Participants and ethics
In our hospital, pediatric osteomyelitis patients were recommended for surgery under the following conditions: unresponsive to empirical antibiotics, persistent pain and fever after 48–72 h of antibiotics, radiographic identification of bone lesions or evidence of purulent material. The decision to apply vancomycin-loaded calcium sulfate during surgery is dependent on the parent of the patient. This option is provided upon consultation and after determination that surgery is required. If the parent does not opt into the local vancomycin treatment, conventional surgery is conducted.
Data from 60 pediatric patients in our hospital were retrospectively retrieved from the electronic pharmacy database from January 2017 to January 2020. Inclusion criteria consisted of (1) patients with AHOM who were under the age of 18 years old, (2) received vancomycin-impregnated calcium sulfate combined with NPWT or received conventional surgical treatment followed by NPWT, or (3) had a minimum follow-up time of one year. Herein, the two groups are termed as the vancomycin group or conventional group.
Exclusion criteria consisted of patients who had other co-morbidities, patients who did not undergo surgery, patients who received other forms of treatment and patients who were on other medication. After screening, 28 patients were included in the conventional group and 32 patients were included in the vancomycin group. This study was approved by the Ethics Review Committee of Children’s Hospital Affiliated to Zhengzhou University. Informed consent to participate was waived as this was a retrospective study and no personal information was included in the data.
Data collection
Baseline demographic and characteristic data was collected from the patients consisting of age, sex, white blood cell (WBC) count, C-reactive protein (CRP) level, erythrocyte sedimentation rate (ESR), bacterial culture results and location of osteomyelitis. Outcomes collected consisted of eradication time of the infection (days to 50% reduction of CRP, normal CRP, normal WBC and normal ESR), number of surgical procedures, incidence of complications during procedure, and recurrence or complications documented at last follow up. Good outcomes were defined as no complications without recurrence at the latest follow-up.
Intervention with vancomycin-loaded calcium sulfate or conventional surgery
Once admitted to the inpatient department, antibiotic therapy was initially administered with an IV broad-spectrum antibiotic immediately after blood sample collection. Surgical intervention was warranted if the child had persistent pain and fever after 48 h or subperiosteum as well as intraosseous abscess formation identified by MRI.
The first stage surgery was performed following a standardized sequence of fenestration of necrotic infected bone, collection of pus sample, debridement of surrounding soft tissue, and opening and washing of the medullary canal (Fig. 1A). On another and sterile work station, calcium sulphate antibiotic impregnated beads were prepared using a 5-ml kit of pharmaceutical-grade calcium sulfate alpha-hemihydrate (Stimulan; Biocomposites Ltd., Staffordshire, England) mixed with 500 mg vancomycin powder plus 5 ml of sterile water for injection. A smooth paste was first formed mixing all components for 60 s and then kneaded into pellets. After changing the gloves and drapes, the pellets were placed into the bone window and the NPWT device (Zhongxinkang, Foshan, China) was applied (Fig. 1B). If concomitant septic arthritis was present, the joint was debrided simultaneously and covered with the NPWT device. Pellets were not placed into the joint. For the second surgery, if no pus or pus mosses were found during operation, the pellets were replaced and the NPWT device was removed. Otherwise, both of the pellets and NPWT device were replaced and an additional surgery was required to remove the NPWT device.
Postoperatively, the limb was immobilized in a posterior brace to protect it from pathological fractures (Fig. 1C). Parenteral vancomycin was administered for a standard duration of four weeks and later switched to oral administration of cefixime. If laboratory markers did not return to normal after four weeks, parenteral vancomycin was administered until blood parameters (WBC, CRP, ESR) normalized. The NPWT device was irrigated intermittently with 100 ml of 0.9% sodium chloride solution twice a day. WBC and CRP counts were noted every four or five days postoperatively.
Aside from the addition of calcium sulfate antibiotic-loaded pellets, the same procedure was conducted for the conventional surgery. After surgical incision, fenestration of necrotic infected bone, collection of pus sample, debridement of surrounding soft tissue, and opening and washing of the medullary canal was conducted. The NPWT device was applied without the addition of localized antibiotics. If pus or pus mosses were present in subsequent surgeries, the NPWT device was replaced and additional surgery was required to remove the NPWT device.
Statistical methods
Descriptive statistics of baseline values and outcomes were tabulated. A chi-square test was conducted to determine statistical differences in the number of surgeries performed per patient as well as the rate of complications between the treatment and control group. Normality of distribution was examined for the outcomes describing the time to recovery, which consisted of days to 50% CRP reduction, days to normal CRP levels, days to normal WBC levels and days to normal ESR levels. A non-normal distribution was found in all outcomes. Homogeneity of variance was conducted, which verified that there were dissimilar distributions of shape between the control and treatment group. Thus, bootstrap t-tests were conducted to compare means. One-way ANCOVA was conducted to determine the dependency of recovery times on covariates. P-values of less than 0.05 were taken to be statistically significant. All statistical methods were implemented in SPSS.