Orthopedics

Review Article 

A Systematic Review and Meta-Analysis of Procalcitonin as a Marker of Postoperative Orthopedic Infections

Ross B. Ingber, BS; Abduljabbar Alhammoud, MD; Daniel P. Murray, BA; Roby Abraham, MD; Anant Dixit, MD; Qais Naziri, MD, MBA; Ghalib Ahmed, MD; Carl B. Paulino, MD; William P. Urban, MD; Chad Craig, MD; Aditya V. Maheshwari, MD; Bassel G. Diebo, MD

Abstract

Procalcitonin is a serologic marker that increases in response to inflammatory stimuli, especially those of bacterial origin. Postoperative orthopedic periprosthetic infections are often difficult to diagnose. This study systematically reviewed the literature to evaluate the statistical measures of performance of procalcitonin as a marker of postoperative orthopedic infection. This study showed that procalcitonin has a weighted pooled sensitivity of 67.3%, specificity of 69.4%, positive likelihood ratio of 1.778, negative likelihood ratio of 0.423, and diagnostic odds ratio of 5.770. These results illustrate that procalcitonin is an effective serologic marker for postoperative bacterial infections. [Orthopedics. 2018; 41(3):e303–e309.]

Abstract

Procalcitonin is a serologic marker that increases in response to inflammatory stimuli, especially those of bacterial origin. Postoperative orthopedic periprosthetic infections are often difficult to diagnose. This study systematically reviewed the literature to evaluate the statistical measures of performance of procalcitonin as a marker of postoperative orthopedic infection. This study showed that procalcitonin has a weighted pooled sensitivity of 67.3%, specificity of 69.4%, positive likelihood ratio of 1.778, negative likelihood ratio of 0.423, and diagnostic odds ratio of 5.770. These results illustrate that procalcitonin is an effective serologic marker for postoperative bacterial infections. [Orthopedics. 2018; 41(3):e303–e309.]

Postoperative orthopedic infections are relatively common and often difficult to diagnose. Due to the overlapping signs and symptoms of infection and other causes of pain, a prompt and accurate clinical diagnosis is often elusive in this patient population. In general, the differential diagnosis for such patients includes infection, inflammation, instability, malalignment, aseptic loosening, septic loosening, and hypersensitivity to metal or cement.1 Because of associated detrimental outcomes and morbidity, postoperative infections must be diagnosed in a timely manner. However, current serologic markers of infection such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell count have limitations, as they are often elevated in complication-free surgery and may remain elevated for months. A more effective method of diagnosis is joint aspiration and culture; however, this is not without risks, and results may not be available for several days, creating challenging clinical scenarios.2

In the current literature, a rise in procalcitonin (PCT) has been suggested as an effective serologic marker for the early diagnosis of bacterial infection in postoperative patients. Procalcitonin is a precursor molecule to calcitonin and is produced by C cells in the thyroid gland.3 Procalcitonin is encoded by the CALC-1 gene, which is suppressed everywhere in the body except for the thyroid gland in the absence of bacterial infection. The thyroid gland converts PCT to calcitonin. Therefore, in healthy individuals, PCT is not present in the blood. Bacterial infection causes a drastic upregulation of the CALC-1 gene (likely due to the presence of bacterial lipopolysaccharides), releasing large amounts of PCT into the serum.4 However, PCT does not rise in response to aseptic inflammation or viral infection. In viral infection, viruses stimulate the release of interferon-γ, which suppresses the CALC-1 gene. Minimal rises in PCT following traumatic surgery typically resolve within days.

In cardiothoracic, neurosurgery, and other surgical literature, PCT has been shown to be the most effective serologic marker for accurately diagnosing postoperative infection.5,6 In addition to discerning between bacterial and viral infection, use of PCT has been shown to be relatively inexpensive. One study of patients with acute respiratory tract infections suggested that serum PCT tests are cost-effective in a number needed to treat model—with respect to avoiding antibiotic resistance and its associated costs.7 In orthopedics, although traditional markers such as ESR did not show a significant difference between infected and noninfected patients until 12 to 14 days postoperatively,8 little research has been performed regarding a potential diagnostic role of PCT.

