Intrathecal drug delivery for the treatment of pain is an outgrowth of pioneering research into spinal cord mechanisms and receptors that modulate nociceptive signal transmission to and within the central nervous system. Starting in the 1970s, researchers identified the peptide substance P in small dorsal root ganglion cells and observed that opiates with a spinal action would produce selective analgesia. These findings as well as the observation that substance P release evoked from C fibers is blocked by morphine marked the beginning of the current trend toward focusing on the selective regulation of spinal afferent processing. Since then, there has been much research in this area, leading to widespread use of intrathecal therapy to treat chronic intractable pain. Although this therapy has proven to be extremely beneficial to many patients, it is also associated with high costs, high maintenance, and risks requiring careful patient selection. This chapter focuses on the patient selection, technologies used for chronic intrathecal drug delivery, system implantation, and troubleshooting.
Intraspinal Drug Delivery Techniques
There are essentially three methods of intraspinal drug delivery: (1) externalized systems, (2) partially externalized systems, and (3) totally implanted systems. The selection of the technique will depend on the goal of therapy, ranging from short-term intrathecal trialing to long-term therapy to treat chronic pain. Each technique has associated risks and costs, which are taken into account when selecting the technique.
Externalized Systems
An externalized intrathecal catheter is the most widely used technique for trialing intrathecal drugs. Such a system is designed for short-term (hours to days) use but has been successfully used for longer-term delivery in the terminally ill (weeks to months). The use of an externalized catheter raises the concern for a local infection at the catheter exit site or that the catheter may serve as a conduit for either epidural or intrathecal spread of infection. Using strict aseptic technique for catheter insertion can minimize the risk of insertion site infection, and the placement of a chlorhexidine impregnated catheter patch at the insertion site further reduces this risk ( Fig. 70.1 ). There is no evidence that systemic antibiotics prevent these infections. In the terminally ill patient, the risk-benefit ratio may favor the use of an externalized catheter for intrathecal drug delivery. Some practitioners tunnel these catheters under the skin for some distance to provide stability and reduce the likelihood of inadvertent catheter withdrawal. Currently, the U.S. Food and Drug Administration (FDA) has not approved any externalized intrathecal catheters, and it is common to use epidural or lumbar drain catheters for this purpose.
Partially Externalized Systems
Partially externalized systems are those in which the intrathecal catheter is tunneled and connected to a subcutaneous port. A needle then accesses the port percutaneously and medication is delivered with tubing connected to an external pump. Compared to an externalized system, this permits greater patient freedom of movement and reduces the risk of inadvertent catheter removal. Partially externalized systems are intended for single bolus injections. There continues to be a risk of colonization and infection of the port pocket with prolonged or repeated needle access. These systems have limited application for long-term, continuous intrathecal drug delivery.
Implanted Intrathecal Infusion Pumps
Totally implanted systems are those with both the catheter and pump system completely internalized by surgical implantation. They have the advantages of a lower risk of infection, potential for long-term use, and much greater freedom of movement for the patient. The totally implanted systems are more expensive and require a more invasive surgical procedure than the externalized or partially externalized systems. Although it is generally recommended that implanted pumps be reserved for patients with a life expectancy of greater than 6 months, this therapy may improve quality of life in patients with life expectancy as low as 2 to 3 months. However, patients with a life expectancy of less than 6 months should be carefully selected.
There are two main types of intrathecal infusion systems, fixed-rate and variable-rate programmable pumps. All implanted pumps have two percutaneous access ports. One port is for refilling the drug reservoir, and there is a separate intrathecal catheter access port for bolus injection through, or aspiration from, the catheter. The infusion systems are generally implanted in the lower abdominal subcutaneous fat and connected to a catheter, which is tunneled subcutaneously around the abdominal wall into the intrathecal space. However, the pumps may also be implanted subcutaneously in the lower back in rare, selected patients (see the discussion that follows). The septum puncture life of the percutaneous refill port is very high: 500 punctures for the most commonly used, Medronic Synchromed II, pump. It would be unlikely to reach this number of punctures during the life of the pump (7 years for the Synchromed II).
