Limiting morbidity in neuromodulation

Preventing infection

Although it would be expected for neuromodulation procedures to have a higher infection rate due to the presence of implanted devices and the frequency of externalized components (like electrodes or extensions during stimulator trials, and temporary epidural or IT catheters used in screening for subsequent pump insertion), the incidence of infection is comparable to other surgical interventions. Contemporary studies have reported an overall SCS infection rate between 2% and 3% [ , ]. This may be, at least in part, explained by the elective nature of neuromodulation procedures. Since insertion of the neuromodulation device is usually done in a planned fashion, there is also a time interval that allows the implanter to check the patient for ongoing infection or bacterial colonization, and to treat this infection to decrease the chances of the new device becoming colonized. IT pumps tend to have a higher reported infection rate, ranging from 3% to 10% [ , ], particularly in patients with spasticity, whose comorbidities may include frequent decubitus ulcers and urinary tract infections.

Prospective neuromodulation patients should be assessed for signs of infection, which includes obtaining routine laboratory tests (peripheral white blood cell count, urinalysis, etc.), and postponing surgery, including trial procedures, until any infection is treated adequately. In addition, in patients with known exposure or infection with methicillin-resistant Staphylococcus aureus (MRSA), it is generally recommended to obtain nasal swabs and treat those who are positive with chlorhexidine showers until the colonization and/or infection are eliminated [ ]. It is recommended to follow an institutional protocol for addressing MRSA infections if one exists. Similar strategies may be used when methicillin-sensitive Staphylococcus aureus is identified on nasal swabs. In fact, all our patients are provided chlorhexidine wipes preoperatively.

To prevent infectious complications and promote better long-term prognosis, there are also modifiable lifestyle factors which should be considered, such as obesity and tobacco use.

Obesity

Obesity is often associated with diabetes, and it is known that patients with poorly controlled diabetes have a higher incidence of postoperative complications [ , ], such as infection secondary to poor wound healing. These can include superficial surgical site infections, deep wound infections, or surgical space abscesses [ ]. In particular, the relationship between diabetes and postoperative complications has been well-documented in patients undergoing elective spine surgery [ , , ]. Although recent large retrospective series of SCS infection rates did not identify diabetes as a risk factor, these studies did not differentiate between poorly and well-controlled disease [ , ]. Furthermore, prospective studies have not been performed.

As such, preoperative screening for diabetes to ensure optimized blood glucose levels is recommended prior to all neuromodulation procedures. It is common to evaluate blood glucose and/or hemoglobin A1c levels; although these are strongly associated, it is not clear if one measure is superior in predicting morbidity and mortality [ ]. Recommendations have been established to target perioperative blood glucose between 80 and 180 mg/dL, although a target for A1c has not been established [ , ]. In general, A1c levels less than 7% in young, healthy patients and less than 8% in older patients with multiple comorbidities are recommended for reducing the risk of heart disease and stroke [ ].

Compared with other comorbidities, obesity (defined as a BMI ≥30 kg/m ² ) has been identified as an independent predictor of unplanned readmission within 30 days of SCS implantation [ ]. The other pertinent predictor of readmission in this study was infection, either related to the SCS hardware or an unrelated systemic infection. As these data were collected from a national database, further interpretation regarding the reason for readmission could not be determined. High readmission rates have been associated with a negative impact on patient outcomes such as increased mortality [ , ]. Unplanned readmissions, including those due to postsurgical infection, contribute to growing national healthcare costs. In recent years, a major healthcare goal has become the reduction in unplanned readmission rates.

Interestingly, obesity has been reported to be predictive of outcomes following SCS as well. In a retrospective review of 77 patients undergoing SCS implantation, it was reported that patients with BMI ≥36.5 kg/m ² demonstrated a 48% worse score on the Beck Depression Inventory (BDI) at 1 year, compared to a 23% improvement in patients with BMI ≤36.5 kg/m ² [ ]. Similar trends were also seen in the Pain Catastrophizing Scale (PCS) scores at 1 year, with 51% improvement in low BMI patients and 49% worsening in high BMI patients. Moreover, it has been reported that with every unit of BMI increase, the efficacy of SCS therapy is reduced by 2%, whereas patients with normal BMI have a 20% better response than in those who are morbidly obese [ ].