The aim of this study was to review the current orthopedic literature for comparisons of PCT values in patients with and without infections following joint surgery (knee, hip, and shoulder). A meta-analysis was performed on studies that contained sensitivity and specificity values, which were used to calculate the negative likelihood ratio (NLR), positive likelihood ratio (PLR), and diagnostic odds ratio (DOR). Studies that did not include statistical parameters of interest were systematically reviewed.

Materials and Methods

A comprehensive literature search was performed to include all relevant studies available through December 2016. PubMed, Embase, and Cochrane databases were searched following the Preferred Reporting Items for Systematic Review and Meta-Analysis guidelines.9 The search string used was “(procalcitonin AND (surgery OR orthopaedic OR orthopedic OR knee OR hip OR spine OR shoulder OR ankle OR hand OR wrist OR foot)).”

The studies included were those determined to have data regarding the use of PCT in detecting postoperative orthopedic infections. Studies that did not compare infected patients with noninfected patients, did not evaluate PCT as a serologic marker, and analyzed patients following non-joint orthopedic surgeries were excluded. The literature search was conducted independently by 2 of the authors (R.B.I., B.G.D.). Any disagreements were resolved by meetings with the investigators.

Data Extraction

All data from the studies were entered into an electronic spreadsheet for further analysis. Data extraction was performed individually by 2 of the authors (R.B.I., B.G.D.). Data regarding sensitivity and specificity of the biomarkers for detecting infection were extracted from the original studies. For completion of the meta-analysis, sensitivities and specificities were used to generate two-by-two tables to calculate the number of true positives, true negatives, false positives, and false negatives. These values were subsequently used to calculate the NLR, PLR, and DOR for the pooled studies.

Statistical Analysis

Sensitivity and specificity values were pooled and presented as a summary forest plot and a summary receiver operating curve. The cutoff point of the index test varied between each study. In studies with multiple cutoff points, the lowest value was selected and analyzed. The PLR, NLR, and DOR, along with their 95% confidence intervals, were calculated for PCT. Statistical heterogeneity across the studies was tested using I2, which reflects the percentage of variance across the studies due to heterogeneity. In case of no statistically significant heterogeneity, pooling was performed according to the fixed effects model. Using meta-regression, the influence on diagnostic accuracy of different sources of variation was evaluated. Statistical analysis was completed using Open Meta-Analyst Software (Brown University, Providence, Rhode Island) and SPSS version 19 software (IBM Corp, Armonk, New York). Publishing bias was not assessed because of the small number of studies included in this meta-analysis.

Results

The search string revealed a total of 762 studies. All studies were screened by title and abstract, yielding a total of 26 articles for full-text review. After complete review of the 26 articles, 15 were excluded based on the exclusion criteria (Figure 1). All articles were cross-referenced to ensure complete inclusion of all relevant studies. The authors included 11 studies in this systematic review and 6 studies in this meta-analysis. Assessment of the studies included in this meta-analysis showed that all 6 studies contained level II evidence, according to North American Spine Society level of evidence guidelines.

Figure illustrating the study selection process.

Figure 1:

Figure illustrating the study selection process.

Sensitivity and Specificity

The weighted random effect pooled sensitivity was 67.3% (95% confidence interval, 35.0%–88.7%; P<.001), and the weighted random effect pooled specificity was 69.4% (95% confidence interval, 37.6%–89.5%; P<.001) (Figures 23). Results of tests for heterogeneity were highly significant for both sensitivity and specificity. The summary receiver operating curve showed PCT to be an accurate test (Figure 4).

Data from each study regarding sensitivity of procalcitonin and the pooled sensitivity of the studies. The pooled sensitivity was 67.3% (95% confidence interval [CI], 35.0%–88.7%; P<.001). Abbreviations: FN, false negative; TP, true positive.

Figure 2:

Data from each study regarding sensitivity of procalcitonin and the pooled sensitivity of the studies. The pooled sensitivity was 67.3% (95% confidence interval [CI], 35.0%–88.7%; P<.001). Abbreviations: FN, false negative; TP, true positive.

Data from each study regarding specificity of procalcitonin and the pooled specificity of the studies. The pooled specificity was 69.4% (95% confidence interval [CI], 37.6%–89.5%; P<.001). Abbreviations: FP, false positive; TN, true negative.