The first intrathecal pump, manufactured by Shiley Infusaid, became commercially available in 1982. The Infusaid and other early pumps were nonprogrammable fixed-rate infusion systems. These pumps were simple devices without batteries, in which pressurized gas pushes the drug from a reservoir through a valve and into the catheter. The chief advantage of these pumps is simplicity of design and lack of batteries. Unfortunately, they are unable to vary the dose delivered without changing the concentration of drug in the reservoir. Currently the Codman 3000 ( Fig. 70.2 ) is the only fixed-rate infusion pump on the market in the United States and is no longer used with significant frequency. The pump is available in three sizes (16, 30, and 50 mL reservoir), and each size is available in three flow rates. The low and medium flow rates are 0.5 and 1 mL/day, respectively, for all sizes. The high flow rates are 1.3, 1.7, and 3.4 mL/day, respectively, for the 16, 30, and 50 mL pumps. The flow rate is controlled by a flow restrictor. Flow rates may vary for different drugs, drug combinations, and especially with high drug concentrations, based on differences in viscosity. In addition, changes in body temperature and atmospheric pressure will change flow rate by affecting the gas pressure within the system.
Programmable pumps, on the other hand, allow the clinician to vary infusion rates to increase or decrease the dose without changing the concentration of the drug or drugs in the pump. The original programmable pump was the Medtronic Synchromed, first introduced in 1991. More recent programmable pumps such as the Medtronic Synchromed II ( Fig. 70.3 ) allow patient-controlled intrathecal bolus dosing via a remote control device. The Synchromed II uses a peristaltic roller system ( Fig. 70.4 ) to move the drug from the reservoir to the implanted intrathecal catheter. The Synchromed II is available in a 20-mL and a 40-mL reservoir size and is FDA approved for baclofen, morphine, and ziconotide.
In 2012, the FDA gave marketing approval for the Flowonix Prometra implantable pump ( Fig. 70.5 ). This device uses a pressurized gas chamber as the driving force. Unlike the older fixed-rate infusion pumps, a programmable flow-metering valve allows for a variable drug delivery rate. The Prometra pump’s low energy requirement is its main advantage. Since the batteries are used to control the electronics only, not to pump the drug, the battery has a 10-year life span before the need to surgically replace the pump. The Prometra is available in a 20-mL reservoir and is FDA approved for morphine. The issues with variable flow resulting from viscosity, temperature, and residual reservoir volume have been addressed by the advanced flow-metering valve.
The MedStream Programmable Infusion System is another implantable device currently FDA approved only for the delivery of intrathecal baclofen ( Fig. 70.6 ). This implantable drug delivery system offers improved catheter technology and pump durability. Its intrathecal catheter has been designed to resist kinking and tearing (similar to the newest Medtronic catheters). The MedStream uses compressed gas as the driving force and a ceramic drive flow valve system to maintain the infusion rate. The pump has either a 20-mL or a 40-mL reservoir and measures 76 mm by 21.6 mm (20 mL) or 28.2 mm (40 mL).
The Medallion is an implantable drug delivery system currently in clinical trials. The Medallion system offers safety improvements with the use of a negative pressure reservoir. The negative pressure draws medication from the syringe during pump refills rather than requiring positive pressure from the syringe plunger. This reduces the risk of inadvertent injection into the pump pocket rather than the pump reservoir. The Medallion will be available in either a 20- or a 40-mL reservoir.
Patient Selection
Pain Type
Intrathecal drug delivery systems (IDDSs) are reserved for treating severe and refractory pain in two types of patients: those with cancer pain and those with noncancer pain who have failed more conservative therapies. The cost and invasive nature of IDDSs necessitate a careful evaluation to determine whether each individual patient is appropriate and likely to achieve benefit. A thorough assessment includes an evaluation of pain-generating pathology, medical comorbidities, psychosocial issues, patient compliance history, economic and health care coverage, and anatomic and technical considerations.