Obesity is also known to be associated with mechanical complications and technical obstacles during an operation. Patient positioning, tissue depth, and other intraoperative challenges can interfere with lead and generator placement [ ]. Obesity has also been reported to be predictive of revision SCS surgery for lead migration and hardware malfunction [ ]. Regardless, obese patients comprise a large cohort of those considered for neuromodulation surgery, which necessitates special considerations [ ]. Often special surgical equipment, such as operating tables capable of accommodating large weight, have to be arranged for patients with markedly elevated BMIs. For example, we are unable to use the Wilson frame in patients over 300 pounds, and chest rolls limit the amount of flexion possible. As such, we tend to place the SCS lead slightly higher by about one-third to one-half of a vertebra, so that the retraction that occurs with flexion does not pull the lead into an unwanted position [ ]. It is prudent to consider reinforcing the anchoring devices with additional sutures or fibrin glue if appropriate. It is also imperative to have family members or caregivers instructed in proper wound care management.

Tobacco use

Tobacco consumption is the number one preventable cause of death in the United States. According to the Centers for Disease Control and Prevention, tobacco related illnesses in the United States costs more than $300 billion each year [ ]. In addition to obesity, a previous comprehensive review of the literature and postmarketing surveillance data suggested stopping tobacco use for 1 month prior to surgery to further mitigate against infection [ ]. Studies have consistently demonstrated the analgesic effects of tobacco [ ]. However, regular tobacco consumers have reported adverse pain symptoms and impairment compared to nonsmokers.

There is an association between tobacco usage with surgical infection and delayed wound healing. Studies have found that wound contraction is heightened in smokers and may be related to altered myofibroblast function, alterations in epidermal regeneration, neovascularization, and damaging vasoactive effects [ ]. These healing issues may lead to more frequent revisions [ , ]. A 2015 study demonstrated that in a cohort of 20 smokers with SCS devices, two required device explant surgery, two required revision due to lead migration, and three required revision due insufficient pain management and worsening symptoms [ ]; these failures were possibly due to impaired wound healing and aberrant pain processing. Smoking has also been associated with an increased risk of infection in DBS [ ].

Preoperative comorbidities, such as smoking and obesity, are modifiable risks. Controlling them allows physicians to mitigate many perioperative and postoperative complications. That tobacco usage correlates and early failure of SCS [ ] may be explained not only by physiological issues [ ], but psychological factors as well. These findings should prompt physicians to better counsel their patients regarding tobacco cessation prior to surgery to avoid surgical complications in the postoperative setting. Indeed, a short 3 minute physician dialogue has been reported to increase smoking cessation rates by up to 10% [ ], but surgical trainees tend to be inadequately mentored about the importance of these discussions [ ].

In addition to modifiable factors, nonmodifiable factors also contribute to the development of perioperative complications. Such factors, including the use of anticoagulation therapy and the psychological states of patients, need to be optimized preoperatively to mitigate complications and ensure improved long-term outcomes.

Anticoagulation medications

Neuromodulation procedure outcomes may be complicated by use of anticoagulation medications. The Neuromodulation Appropriateness Consensus Committee (NACC) considers the patients on anticoagulation treatments at high risk due to the increased incidence of spontaneous bleeding and epidural hematoma formation [ , ]. The use of anticoagulation therapy, especially in the elderly patient population, has become common clinical practice for the prevention of thromboembolic events [ , ], treatment and prophylaxis of myocardial infarctions, cerebrovascular events, hypercoagulable states, and deep vein thromboses. Thus, it is a careful balancing act between procedural safety and patient risk of developing one of these events.

In 2016, the NACC published guidelines regarding the perioperative management of anticoagulants [ ]. They meticulously outlined specific neuromodulation procedures and their stratified bleeding risks while labeling each as high-, medium-, or low-risk procedures. Primarily, they recognized that neuromodulation procedures carry patient- and procedure-specific risk factors. For example, low- and medium-risk procedures (i.e., SCS, PNS, battery placements, and replacements) do not carry the same risk of hemorrhage that is associated with high-risk procedures such as DBS. The highest risk of hemorrhage was seen following DBS (0%–2.6%), with complications including neurological deficit and death. As such, the recommendations are stratified in terms of brain, spinal, and peripheral procedures.

The types of medications outlined in the NACC recommendations include aspirin, nonaspirin nonsteroidal antiinflammatories (NSAIDs), phosphodiesterase inhibitors, anticoagulants (i.e., Coumadin, heparin), fibrinolytic agents, P2Y12 inhibitors, glycoprotein IIb/IIIa inhibitors, and new oral anticoagulants. In general, nonaspirin NSAIDs, phosphodiesterase inhibitors, and anticoagulants are recommended to be stopped between 1 and 10 days, depending on the medication, prior to both brain and spinal procedures. Aside from nonaspirin medications and phosphodiesterase inhibitors, all others are recommended to be stopped 3–7 days prior to peripheral neuromodulation procedures. Table 13.1 illustrates a few of the pertinent recommendations for medication cessation and continuation.