Figure 3:

Data from each study regarding specificity of procalcitonin and the pooled specificity of the studies. The pooled specificity was 69.4% (95% confidence interval [CI], 37.6%–89.5%; P<.001). Abbreviations: FP, false positive; TN, true negative.

Summary receiver operating curve for procalcitonin illustrating the accuracy in diagnosing postoperative infection. As sensitivity increases, specificity decreases; however, the accuracy of procalcitonin remains convincing.

Figure 4:

Summary receiver operating curve for procalcitonin illustrating the accuracy in diagnosing postoperative infection. As sensitivity increases, specificity decreases; however, the accuracy of procalcitonin remains convincing.

Positive Likelihood Ratio and Negative Likelihood Ratio

Patients with infections were found to have a significantly increased likelihood of having an elevated PCT test result, whereas a normal result indicated no infection (PLR, 1.778; 95% confidence interval, 1.184–2.670; P=.005) (NLR, 0.423; 95% confidence interval, 0.267–0.671; P<.001) (Figures 56). The studies were found to have significant heterogeneity (I2=86.89 for NLR and I2=83.01 for PLR) (Table 1).

Data from each study regarding negative likelihood ratio of procalcitonin and the pooled negative likelihood ratio of the studies. The pooled negative likelihood ratio was 0.423 (95% confidence interval [CI], 0.267–0.671; P<.001). Abbreviations: Di−, disease absent; Di+, disease present; FN, false negative; TN, true negative.

Figure 5:

Data from each study regarding negative likelihood ratio of procalcitonin and the pooled negative likelihood ratio of the studies. The pooled negative likelihood ratio was 0.423 (95% confidence interval [CI], 0.267–0.671; P<.001). Abbreviations: Di−, disease absent; Di+, disease present; FN, false negative; TN, true negative.

Data from each study regarding positive likelihood ratio of procalcitonin and the pooled positive likelihood ratio of the studies. The pooled positive likelihood ratio was 1.778 (95% confidence interval [CI], 1.184–2.670; P=.005). Abbreviations: Di−, disease absent; Di+, disease present; FP, false positive; TP, true positive.

Figure 6:

Data from each study regarding positive likelihood ratio of procalcitonin and the pooled positive likelihood ratio of the studies. The pooled positive likelihood ratio was 1.778 (95% confidence interval [CI], 1.184–2.670; P=.005). Abbreviations: Di−, disease absent; Di+, disease present; FP, false positive; TP, true positive.

Pooled Negative Likelihood Ratio, Positive Likelihood Ratio, and Diagnostic Odds Ratio

Table 1:

Pooled Negative Likelihood Ratio, Positive Likelihood Ratio, and Diagnostic Odds Ratio

Diagnostic Odds Ratio

The pooled unweighted DOR of PCT was 5.770 (95% confidence interval, 2.512–13.251; P<.001) (Figure 7). The studies were found to be heterogeneous with (I2=48.21).

Data from each study regarding diagnostic odds ratio of procalcitonin and the pooled diagnostic odds ratio of the studies. The pooled diagnostic odds ratio was 5.770 (95% confidence interval [CI], 2.512–13.251; P<.001). Abbreviations: FN, false negative; FP, false positive; TN, true negative; TP, true positive.

Figure 7:

Data from each study regarding diagnostic odds ratio of procalcitonin and the pooled diagnostic odds ratio of the studies. The pooled diagnostic odds ratio was 5.770 (95% confidence interval [CI], 2.512–13.251; P<.001). Abbreviations: FN, false negative; FP, false positive; TN, true negative; TP, true positive.

Discussion

Overall, serum PCT level performs well as a marker for postoperative bacterial infection based on the results of this meta-analysis. This study suggests that PCT has a weighted pooled sensitivity of 67.3%, specificity of 69.4%, PLR of 1.778, NLR of 0.423, and DOR of 5.770. These performance measures of PCT suggest it may have a role in identifying infections after orthopedic procedures. Notably, the studies performed by Bottner et al10 and Randau et al11 produced outliers; however, the study by Randau et al used a much higher cutoff value as compared with other studies (Table 2). Analyzing the data without these outliers yields an elevated pooled sensitivity compared with the overall pooled sensitivity.

