DOI: https://doi.org/10.21203/rs.3.rs-1417741/v1
Delayed diagnosis of surgical site infection (SSI) in lumbar surgery may eventually result in surgical failure. In this prospective two-centers study, we introduced a method to identify and control likely SSI following OPLS at early stages.
Patients with likely SSI according to our criteria were enrolled. A bedside irrigation treatment with no need for anesthesia would then be placed. Clinical information including the pre-operative characteristics and post-operative outcomes were collected and fully analyzed. All patients were followed up for 1 year after their hospital discharge.
Among 672 patients who received OPLS, there are 72 patients (10.7%) diagnosed with likely SSI. Lumbar spinal stenosis (41 cases, 56.9%) accounts for the most common primary diagnosis. Only 14 of the 72 patients (19.4%) have a positive microbiological tests result. During the bedside irrigation, the level of inflammatory biomarkers demonstrated a rapid decrease. The irrigation lasted for an average time of 5.4 ± 1.2 days. The average time for antibiotics administration is 13.8 ± 4.3 days. The average length of hospital stay is 27.1 ± 5.6 days. The fusion rates at 6-months and 12-moths are 84.7% and 97.2% respectively. None patient had even a minimal disability at 1-year follow up according to the Oswestry Disability Index.
Early identification and control of likely SSI following OPLS is valuable in clinical practice and the bedside irrigation system seems to be a safe and effective choice. Future large sample sized control studies are awaited to provide stronger evidence for its efficacy.
Currently, traditional open posterior lumbar surgery (OPLS) is used to manage a variety of spinal pathologies, including degenerative diseases, severe instability, spondylolisthesis, deformity, and trauma.1 Due to procedure-related increase in operative time, blood loss, and tissue damage, the incidence of surgical site infection (SSI) following OPLS is relatively higher than that of minimally invasive surgery.2 Despite significant efforts toward mitigation, SSI remains a common and one of the fatal complications following OPLS. Considerable morbidity may be associated with SSI, including hospital readmission, revision surgery, and delayed rehabilitation.3 In previous reports, the incidence of SSI has ranged from 1 to 4%.1,4,5
Diagnosis of SSI depended on either definite pus discharge from the wound or identification of germs in exudates.6 Nevertheless, it would be ideal to start the treatment before the infection results in local pus. Successful management of SSI following OPLS has been reported in some studies using approaches such as multiple debridements with continuous irrigation, vacuum-assisted wound closure, local tissue flap coverage, implantation of polymethyl-methacrylate beads impregnated with antibiotics and long-term use of intravenous antibiotics followed by oral intake.4,7−10 However, none of those treatments started at early stages of the infection and most of them need to be performed in the operating room with anesthesia, which significantly increases the total cost.
It is obvious that the proper management of SSI following OPLS is far from being standardized. Herein we introduced a method to recognize those patients who very likely have a SSI at early stages. An easy bedside irrigation system, which may help avoiding the risk of debridement and decreasing the medical cost, is also described to manage the wound.
This is a prospective study conducted in two clinical centers in mainland China between July 1st of 2016 and Jun 30st of 2019. The study protocol was made by the hospital committee of “fight with SSI in spine surgery project”, ethically approved by the Human Research Ethics Committee of The First Affiliated Hospital of Nanchang University and Beijing Tsinghua Changgung Hospital respectively. Informed consent was obtained from all participants.
A total of 672 consecutive patients underwent OPLS in two medical centers (Fig. 1). During the end of OPLS operation, a drain would be positioned in the incision connecting with a suction reservoir (Hemovac or Jackson-Pratt–like suction drainage) before the wound was tightly closed in layers. The drain would be removed if its volume is less than 50mL over 24 hours, which was usually possible after 4 to 5 days post-operation. On the other hand, vein blood samples would be collected for testing inflammatory indicators at day 1, 3, 5 post-operation.
As shown in Fig. 1, the patient will be considered as likely SSI with one of the following conditions: 1) Volume of drain is still over 100 ml over 24 hours and color of drain turns turbid/crimson at day 5. Dural damage should be excluded in this condition. 2) Prolonged fever or escalating level of inflammatory biomarkers after 5 days post-operation. Other sources of infections such as pneumonia or urinary tract infection should be excluded in this condition. 3) Contaminated dressing by oozing for 3 consecutive days after the drain tube was removed.