In advanced stages of cancer, the prevalence of pain exceeds 60%. One third of those patients report moderate to severe pain. Comprehensive medical management (CMM) can significantly reduce pain and manage the adverse effects of opioids and other analgesics. Despite aggressive therapy, however, some patients continue to endure severe pain. Others are unable to tolerate systemic opioids due to toxicity. A 2002 multicenter randomized trial compared CMM alone versus CMM plus an IDDS. They found that cancer patients receiving CMM plus an IDDS had reduced pain, fewer drug toxicities, and prolonged survival compared to controls. Patients with significant toxicity or with refractory pain that has been inadequately controlled by the systemic administration of opioids and other analgesics are potential candidates for an IDDS. Intrathecal therapy may be especially valuable in patients with neuropathic pain from tumor invasion of neural plexuses, unstable pathologic fractures, painful impending spinal cord compression, and visceral tumors with or without autonomic dysfunction that results in gut dysmotility.
In patients with a life expectancy of less than 2 to 3 months, a percutaneous catheter and pump is generally used. These patients may be discharged from the hospital with hospice care or home-health services for the care of the catheter and infusion if appropriate services are locally available. When life expectancy is greater than 3 to 6 months, an implanted system is generally recommended.
For the patient with chronic noncancer pain, a comprehensive evaluation to establish the etiology of the pain should be performed as part of the decision process regarding whether an intrathecal pump is appropriate. The evaluation should include a detailed history, physical examination, radiologic studies, and assessment of prior therapies and their outcomes. Many diagnoses and painful conditions may be amenable to intrathecal drug delivery in certain cases. These include neuropathic syndromes such as spinal cord injury, diabetic peripheral neuropathy, complex regional pain syndrome, postherpetic neuralgia, and phantom limb pain. Additionally, mixed nociceptive and neuropathic conditions such as postlaminectomy pain syndrome or chronic pancreatitis as well as nociceptive pain conditions such as chronic vertebral compression fractures and severe spinal degeneration may be appropriate for an IDDS in some circumstances.
Implanted drug delivery therapy should be reserved for those patients who are refractory from more conservative treatment. A stepwise approach starting with the least invasive and costly therapies is recommended. This may begin with exercise programs, relaxation techniques, and over-the-counter analgesics. Subsequent steps include adjuvant analgesics, muscle relaxants, physical therapy, and psychological therapies. More invasive therapies include neuraxial somatic and sympathetic blocks and spinal cord stimulation. More aggressive medication therapy can then be introduced with oral opioids. The top tier of this stepwise treatment pyramid is reserved for the most invasive therapies, including intrathecal drug delivery and neurodestructive procedures. There are few controlled trials of intrathecal therapy for chronic noncancer pain, but observational studies suggest such therapy can provide pain relief in certain patients who have failed to achieve relief with more conservative treatments. Determining the efficacy of long-term intrathecal drug therapy remains one of the greatest challenges in chronic noncancer pain management. One of the reasons for the lack of randomized controlled trials is that the high cost and invasiveness of such treatment pose ethical challenges in designing a double-blind, placebo-controlled study. Therefore, one can only rely on the retrospective reviews of patients receiving this therapy and attempt to make conclusions on the efficacy of this treatment. Many studies have attempted to determine the efficacy of long-term intrathecal morphine therapy based on classification of the pain. However, a review of these articles reveals large discrepancies in pain classification between studies. Because pain is often multifactorial (e.g., failed back syndrome versus HIV neuropathy versus complex regional pain syndrome), it is difficult to determine what pain syndromes are the most responsive to this therapy.
Patient Comorbidities
The aging chronic pain patient may experience some of the same age-related comorbidities as any individual. Many of these comorbid conditions can complicate the implantation and use of an intrathecal pump. These medical issues and their potential interaction with intrathecal therapy must be considered during the patient selection process.