Demographics and Raw Data From the Studies Included in the Meta-analysis

Table 2:

Demographics and Raw Data From the Studies Included in the Meta-analysis

As detailed in Table 3, several other serologic markers have been investigated for their role in diagnosing postoperative orthopedic infection. Notably, from the studies included in this meta-analysis, interleukin-6 and tumor necrosis factor-α show promise in helping to accurately diagnose infections. Additional research is needed to fully evaluate the utility of these serologic markers, along with PCT, in diagnosing postoperative orthopedic infections. To this point, CRP has been the most useful and widely implemented serum marker for screening postoperative orthopedic patients for infection. However, evidence indicates that in the immediate postoperative period, CRP will be significantly elevated for at least 24 to 48 hours in all patients and may remain elevated above preoperative values for approximately 60 days postoperatively.12 C-reactive protein is also widely considered to be a sensitive but nonspecific marker of inflammatory states. In a prospective study, Mok et al13 determined that a CRP level higher than expected in postoperative orthopedic patients had 82% sensitivity but only 48% specificity for detecting infectious complications. Bottner et al10 suggest that PCT can be used as an adjunct to CRP and other serologic markers. They conclude that an elevated CRP suggests infection and that confirmation with an elevated PCT level is specific for periprosthetic infections. Additionally, they suggest that preoperative aspiration of the joint can help accurately diagnose infections. Erythrocyte sedimentation rate is another serum marker commonly used to detect infection in postoperative patients, but it has been shown to be less reliable than CRP. Mok et al13 also investigated the diagnostic value of a postoperative peak in ESR. They detected a postoperative peak in fewer than half of their patients, had a wide variability regarding on what postoperative day those peaks occurred, and noted that creating a curve to predict ESR values for determining “normal” vs “abnormal” values was not possible.13 Similarly, Lee et al8 found that ESR did not show a significant difference between infected and noninfected spine patients until 12 to 14 days postoperatively. Although little data exist on the subject, Jacovides et al14 showed synovial fluid measurements of ESR, CRP, white blood cell count, and leukocyte percentage to be significantly higher in infected joints compared with noninfected joints.

Data Regarding Procalcitonin, White Blood Cell Count, C-Reactive Protein, Interleukin-6, and TNF-α Pooled From the Studies Included in the Meta-analysis

Table 3:

Data Regarding Procalcitonin, White Blood Cell Count, C-Reactive Protein, Interleukin-6, and TNF-α Pooled From the Studies Included in the Meta-analysis

Regarding PCT, research has been performed on its role as an infectious marker among postoperative patients in other specialties. The pharmacokinetics of PCT aid in its utility for detecting infection. Procalcitonin may rise slightly following surgery, peaking on postoperative day 1. However, levels in noninfected patients are typically lower than a theorized cutoff of 0.5 ng/mL, and the short half-life of PCT allows for trending of the value day by day.15 Studies involving cardiac surgery patients have indicated that serum PCT values are a good indicator of postoperative infection. In these studies, there was inconsistency regarding what serum PCT level reflects a positive test value. However, they have shown that patients with confirmed infection following cardiac surgery have significantly higher serum PCT levels overall.3 Additionally, a study by Kassir et al16 showed that mean PCT levels were elevated in sleeve gastrectomy patients with either fistulae formation or infectious complications when compared with uncomplicated postoperative patients. Perrakis et al17 showed similar results among liver transplantation patients, citing both the peak value and the postoperative trend of PCT as valuable diagnostic tools for detecting infectious complications. Furthermore, Laifer et al6 established that among uncomplicated postoperative neurosurgical patients, PCT levels were not elevated compared with preoperative levels. In that study, the same patients exhibited elevated postoperative leukocyte and neutrophil counts and CRP levels compared with preoperative levels. This could indicate that PCT may prove to be more specific for postoperative infection than those other serological markers.6 With such widespread evidence that PCT can be a valuable serological marker in postoperative patients in other fields, evaluation among orthopedic patients is a logical next step.

