Infection in neuro‐muscular scoliosis deformity correction

Abstract

Assess the outcome of a standardised protocol for the treatment of post‐operative wound infection in patients undergoing deformity correction for neuro‐muscular scoliosis (NMS). Retrospective review of 443 consecutive patients with a minimum 18 months’ follow‐up, following a primary posterior deformity correction for NMS. In patients who developed a wound complication, the patient demographic and comorbidities, causative pathogen, number of re‐operations, length of stay (LOS), rate of cure, and complications were analysed. Forty‐four patients (9.9%) developed a wound infection. Marginally more infections were mono‐microbial (23) than poly‐microbial (21). Coagulase negative staphylococcus and Staphylococcus aureus were the most commonly cultured pathogens. Seventeen patients were treated with antibiotics alone, while 27 patients also required surgical debridement. The average LOS for those treated with antibiotics alone was 12 days (range: 9‐15 days), in contrast to those requiring debridement, which was 35 days (range: 35‐70 days). All patients were cured from their infection and ultimately achieved fusion. Infection is common in NMS deformity correction. This is marginally more common as a mono‐microbial than poly‐microbial infection with most pathogens being staphylococcal in origin. Our defined treatment strategy resulted in a cure for all patients and capacity for all patients to achieve fusion.

1. INTRODUCTION

Neuro‐muscular scoliosis (NMS) is a challenging condition to treat. In general, these curves present early, are more severe and progress more rapidly than idiopathic early onset scoliosis. The need to maintain growth while avoiding curve progression in order to ensure normal pulmonary and solid organ development is critical in managing this condition.1, 2, 3, 4, 5 However, because of patient comorbidities, reduced immunity, poor skin and soft tissue quality, long procedure times, significant blood loss, and necessity for longer lengths of fusion these patients have high complication rates, particularly infection. The rate of infection in the surgical management of NMS ranges from 4% to 20%6, 7, 8 compared with syndromic scoliosis (7%‐10%)9 and idiopathic scoliosis (2%‐3%).10

Despite the relatively high rate of infection in this group of patients, limited literature on the causative pathogen, treatment, and subsequent outcomes exist. In the face of this paucity of information, we have established a protocol for treating such patients based on basic clinical reasoning. This protocol involves early microbiologist and infectious disease input in order to optimise the antibiotic therapy for these patients, which is based on culture and sensitivity results from a wound swab and blood cultures. In patients with superficial infections without wound dehiscence or systemic upset, intravenous antibiotics for 5 days followed by oral antibiotics are administered until the infection is clinically settled. In contrast, patients with wound dehiscence or systemic upset, operative intervention with wound debridements and intravenous antibiotics for 6 weeks through a peripherally inserted central catheter line is performed. This is followed by oral antibiotics for a further 3 months. Wound debridements involve the debridement of the entire surgical site with removal of all the necrotic and devitalised unhealthy tissue followed by pulsed lavage. In early infections (within 3 weeks of the index procedure) bone graft is retained. In contrast, in late infections bone graft that is noted to be sheet like (granular) is retained as this implies graft incorporation, while soft putty‐like graft is removed. Negative pressure wound therapy (NPWT) is then used (125 mmHg continuous suction) to aid healing. Debridements are performed every 72 hours until the wound is macroscopically sterile and the patients are systemically well. Subsequently, NPWT dressings are performed weekly with adjunctive medical honey or silver nitrate dressings until there is evidence of re‐epithelialisation and coverage of the metalwork. Wounds are then dressed with standard dressings and allowed to heal with secondary intention. Medical honey and silver nitrate dressings were used by our tissue viability team to preserve the moisture in the wound, enhance healing and as an adjunct to antibiotics to reduce infection and allow the NPWT to continue to be used at the same time.

To our knowledge, this is the first study to assess the outcome of a standardised protocol for the treatment of post‐operative wound infection in patients with NMS who underwent deformity correction.