Once a patient was considered with likely SSI, a bedside operation with no need for anesthesia would be performed by an experienced attending doctor. As shown in Fig. 2 and Fig. 3, sutures at the end of the incision need to be removed in order to insert a soft nasal tube through the lumbodorsal fascia. Usually there is a breakthrough feedback from the end of the tube once it get through the lumbodorsal fascia layer. According to our experience, the end of the tube should be placed through the lumbodorsal fascia and then goes deeper for about 2 cm. Since the tube we used in our study is soft and easily get twisted, we do not recommend placing it against the lamina or the implants. After that, injection of a few normal saline was needed to make sure the tube is not blocked or twisted. After the successful assembling of the whole devices, pure normal saline (500 ml, bid) will be irrigated into the incision slowly, and then sucked out through the tube using a suction reservoir (Hemovac or Jackson-Pratt–like suction drainage).
During the continuous irrigation, patients were asked to stay on the bed with lateral or prone position. Nurses will make sure the tubing remains well immobilized after the irrigation. Then the patient would be allowed to change position or even ambulate with the tubes.
Also, intravenous antibiotics that cover the likely causative organisms were given simultaneously. Drainage samples were sent for microbiological tests and inflammatory indicators were measured every day after the irrigation, including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), white blood cell count (WBC) and procalcitonin (PCT). Indications of removing the tube include: 1) the irrigated drains turned clear on a visual examination; 2) normalized inflammatory indicators level; 3) and absent microorganisms were identified on microbiological tests. Intravenous antibiotics were continued to the time of two consecutive results of normalized inflammatory indicators.
Patients were free to choose to stop the treatment if they showed poor tolerance to the irrigation/draining process. On the other hand, if the infection still becomes out of control after 5 days of irrigation, the irrigation treatment would be considered failed and the patient would be sent to the operating room. The treatment failure criteria including dehiscence of the wound, definite pus discharge from the incision, recurrent fever, or continuous increasing level of inflammatory indicators.
Once a patient was enrolled, his/her clinical information would be collected and fully analyzed. Those information included patient characteristics, admission lab results, details of the operation (primary diseases, comorbidities, levels of lumbar spine involved, operation time), and details of the treatment (elapsed time to diagnosis, causative microorganisms, duration for continuous irrigation, duration of antibiotics administration, length of hospital stay and their medical costs).
All patients were followed up for 1 year after their hospital discharge. They received tests for lumbar X-ray and inflammatory indicators test every 3 months after discharge. Criteria for fusion included 1) solid posterolateral, facet, or disk space bridging bone, 2) no translational or angular motion on flexion/extension X-rays, and 3) intact posterior hardware without evidence of screw loosening or breakage. It was classified as fused when satisfied all three criteria. Patient would be considered unfused when bone continuity was interrupted, and uncertain when bone continuity across the operated segment was questionable. Successful treatment of likely SSI was defined as no more clinical, laboratory, and image finding of SSI at the 1-year follow-up.
Statistical analysis was performed using SPSS 13.0 (SPSS Inc., Chicago, IL, USA). Patient characteristics were summarized as counts with percentages or means with standard error of mean as appropriate. Student’s t-test for independent samples was used to assess continuous data.
There were 72 patients diagnosed with likely SSI, the incidence of which is 10.7%. The average elapsed time from operation to the diagnosis is 7.1 ± 1.3 days. During the irrigation treatment, seven cases dropped off because of poor tolerance to the irrigation system (need to change the wound dressing 3–4 times a day since it’s easy to get wet after irrigation), and received debridement in operation room. The rest 65 patients received the whole treatment.
Demographics and clinical parameters of the included patients were listed in Table 1. The average age is 53.6 ± 7.5 years old and 44.4% of enrolled cases are male. We also looked at the primary diagnosis: there were 41 cases (56.9%) of lumbar spinal stenosis, 21 cases (29.2%) of lumbar spondylolisthesis, 8 cases (11.1%) of degenerative scoliosis, and 2 cases (2.8%) of metastatic tumor. Near half of the operations (35 cases, 48.6%) have 2 levels involved, 23.6% of them (17 cases) involved 1 level, and 13.9% of them (10 cases) involved 3 or more than 4 levels. Instrumented spinal fusion was involved in all cases.