In the orthopedic literature, a meta-analysis by Shen et al18 showed that serum PCT can be a valuable marker for nonsurgical orthopedic infections, being superior to CRP in certain aspects. In analyzing serum PCT as a marker for osteomyelitis and septic arthritis, serum PCT displayed lower sensitivity than CRP but higher specificity, PLR, NLR, area under the curve, and DOR at a cutoff value of 0.5 ng/mL. Procalcitonin performed even better when using a lower cutoff value of 0.2 to 0.3 ng/mL.18 Additionally, a prospective trial by Maharajan et al19 showed similar findings for serum PCT in nonsurgical patients as a sensitive and specific marker for osteomyelitis and septic arthritis using a ”positive test” cutoff value of 0.4 ng/mL. There have been some limited investigations into the utility of serum PCT as an infection marker in postoperative orthopedic patients that have yielded mixed results.

Aside from the studies included in this meta-analysis, an investigation performed by Nie et al20 suggested that there may be a role for PCT in patients undergoing spine surgery following acute spinal cord injury. This study investigated the efficacy of serum PCT as a marker for early infectious postoperative complications in acute traumatic spinal cord injury patients, finding it to be superior to CRP when using a PCT cutoff value of 0.5 ng/ mL. This study showed that postoperative PCT levels were significantly higher in infected patients than in noninfected patients; at a cutoff value of 0.5 ng/mL, serum PCT had an 88% sensitivity, 74% specificity, 3.42 PLR, and 0.16 NLR.20 Further investigation of PCT following spine surgery was conducted by Chung et al.21 They found PCT to be significantly higher in patients who had longer spine surgeries, cervical operations, and use of instrumentation. However, they suggested that PCT was still useful in determining infection in neurosurgical patients with fever of unknown origin. A study by Uçkay et al22 showed that only 25% of patients with blood culture–negative postoperative orthopedic infections had elevated serum PCT levels at any point. Their results also showed that the median PCT level was elevated only on postoperative day 1, similar to CRP. However, this study had several limitations, including being retrospective, having a small sample, and not including patients without infection.22 Worthington et al23 investigated various serologic markers in patients with total joint prosthesis revisions. Their results suggest that PCT is not helpful in differentiating aseptic loosening from septic loosening.23 However, this study had a limited sample and did not discuss the average PCT levels of the infected and noninfected cohorts; rather, it simply used a single cutoff of 0.5 ng/mL for detection of infection.

The significance of finding a novel, reliable serum marker for postoperative orthopedic infections cannot be overstated. Symptoms of orthopedic surgical site infection, such as periprosthetic joint infection (PJI), can often mirror a normal, uninfected postoperative course. In fact, just establishing diagnostic criteria for PJI has proven difficult. The Musculoskeletal Infection Society released new diagnostic criteria for PJI in 2011; however, the authors acknowledged that PJI may be present even without satisfying some of the criteria, and that in low-grade PJI, “several of these criteria may not be routinely met.”1 For detecting PJI, methods other than serum markers, such as synovial fluid analysis, have also been investigated and have shown promise. However, the invasiveness, difficulty, and cost of synovial fluid analysis make a reliable serum marker far more desirable.14

The current meta-analysis study is the first to indicate the potential utility of serum PCT in identifying postoperative orthopedic infections. The lack of overall data and the inconsistency of data across studies on the topic of serum PCT as an orthopedic postoperative infectious marker are both potential shortcomings of this study. In this meta-analysis, only 6 studies with a total of 319 patients were included. There was significant variability between these studies, as seen in the results of the heterogeneity analysis. Five of the 6 studies examined patients undergoing several different types of orthopedic surgeries. Bottner et al,10 Randau et al,11 and Glehr et al24 included both revision total hip arthroplasty and total knee arthroplasty patients. Similarly, Ettinger et al25 included total joints but also added some revision shoulder arthroplasty patients. Hunziker et al26 examined both arthroplasty and fracture patients. Only Yuan et al27 investigated patients undergoing one procedure type, which was revision total hip arthroplasty. Additionally, as mentioned above, all 6 studies used different cutoff values for a positive test result for serum PCT levels, yielding high I2 values. The inclusion of both acute and chronic infections could have an effect on the data. Given the lack of studies available, all infections were included in this analysis. However, on further investigation, stronger conclusions could be reached based on time from surgery. Given the results seen in other surgical fields, however, the authors believe that the lack of available data for this meta-analysis only strengthens the argument for further research to be performed.