2. METHODS

A retrospective review of prospectively collected data (October 2008‐October 2013) of 443 consecutive patients with NMS who underwent a primary posterior correction and instrumented fusion at two tertiary referral centres and a minimum follow‐up of 18 months was performed. Ethical approval was sought, but deemed unnecessary as this was a retrospective review of previously collected data and no new data was obtained from the previous audit. All procedures used a standard posterior midline approach and preferentially pedicle‐based fixation techniques with local autograft and the addition of an osteoconductive resorbable matrix (ProOsteon 500R, BIOMET) to aid fusion. Patients were transferred post‐operatively to the paediatric intensive care unit (PICU) where they were stabilised for pain and physiological parameters. All patients were encouraged to mobilise within the first 24 hours, and transferred to the paediatric ward when physiologically stable.

A modified ASEPSIS scoring method was used to determine wound infection.11 Early wound infection was defined as that occurring within 3 weeks of surgery. In contrast late infection was defined as occurring after 3 weeks of surgery. We defined wound dehiscence as a wound, where a wound swab could be inserted up to the metalware. The clinical notes of those who subsequently developed an infection were reviewed for the patient age at index surgery, sex, comorbidities, body mass index (BMI), infection type, number of subsequent operations, length of hospital stay, and complications.

All patients had serial radiographs at 6 weeks, 3 months, 6 months, 1 year, and then annually. Haematological inflammatory markers (Complete blood count, Erythrocyte sedimentation rate, and C‐reactive protein) were performed in all patients at 1 year and a computed axial tomography (CAT) scan of the spine performed at 18 months to determine fusion.

3. RESULTS

Of the 443 patients, we identified 44 patients (9.9%) who were diagnosed with early wound infection. Of these, 20 were male (mean age 13.8 years, range: 7‐21 years) and 24 were female (mean age 13.5 years, range: 8‐20 years) with an average follow‐up of 26 months (range: 18‐35 months). There were no cases of late infection.

The underlying cause of neuro‐disability in those with an infection is shown in Table 1.

Table 1.

The underlying cause of neuro‐muscular scoliosis in our cohort

Neuro‐disability	Number of patients
Cerebral palsy	19
Duchenne muscular dystrophy	6
Spinal muscle atrophy	4
Myopathy	4
Spina bifida	3
Cord injury	1
Others	7

Patient comorbidities are shown in Table 2

Table 2.

Comorbidities of the patients included in our cohort

		Number of patients
Comorbidities	Recurrent chest infection	22
	Epilepsy	17
	Restrictive lung disease	15
	Obstructive lung disease	2
	Cardiomyopathy	5
	Metabolic disorders	6
	Coagulopathy	11
Indwelling devices	Percutaneous Endoscopic Gastrostomy	37
	Tracheostomy	4
BMI	Overweight	17
	Normal	23
	Underweight	4

Abbreviations: BMI, body mass index.

In all patients with a wound dehiscence, the wound dehiscence was identified at 5 to 7 days post‐operatively (Figure 1).

Figure 1.

Representative photograph of a patient with early wound infection exuding purulent material

Microbiological samples yielded poly‐microbial growth in 21 patients and mono‐microbial growth in 23 cases. Table 3 lists the microorganisms isolated.

Table 3.

Bacteriology results of the infected cohort

	Type of organism	Number of cases
Mono‐microbial	Coagulase negative staphylococcus	6
	Staphylococcus aureus	5
	Streptococcus	3
	Enterococcus coli	4
	Enterococcus (other)	3
	Enterobacter	1
	Pseudomonas	1
Poly‐microbial		21

Seventeen patients were treated with antibiotics alone, while 27 patients also required surgical debridement (Table 4). As no late infections were encountered, in all patients the bone graft was left in situ. One patient who required surgical debridement and NPWT had to be excluded from the study as the parents wished for the wound to be surgically closed and the treatment discontinued.

Table 4.