| Value | |
|---|---|
| Male (n, %) | 32, 44.4% |
| Age | 53.6 ± 7.5 |
| Primary disease | |
| Lumbar spinal stenosis | 41, 56.9% |
| lumbar spondylolisthesis | 21, 29.2% |
| Degenerative scoliosis | 8, 11.1% |
| Metastatic tumor | 2, 2.8% |
| Number of levels involved (n, %) | |
| 1 | 17, 23.6% |
| 2 | 35, 48.6% |
| 3 | 10, 13.9% |
| ≥ 4 | 10, 13.9% |
| Elapsed time to diagnosis (days) | 7.1 ± 1.3 |
We also analyzed the changing tendency of different inflammatory biomarkers during the bedside irrigation treatment (Fig. 4). The biomarkers CRP, ESR, WBC and PCT were measured at various time points before and after the start of irrigation. The mean CRP and ESR values were significantly higher than the normal level before the start of irrigation and dropped rapidly during the following treatment. The average level of WBC and PCT did not show significant changes.
Clinical outcomes of enrolled patients were further analyzed, as shown in Table 2. The irrigation lasted for an average time of 5.4 ± 1.2 days. Sutures were removed averaged on 19.5 ± 3.2 days post-operation. The average time for antibiotics administration and fully SSI elimination is 13.8 ± 4.3 days and 12.6 ± 1.7 days respectively. The average length of hospital stay is 27.1 ± 5.6 days. The average medical cost of enrolled cases is 13,274 ± 3,471 US dollars. Only 14 patients (19.4%) have a positive microbiological test result. There were 5 cases each of methicillin-sensitive staphylococcus aureus (MSSA) and methicillin-resistant staphylococcus aureus (MRSA), and 4 cases of Escherichia coli (E-Coli). The fusion rates at 6-months and 12-moths are 84.7% and 97.2% respectively. None patient had minimal disability at 1 year follow up according to the Oswestry Disability Index (ODI).
| Value | |
|---|---|
| Time for irrigation (days) | 5.4 ± 1.2 |
| Time for antibiotics administration (days) | 13.8 ± 4.3 |
| Time from surgery to suture removal (days) | 19.5 ± 3.2 |
| Time from surgery to SSI elimination (days) | 12.6 ± 1.7 |
| Length of hospital stay (days) | 27.1 ± 5.6 |
| Medical costs (US dollars) | 13,274 ± 3,471 |
| Microbiological exam (n, %) | |
| Negative | 58 (80.6%) |
| MSSA | 5 (6.9%) |
| MRSA | 5 (6.9%) |
| E-Coli | 4 (5.6%) |
| Rates of fusion at 6 months (n, %) | 61 (84.7%) |
| Rates of fusion at 12 months (n, %) | 70 (97.2%) |
SSI is one of the most frequent healthcare-associated infections.11 It significantly hampers the benefits of surgical management in addition to reducing satisfaction from patients.12–14 As we know, frequent symptoms of spinal postoperative SSI, including incision pain, fever, and oozing, can rarely be identified at early stages. Delayed diagnosis of SSI may eventually result in surgical failure. Identification of SSI for early intervention becomes valuable in clinical practice since infection control should be lunched as early as possible. In this study, we tried to distinguish the likely SSI at very early stages and prevent it from developing into confirmed SSI.
In current study, we found a likely SSI rate of 10.7%, which is relatively higher than the results from previous studies.11,15−17 This can be explained by the fact that, the enrolled criteria in current study are actually wider than the standard diagnosis criteria of SSI. Actually there are variations in the definitions of SSI and the frequency of SSI after spine surgery varied according to different definitions.18,19 Positive microbiological evidence is needed to diagnose a SSI in most international criteria, while here in this study we used evidence of clinical signs and symptoms of infection. We noticed that most of the patients did not have culture positive laboratory values in our study. A possible reason is that, in order to have a positive microbiological test, it requires a certain number of bacteria to be grown in those cultures. While in our study the infection was identified and treated at early stages, results in a relative small number of bacterial colonization, which is not enough for a positive culture result.
Previous studies showed that the majority of SSI become apparent within 30 days of an operative procedure, and longer in instrumented surgeries. While some cases had their infections present clinically even after the initial 3-month period.12,17 In this study, we differentiated those patients at very early stages. The average elapsed time from operation to the diagnosis of likely SSI is 7.1 ± 1.3 days, which is significant earlier than that of published studies.