Current literature has shown that PCT itself has several limitations. One study suggested that PCT cannot be used alone to discriminate between aseptic loosening and low-grade infection.25 However, this conclusion was based on a low specificity (only 27.8%); the current meta-analysis found a pooled weighted specificity of 69.4%. Another limitation to using PCT is that it is often elevated in patients with renal dysfunction. Amour et al28 found PCT to be significantly higher in both infected and noninfected cohorts of patients with renal impairment. This can certainly affect orthopedists' patients, as many patients who require arthroplasties are older and may have baseline renal impairment.

Conclusion

This study sheds some light on the potential role of PCT in diagnosing infections in postoperative orthopedic patients. Sufficient evidence seems to support PCT as a standalone marker. Further investigation may reveal additional diagnostic value when used in conjunction with other markers. Although not specific for joint infection, in the absence of other clinical findings of bacterial infection elsewhere in the body, PCT appears to be a highly sensitive and specific initial test. However, for serum PCT to be used as the sole marker for postoperative infection, studies must be conducted investigating the appropriate cutoff values for diagnosing infection. These cutoff values may vary based on surgery performed; however, given that PCT solely reflects infection and not inflammation, a standard cutoff value has the potential to be established. More homogeneous studies with higher levels of evidence, consistent cutoffs, and multiple markers on the same patient populations are of utmost importance. Although PCT appears to be accurate in diagnosing postoperative infection, arthrocentesis is still required to confirm the diagnosis and to allow for more individualized antibiotic treatment.