Treatments of infected cohort

Treatment	Number of patients	Percentage
Antibiotics alone	17	38.6
Washouts and antibiotics	27	61.4

Of the remaining 26 patients who underwent surgical debridement, 25 required less than three debridements until wound closure was achieved (Table 5).

Table 5.

Number of washouts

Number of debridements	Number of patients
1	7
2	13
3	5
7	1

Figure 2 offers a representative example of a patient requiring two debridements and NPWT. In contrast, Figure 3 provides the clinical photographs of the patient requiring seven debridements and NPWT.

Figure 2.

Representative example of a patient undergoing two debridements and to ultimately close an infected wound with exposed metalware after NMS correction.NMS, neuro‐muscular scoliosis; NPWT, negative pressure wound therapy

Figure 3.

Representative example of a patient undergoing seven debridements to ultimately close an infected wound dehiscence after NMS correction. NMS, neuro‐muscular scoliosis

The average length of stay in those with an infection requiring debridement was 35 days (range: 35‐70 days) in contrast to those patients treated with antibiotics alone which was 12 days (range: 9‐15 days). Sixteen patients had a concomitant infection as an inpatient (respiratory 14, urinary 1, and gastrointestinal 1). Of these patients, nine required readmission to PICU after ward transfer.

Inflammatory markers at 1 year in all patients were within normal limits. In all patients, fusion was confirmed at 18 months and there were no complications identified at final follow‐up.

4. DISCUSSION

Our study confirms that wound infection after deformity correction is common in NMS, with the rate of infection in our cohort being 9.9%, which is consistent with recent literature suggesting a rate of 5.3% to 15.2%.12, 13, 14, 15 In our study, we were able to identify the causative organism in all of our patients. Marginally more infections were mono‐microbial than poly‐microbial, with the most common organisms isolated being coagulase negative staphylococcus and Staphylococcus aureus. Other authors have had a similar experience with the most common organisms being S. aureus and Staphylococcus epidermidis.16, 17, 18, 19, 20

We treated our patients with a consistent protocol. All patients received antibiotic therapy in consultation with our microbiological and infectious diseases colleagues. In patients with wound dehiscence or deep infection causing severe systemic upset, operative intervention with thorough debridement of all the necrotic and devitalised unhealthy tissue was performed and NPWT used. Further debridement procedures were performed on clinical grounds and aided by evaluation of biochemical parameters. With this protocol, we were able to cure all infections and discharge patients requiring antibiotic therapy alone at an average of 12 days and those requiring operative debridement at 35 days after an average of two debridements. We observed re‐epithelialisation and coverage of metalwork after debridements and NPWT dressings in all cases. Furthermore, there were no cases of metalwork failure on serial radiographs and all patients achieved fusion with no evidence of a pseudarthrosis on a CT scan at 18 months. This suggests that the treatment protocol, we have established offers merit in treating these patients.

However, this study has a number of limitations that need to be recognised when interpreting the results. Firstly, although to our knowledge this is the largest series of its kind, it is still affected by low numbers, which is particularly important in this cohort because of the variable causes for the underlying NMS which themselves may affect healing capacity. Secondly, no patients presented with a late infection and therefore this study fails to determine the value of our protocol in patients with late infection. Lastly, there are no comparative treatment algorithms offered in this study and therefore the value of this protocol in contrast to another cannot be discerned. However, we believe that the information offered by this research provides direction for further research into this challenging field in order to ensure optimal outcomes for patients with infection following NMS deformity correction.

In conclusion, infection in deformity correction for NMS is common. This is marginally more common as a mono‐microbial than poly‐microbial infection with most pathogens being staphylococcal in origin. Our defined treatment strategy resulted in a cure for all patients and capacity for all patients to achieve fusion.

CONFLICT OF INTEREST

The authors declare no conflicts of interest.

Haleem S, Edwin J, Bashir MA, Soltani S, Nadarajah R, Kieser DC. Infection in neuro‐muscular scoliosis deformity correction. Int Wound J. 2020;17:729–734. 10.1111/iwj.13332