The American Society of Regional Anesthesia and Pain Medicine (ASRA) also developed guidelines for patients on anticoagulation therapy undergoing interventional pain management procedures [ ]. They have released an app, ASRA Coags 2.0 that contains the ASRA Anticoagulation Guidelines and serves as a quick and easy reference for physician practices. Along with the guidelines the app provides quick access to drug-specific information.

In a retrospective analysis of patients undergoing SCS implantation or revision, those who had their anticoagulation medication suspended did not have an increased risk of perioperative hemorrhagic or thromboembolic adverse effects following surgery [ ]. Although no single anticoagulant contributed to adverse events, a small subset of patients demonstrated a higher incidence of adverse events when taking anticoagulants in combination with enoxaparin [ ]. The anticoagulation guidelines used in this study were more conservative compared to the NACC recommendations, notably stopping new oral anticoagulants 5 days preoperatively and resuming any anticoagulation after 5 days, if feasible. Due to the complications that can occur, it is important that physicians follow a set of guidelines on medication suspension; these can be based on the NACC recommendations, ASRA guidelines, the surgeon's personal experience, or a combination thereof.

Psychological evaluations

An important nonmodifiable factor that should be considered during patient selection is the psychological evaluation. This evaluation may serve as a selection tool and provides significant insight on surgical outcomes.

Due to the elective nature of neuromodulation surgery, it is standard for patients undergoing these procedures to have a multidimensional psychological evaluation performed preoperatively. Psychological testing can encompass a wide variety of measures such as Minnesota Multiphasic Personality Inventory (MMPI), the PCS, and the BDI. These measures can provide physicians with insight regarding patient characteristics and potentially predict the long-term prognosis and outcome of the procedure. Currently, it is well recognized that patients with untreated major depressive disorder, psychosis, and history of substance abuse are not promising candidates for SCS. For example, SCS patients with a psychiatric comorbidity (including depression, anxiety, substance abuse, and a history of physical or sexual abuse) have been identified to be more likely to undergo removal of their device within 1 year, primarily due to lack of efficacy [ ]. Although there is limited information regarding which specific psychological factor serves as the most accurate predictor of SCS outcomes, physicians agree that a comprehensive assessment is of value [ ].

In 2016, researchers conducted a study to identify the predictive value of psychological evaluations prior to SCS [ ]. They examined MMPI, BDI, and PCS findings and analyzed outcome measures including the visual analog scale, McGill Pain Questionnaire, and Oswestry Disability Index. At 1 year, patients without preexisting psychological conditions recorded a 2.4 times improvement in PCS scores compared to patients with psychological conditions. Depressed patients using antidepressants displayed significant improvements in their BDI score, at 1 year, compared to patients with depressive symptoms that were not using antidepressant medications. No difference was found in revision rates between patients with and without preexisting psychiatric conditions [ ].

Maximizing neuromodulation benefits requires a collaborative approach to patient care including medical treatment, surgical intervention, psychological counseling, and physical therapy. Patients have various aspects of their lives that interact with each other and contribute to the success and failure of their surgical interventions. Specifically, SCS outcome studies have found that multidimensional pain inventories and psychological evaluations provide more comprehensive information regarding patients' characteristics that could aid in the development and implementation of more effective treatments [ , ].

Pain heterogeneity

Pain is highly subjective, and our understanding of the pathophysiology underlying most pain conditions is incomplete. There are excellent data regarding the outcomes of SCS for pain disorders such as failed back surgery syndrome (FBSS) and complex regional pain syndrome [ , ]; however, pain is highly heterogeneous with multiple etiologies and symptoms. As such, one of the most difficult components of the preoperative workup is selecting the appropriate candidate based on their pain presentation.

Pain is primarily measured based on patient accounts and cannot be easily identified by standard examination [ ]. Therefore, misdiagnoses of pain-related diseases or locations are very common. Facial pain is particularly challenging, such as distinguishing between diagnoses. For example, PNS is of benefit in some cases of trigeminal neuropathic pain but not in trigeminal neuralgia. Furthermore, the proportion of responders and nonresponders among those with facial pain is roughly equal. Similar issues may occur in thoracic pain or visceral pain. Notably, there is currently excitement about the use of neuromodulation for pelvic pain.