References

  1. Workgroup Convened by the Musculoskeletal Infection Society. New definition for periprosthetic joint infection. J Arthroplasty. 2011; 26(8):1136–1138. doi:10.1016/j.arth.2011.09.026 [CrossRef]
  2. Hansford BG, Stacy GS. Musculoskeletal aspiration procedures. Semin Intervent Radiol. 2012; 29(4):270–285. doi:10.1055/s-0032-1330061 [CrossRef]
  3. Sharma P, Patel K, Baria K, et al. Procalcitonin level for prediction of postoperative infection in cardiac surgery. Asian Cardiovasc Thorac Ann. 2016; 24(4):344–349. doi:10.1177/0218492316640953 [CrossRef]
  4. Shaikh MM, Hermans LE, van Laar JM. Is serum procalcitonin measurement a useful addition to a rheumatologist's repertoire? A review of its diagnostic role in systemic inflammatory diseases and joint infections. Rheumatology (Oxford). 2015; 54(2):231–240. doi:10.1093/rheumatology/keu416 [CrossRef]
  5. Jebali MA, Hausfater P, Abbes Z, Aouni Z, Riou B, Ferjani M. Assessment of the accuracy of procalcitonin to diagnose postoperative infection after cardiac surgery. Anesthesiology. 2007; 107(2):232–238. doi:10.1097/01.anes.0000271871.07395.ad [CrossRef]
  6. Laifer G, Wasner M, Sendi P, et al. Dynamics of serum procalcitonin in patients after major neurosurgery. Clin Microbiol Infect. 2005; 11(8):679–681. doi:10.1111/j.1469-0691.2005.01205.x [CrossRef]
  7. Michaelidis CI, Zimmerman RK, Nowalk MP, Fine MJ, Smith KJ. Cost-effectiveness of procalcitonin-guided antibiotic therapy for outpatient management of acute respiratory tract infections in adults. J Gen Intern Med. 2014; 29(4):579–586. doi:10.1007/s11606-013-2679-7 [CrossRef]
  8. Lee JH, Lee JH, Kim JB, Lee HS, Lee DY, Lee DO. Normal range of the inflammation related laboratory findings and predictors of the postoperative infection in spinal posterior fusion surgery. Clin Orthop Surg. 2012; 4(4):269–277. doi:10.4055/cios.2012.4.4.269 [CrossRef]
  9. Panic N, Leoncini E, de Belvis G, Ricciardi W, Boccia S. Evaluation of the endorsement of the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement on the quality of published systematic reviews and meta-analyses. PLoS One. 2013; 8(12):e83138. doi:10.1371/journal.pone.0083138 [CrossRef]
  10. Bottner F, Wegner A, Winkelmann W, et al. Interleukin-6, procalcitonin and TNF-alpha: markers of peri-prosthetic infection following total joint replacement. J Bone Joint Surg Br. 2007; 89(1):94–99. doi:10.1302/0301-620X.89B1.17485 [CrossRef]
  11. Randau TM, Friedrich MJ, Wimmer MD, et al. Interleukin-6 in serum and in synovial fluid enhances the differentiation between periprosthetic joint infection and aseptic loosening. PLoS One. 2014; 9(2):1–6. doi:10.1371/journal.pone.0089045 [CrossRef]
  12. Bilgen O, Atici T, Durak K, KaraeminogullariBilgen MS. C-reactive protein values and erythrocyte sedimentation rates after total hip and total knee arthroplasty. J Int Med Res. 2001; 29(1):7–12. doi:10.1177/147323000102900102 [CrossRef]
  13. Mok JM, Pekmezci M, Piper SL, et al. Use of C-reactive protein after spinal surgery: comparison with erythrocyte sedimentation rate as predictor of early postoperative infectious complications. Spine (Phila Pa 1976). 2008; 33(4):415–421. doi:10.1097/BRS.0b013e318163f9ee [CrossRef]
  14. Jacovides CL, Parvizi J, Adeli B, Jung KA. Molecular markers for diagnosis of periprosthetic joint infection. J Arthroplasty. 2011; 26(6)(suppl):99–103. doi:10.1016/j.arth.2011.03.025 [CrossRef]
  15. Syvänen J, Peltola V, Pajulo O, et al. Normal behavior of plasma procalcitonin in adolescents undergoing surgery for scoliosis. Scand J Surg. 2014; 103(1):60–65. doi:10.1177/1457496913504910 [CrossRef]
  16. Kassir R, Blanc P, Tibalbo LMB, Breton C, Lointier P. C-reactive protein and procalcitonin for the early detection of postoperative complications after sleeve gastrectomy: preliminary study in 97 patients. Surg Endosc. 2015; 29(6):1439–1444. doi:10.1007/s00464-014-3821-2 [CrossRef]
  17. Perrakis A, Yedibela S, Schellerer V, Hohenberger W, Müller V. Procalcitonin in the setting of complicated postoperative course after liver transplantation. Transplant Proc. 2010; 42(10):4187–4190. doi:10.1016/j.transproceed.2010.08.070 [CrossRef]
  18. Shen CJ, Wu MS, Lin KH, et al. The use of procalcitonin in the diagnosis of bone and joint infection: a systemic review and meta-analysis. Eur J Clin Microbiol Infect Dis. 2013; 32(6):807–814. doi:10.1007/s10096-012-1812-6 [CrossRef]
  19. Maharajan K, Patro DK, Menon J, et al. Serum procalcitonin is a sensitive and specific marker in the diagnosis of septic arthritis and acute osteomyelitis. J Orthop Surg Res. 2013; 8:19. doi:10.1186/1749-799X-8-19 [CrossRef]
  20. Nie H, Jiang D, Ou Y, et al. Procalcitonin as an early predictor of postoperative infectious complications in patients with acute traumatic spinal cord injury. Spinal Cord. 2011; 49(6):715–720. doi:10.1038/sc.2010.190 [CrossRef]
  21. Chung YG, Won YS, Kwon YJ, Shin HC, Choi CS, Yeom JS. Comparison of serum CRP and procalcitonin in patients after spine surgery. J Korean Neurosurg Soc. 2011; 49(1):43–48. doi:10.3340/jkns.2011.49.1.43 [CrossRef]
  22. Uçkay I, Garzoni C, Ferry T, et al. Postoperative serum pro-calcitonin and C-reactive protein levels in patients with orthopedic infections. Swiss Med Wkly. 2010; 140:w13124.
  23. Worthington T, Dunlop D, Casey A, Lambert P, Luscombe J, Elliott T. Serum procalcitonin, interleukin-6, soluble intercellular adhesin molecule-1 and IgG to short-chain exocellular lipoteichoic acid as predictors of infection in total joint prosthesis revision. Br J Biomed Sci. 2010; 67(2):71–76. doi:10.1080/09674845.2010.11730294 [CrossRef]
  24. Glehr M, Friesenbichler J, Hofmann G, et al. Novel biomarkers to detect infection in revision hip and knee arthroplasties. Clin Orthop Relat Res. 2013; 471(8):2621–2628. doi:10.1007/s11999-013-2998-3 [CrossRef]
  25. Ettinger M, Calliess T, Kielstein JT, et al. Circulating biomarkers for discrimination between aseptic joint failure, low-grade infection, and high-grade septic failure. Clin Infect Dis. 2015; 61(3):332–341. doi:10.1093/cid/civ286 [CrossRef]
  26. Hunziker S, Hügle T, Schuchardt K, et al. The value of serum procalcitonin level for differentiation of infectious from noninfectious causes of fever after orthopaedic surgery. J Bone Joint Surg Am. 2010; 92(1):138–148. doi:10.2106/JBJS.H.01600 [CrossRef]
  27. Yuan K, Li WD, Qiang Y, Cui ZM. Comparison of procalcitonin and C-reactive protein for the diagnosis of periprosthetic joint infection before revision total hip arthroplasty. Surg Infect (Larchmt). 2015; 16(2):146–150. doi:10.1089/sur.2014.034 [CrossRef]
  28. Amour J, Birenbaum A, Langeron O, et al. Influence of renal dysfunction on the accuracy of procalcitonin for the diagnosis of postoperative infection after vascular surgery. Crit Care Med. 2008; 36(4):1147–1154. doi:10.1097/CCM.0b013e3181692966 [CrossRef]

