Serum Calcium Level Combined with Platelet Count May Be Useful Indicators for Assisted Diagnosis of Extremity Posttraumatic Osteomyelitis: A Comparative Analysis

Abstract

Background. A previous study had reported that patients with osteomyelitis (OM) appeared to be more likely to develop hypocalcemia before and after surgery. Calcium sulfate (CS) is frequently used as a local antibiotic vehicle in the treatment of OM, which may also affect serum calcium level. However, whether changes of serum calcium level are caused by OM and/or local use of calcium sulfate remains unclear. Also, platelet (PLT) count plays a crucial predictive role in periprosthetic joint infections (PJIs), but its role in assisted diagnosis of OM is largely unknown. The purpose of this study was to determine whether serum calcium level and PLT count may be helpful in assisted diagnosis of PTOM. Methods. Between January 2013 and December 2018, we analyzed 468 consecutive patients (392 males and 76 females), including 170 patients with posttraumatic OM (PTOM), 130 patients with aseptic bone nonunion (ABN), and 168 patients recovered from fractures with requirement of implant removal set as controls. Preoperative serological levels of calcium, phosphorus, and PLT were detected, and comparisons were conducted among the above three groups. Additionally, correlations and receiver operating characteristic (ROC) curves were displayed to test whether calcium level and PLT can differentiate patients with ABN and PTOM. Results. Outcomes showed that the incidences of asymptomatic hypocalcemia (PTOM vs. ABN vs. $\text{controls}=22.94\%$ vs. 6.92% vs. 8.82%, ${\chi }^2=21.098$ , $P<0.001$ ) and thrombocytosis (PTOM vs. ABN vs. $\text{controls}=35.3\%$ vs. 13.84% vs. 12.35%, ${\chi }^2=28.512$ , $P<0.001$ ) were highest in PTOM patients. Besides, the mean serological levels of phosphorus in PTOM and ABN patients were significantly higher than those in the controls ( $P=0.007$ ). The Area Under the Curve (AUC) of the ROC curve outcomes revealed that, with the combination of serum calcium level with PLT count, the predictive role was acceptable (AUC 0.730, $P<0.001$ , 95% CI 0.681-0.780). Also, serological levels of calcium of 2.225 mmol/L and PLT count of $246.5\times {10}^9/\text{L}$ were identified as the optimal cut-off values to distinguish patients with and without PTOM. However, age- and gender-related differences in serum calcium levels (age, $P=0.056$ ; gender, $P=0.978$ ) and PLT count (age, $P=0.363$ ; gender, $P=0.799$ ) were not found to be statistically significant in any groups. In addition, no significant correlations were identified between serum calcium level and PLT count ( $R=0.010$ , $P=0.839$ ). Conclusions. Asymptomatic hypocalcemia and thrombocytosis appeared to be more frequent in this cohort with PTOM. Serological levels of calcium and PLT count may be useful biomarkers in screening patients suspected of PTOM.