Pelvic pain patients are particularly challenging as their duration of pain prior to undergoing neuromodulation is often longer and thus the pain syndrome more complex [ ]. Furthermore, there are many conditions that can lead to pelvic pain—some which are amenable to stimulation (pudendal neuralgia) and some which are not (pelvic floor dystonia). Similarly, pelvic pain is greatly misunderstood and often misdiagnosed. In order to choose an effective treatment, the correct diagnosis must be made. A proper history and exam must be undertaken and a multimodal approach is necessary. The innervation of the pelvis is very complex and thus neurogenic pelvic pain may be the most difficult to diagnose [ , ]. Physical examinations, imaging, differential diagnostic nerve blocks, and electromyography (EMG) have been used to assess the function of muscles and nerves. Using these tools, providers have been able to better pinpoint the nerve(s) responsible for the pain [ ]. This approach improves outcomes, but effectiveness of treatment is also heavily dependent on when the diagnosis is made. Diagnosis from time of symptom onset is typically delayed by as much as 10 years and thus can be accompanied by significant emotional and physiological difficulties [ ]. Our pelvic pain consortium has developed an algorithm using a variety of outcome measures which helps providers develop individualized treatment plans for pelvic pain patients [ ].

In all pain conditions, a multimodal approach is more successful than just using one single modality. How to do this logistically and across providers is a challenge. Also education of primary care providers that neuromodulation devices are but one tool in the tool box and not meant to relieve all pain is essential to patients' continued use of the device. Unfortunately, there is not a single treatment plan that works for every patient. Treatment relies heavily on patient and provider experience. First, providers depend on patient accounts of their pain or discomfort, causing treatment, and the degree of treatment to revolve around patient reports. Second, providers rely on their own past experiences to manage the situation. A doctor-patient relationship is the foundation of a successful treatment plan, especially for pain disorders and all aspects of what is going on in the patient's life must be considered.

General principles

As previously stated, infection is an important complication in neuromodulation surgery. In addition to optimizing comorbidities, it is critically dependent on appropriate sterile preparation and technique. In terms of antibiotic prophylaxis, there is no consensus among neuromodulation implanters regarding the duration, route of administration, and choice of antibiotic regimen. An almost universally accepted practice of administering preoperative antibiotics, usually second generation cephalosporins, within 1 hour of making a surgical incision represents more of a general standard that applies to most surgeries done in the operating room. The choice of antibiotics is further determined by the patient's allergies/weight and local hospital guidelines.

In order to sterilize the surgical field, we clean the surgical area first with 70% isopropyl alcohol, followed by 2% chlorhexidine gluconate/70% isopropyl alcohol formulation (ChloraPrep, CareFusion, San Diego, CA). This final scrub is allowed to dry for 3 minutes prior to draping. We then cover the entire surgical field with iodophor impregnated adhesive (Ioban, 3M) unless the patient is known to have an allergy to any of the mentioned components.

Postoperative antibiotic administration is controversial, with differing opinions regarding length of therapy. There are limited data to support antibiotic duration and class, although a previous retrospective study identified a significant decrease in infection rates with greater than 24 hours of postimplant antibiotics (mean of 7.6 days) [ ]. However, no prospective study has assessed this topic. In general, oral antibiotics tend to be prescribed to all neuromodulation implant patients for the duration of a trial, when the part of implanted device is externalized. The use of vancomycin powder has been suggested to reduce surgical site infections, although high quality studies are lacking [ ].

Skin erosion can occur with any implant, usually as a result of tissue strain and microtrauma or infection. General precautions to prevent erosion are often straightforward: avoid too superficial placement of hardware, close soft tissues over the devices in multiple layers, keep the profile of every implant as low as possible, make the depth of device implantation at the allowable maximum, and try to avoid placing hardware over hard surfaces (such as rib cage or iliac crest). The depth of device implantation may be limited due to telemetry or recharging limitations (1–2 cm for most rechargeable generators) or due to difficulty with pump refill if it is placed too deep in the fat or under abdominal fascia.

One has to avoid stretching the skin over the implanted device as this seems to be the most serious predisposing factor for skin hardware erosion [ ]. This means that the pocket for each device should be large enough to prevent tightness of tissues next to the implanted pump or generator. It is also recommended to avoid placing devices directly under incision lines [ , ], as this may increase the risk of wound dehiscence and create additional areas of discomfort and irritation during the postoperative period.

To avoid fractures and disconnections in neuromodulation devices, one has to be very accurate with each connection or stress point along the path of the device components. Overtightening of the holding screws may result in breaking contacts, while undertightening may result in loosening the connection and eventual pullout of inserted components. Use of strain-relief loops reduces the chance of electrode kinks that would eventually produce metal fatigue and fracture. Similarly, in placing extra loops of electrode or extension cable, one should try to create smooth loops rather than sharp turns and bends in each hardware component. While older anchors carried some risk of crushing the electrode if the holding suture was too strong, newer anchors are designed in a way that even very tight closure will not damage the external or internal electrode structure.