvertently interchanged. The resultant changes in the estimation of the sensitivities and specificities (along with their 95% CIs) are presented here in corrected tables (tables 4 and 5).

Figure 1

Summary receiver operating characteristic (SROC) curves comparing serum procalcitonin (PCT; •) and C-reactive protein (CRP; +) markers for detection of bacterial infections versus noninfective causes of inflammation. Each point contributing to the SROC curve represents 1 study.

Figure 2

Summary receiver operating characteristic (SROC) curves comparing serum procalcitonin (PCT; •) and C-reactive protein (CRP; +) markers for detection of bacterial infections versus viral infection. Each point contributing to the SROC curve represents 1 study.

Table 4

Results derived from the 2 × 2 tables of individual studies involving procalcitonin and C-reactive protein levels as markers for bacterial infections versus noninfective causes of inflammation.

Table 5

Results derived from the 2 × 2 tables of individual studies involving procalcitonin and C-reactive protein levels as markers for bacterial infections versus viral infections.

As a result of these changes, figures 1 and 2 need to be slightly modified, as well. The corrected figures are presented here (figures 1 and 2).

These corrections do not alter the interpretation or the conclusions of the meta-analysis. The authors sincerely apologize for these errors.

• © 2005 by the Infectious Diseases Society of America