On the other hand, we looked at the changing tendency of different inflammatory biomarkers during the irrigation treatment. CRP and ESR values were significantly higher than the normal level before the start of irrigation and dropped rapidly during the following treatment. It seems that CRP and ESR are closely correlated with the severity of infection. This is consistent with the work from Eiichiro et al, among which they looked at CRP in 11 cases of SSI and 130 cases of non-SSI following instrumented spinal fusion.20 They concluded that CRP at day 7 post-operation had high predictive value of an SSI.
Furthermore, treatment options reported in published studies composed of extensive debridement and prolonged antibiotic therapy guided by antimicrobial susceptibility of the isolated bacteria.8,10 Those methods always come with multiple debridements, secondly wound closure and even removal of the implant. However, reopen the incision and doing multiple debridements is painful and leading to a lengthy hospital stay. It is even more difficult to decide whether to preserve or remove the implant in those situations because cage removal is technically demanding and results in total segmental instability and neurological compromise.13
Lee et al.8 found 31% of SSI patients had implant loosening with both pedicle screws and cages at the time of diagnosis. During their treatment, all implants including cages were removed and re-implantation of auto-iliac bone blocks were performed. Tsubouchi et al.13 investigated the risk factors for implant removal after spinal SSI based on data from seven spine centers, they observed that 40% of SSI patients had their implant removal. Study from Mirovsky et al.21 reported that 33.3% of patients with SSI have their cages repositioned in the face of infection, and they concluded that in postoperative uncontrolled deep infection with cages, removal of all hardware including the interbody cages is the treatment of choice. Kim et al.22 reported 10 patients who were referred to them during a 1-year period for deep SSI after OPLS with cages. All 10 patients underwent cage removal via an anterior surgical approach. However, none of the likely SSI patients enrolled in current study had implant loosening and none of them required removal of the instrumentation during the whole treatment.
On the other hand, compared with the prolonged antibiotic therapy (usually 4 to 6 weeks of intravenous antibiotic therapy followed by another 6 to 9 weeks of oral antibiotic administration) in previous studies, the average time of antibiotics administration in our study is only 13.8 ± 4.3 days, which is significant shorter. It is also known that deep SSI leads to a lower incidence of solid fusion.23–26 However, the fusion rate at 6-months and 12-moths in our study are 84.7% and 97.2% respectively, which is significantly higher than most other studies. By notes, the average medical cost of enrolled patients is 13,274 ± 3,471 US dollars, much cheaper than the cost reported by published studies.27–30
Hence, compared with traditional reported methods, we believe this early infection control treatment may help preventing the formation of SSI, decreasing the dose of antibiotic administration, reducing the length of hospitalization or medical costs, and increasing the fusion rate without removing the implant in most patients.4,27,31
Limitations to this study should be noted. First of all, it is a bedside invasive procedure, which means it may trigger secondary infections without carrying out strict septic manipulation. Trained personnel are needed to perform the insertion of the irrigation/draining devices sterilely. Second, it needs a nurse who has received relative training stay beside the patient and control the “T” type valve during the irrigation. Third, the tube is relatively thin and soft, makes it easily twisted or blocked by the debris from the incision, which results to recurrent leaking/oozing and dressing changes.
In conclusion, early identification and control of likely SSI following OPLS is valuable in clinical practice and the bedside irrigation system seems to be a safe and effective choice. Future large sample-sized control studies are awaited to provide stronger evidence for its efficacy.
Consent for publication: All authors have read and approved the manuscript.
Competing interests: None
Funding: This work was supported by the China Postdoctoral Science Foundation Grant. Grant NO: 2018M641389.
Authors' contributions:
Dong Hu and Fei Song contribute equally to this study. Songhua Xiao, conceptualization; Dong Hu, data curation; Suxi Gu and Dong Hu, formal analysis; Songhua Xiao, Dong Hu, and Dong Yang, funding acquisition; Dong Hu and Huawei Liu, investigation; Dong Hu and Fei Song, Methodology; Songhua Xiao and Dong Yang, project administration; Songhua Xiao, supervision; Dong Hu, original draft; Dong Yang and Songhua Xiao, review & editing.