Pooled Negative Likelihood Ratio, Positive Likelihood Ratio, and Diagnostic Odds Ratio

RatioEstimated95% Confidence IntervalPI2P for I2
Negative likelihood0.4230.267–0.671<.00186.89.001
Positive likelihood1.7781.184–2.670<.00583.01.001
Diagnostic odds5.7702.512–13.251<.00148.21.086

Demographics and Raw Data From the Studies Included in the Meta-analysis

Study (Year)CountryNo. of PatientsCutoff, ng/mLSensitivitySpecificityLevel of Evidence

InfectedNoninfected
Bottner et al10 (2007)Germany21570.333%98%II
Hunziker et al26 (2010)Switzerland45580.189%16%II
Glehr et al24 (2013)Austria55290.05581%54%II
Randau et al11 (2014)Germany48724612.9%100%II
Yuan et al27 (2015)China25460.580%73.9%II
Ettinger et al25 (2015)Germany20570.02590%27.8%II

Data Regarding Procalcitonin, White Blood Cell Count, C-Reactive Protein, Interleukin-6, and TNF-α Pooled From the Studies Included in the Meta-analysis

MarkerNo. of StudiesSensitivitySpecificityDORPLRNLR
Procalcitonin667.3%69.4%5.7701.7780.423
White blood cell count360.5%71.5%4.5941.7110.494
C-reactive protein480.6%78.1%17.3053.7460.218
Interleukin-6486.8%74%30.0003.8650.204
TNF-α239%90.4%6.2584.1850.683
Authors

The authors are from the Department of Orthopaedic Surgery (RBI, DPM, RA, AD, QN, CBP, WPU, AVM, BGD), State University of New York, Downstate Medical Center, Brooklyn, and the Department of Medicine, Hospital for Special Surgery, New York, and the Department of Medicine, Weill Medical College, Cornell University (CC), New York, New York; and Hamad General Hospital (AA, GA), Doha, Qatar.

Mr Ingber, Dr Alhammoud, Mr Murray, Dr Abraham, Dr Dixit, Dr Naziri, Dr Ahmed, Dr Urban, Dr Craig, Dr Maheshwari, and Dr Diebo have no relevant financial relationships to disclose. Dr Paulino is a paid presenter and speaker for DePuy and Ethicon.

Correspondence should be addressed to: Bassel G. Diebo, MD, Department of Orthopaedic Surgery, State University of New York, Down-state Medical Center, 450 Clarkson Ave, Box 30, Brooklyn, NY 11203 ( dr.basseldiebo@gmail.com).

Received: May 16, 2017
Accepted: July 31, 2017
Posted Online: April 16, 2018

10.3928/01477447-20180409-07

Sign up to receive

Journal E-contents