Diabetes Mellitus
Diabetes mellitus affects over 8% of the U.S. population and nearly 27% of those over the age of 65. Patients with diabetes experience poor wound healing and suffer from an increased incidence of surgical site infections. Implantable devices pose a special risk of surgical wound infections. With implantable intrathecal pumps, wound infections have been found to be the most common device-related complication. Patients with diabetes, especially if blood glucose control is poor, should be counseled on the increased risk, and implanters should maintain extra vigilance for signs of site infections.
Anticoagulant Therapy
Anticoagulant therapy has become increasingly common as more indications for anticoagulation such as coronary artery stent placement have arisen. Additionally, newer anticoagulant therapies have improved management and dosing of anticoagulants. Patients on anticoagulant therapy are at increased risk of epidural hematoma formation after placement or removal of epidural or spinal catheters. The American Society of Regional Anesthesia publishes guidelines regarding the performance of neuraxial injection and catheter placement with concomitant administration of anticoagulants ( Fig. 70.7 ). These guidelines do not specifically address the permanent implantation of intrathecal drug delivery systems. Scattered case reports of anticoagulant therapy restarted after implantation of intrathecal pumps offered no clear evidence for or against the safe use of intrathecal pumps in these patients.
Infections
Any active infection puts the patient at high risk of seeding the pump with bacteria. The resulting pump pocket infection is virtually impossible to clear completely and will require explanting the entire pump and catheter system. Any active or chronic bacterial or fungal infectious process, particularly with noncancer pain patients, should be regarded as an absolute contraindication to implantation.
Pulmonary Disease
More than 18 million Americans have obstructive sleep apnea (OSA). In addition to the risks of carbon dioxide retention and both systemic and pulmonary hypertension, the OSA patient has increased sensitivity to the respiratory depressant effects of opioids. There are no studies of intrathecal opioid use in OSA patients. However, case reports suggest that chronic use of oral opioids in those with OSA may increase apnea duration and hypoxia severity. Chronic obstructive pulmonary disease (COPD) also results in carbon dioxide retention and increased risk of respiratory depression with administration of opioids. A retrospective case-controlled study of postoperative COPD patients administered opioids found an increased risk of respiratory events (odds ratio [OR] 5.09; 95% confidence interval [CI]) compared to patients without chronic lung disease. Most of these events occurred within 24 hours of surgery. The long-term risk of intrathecal opioid administration in patients with severe pulmonary disease is unknown. However, known risks of perioperative opioid administration in this population warrant careful monitoring during trialing as well as after implantation and initiation of therapy.
Psychological Screening
The Centers for Medicare and Medicaid Services require a psychological assessment prior to the implantation of spinal cord stimulator systems. No such requirement exists for IDDS. Although not required, it is widely accepted that a pre-procedure psychological assessment is warranted and is at least as important for appropriate patient selection for an IDDS as spinal cord stimulation. Ideally, a psychological evaluation would aid in predicting positive response to an implantable therapy. In reality, there is limited evidence regarding the efficacy of psychological screening to predict long-term success. Much of the literature in this area concerns implantable spinal cord stimulators, and it is from this body of literature that guidelines regarding the use of psychological evaluation for IDDS have been extrapolated.
The presence of suicidal or homicidal ideation, uncontrolled depression, active psychosis, severe cognitive deficits, and active drug abuse are absolute contraindications to implantation. In a study by Celestin and colleagues, certain psychological factors and traits were associated with poor analgesic outcome after implantation. These include somatization, depression, anxiety, and poor coping skills. However, a study by Doleys and Brown found that those with mild abnormality on the Minnesota Multiphasic Personality Inventory (MMPI) personality profiles had better long-term analgesic efficacy than those with more normal scores. Consensus guidelines have recommended the use of some form of psychological assessment in the selection process for IDDS candidates but caution against using the results to categorically rule out implantation in patients with psychological issues.
The psychological evaluation should be used to identify factors that may affect the outcome of a trial or long-term therapy, either positively or negatively. In addition to identifying severe psychopathology, the screening examination can evaluate for cognitive deficits, coping mechanisms, social supports, patient expectations, and comprehension of treatment goals that may affect therapy. This screening may also identify the need for psychological treatment that can improve the outcome of therapy. The psychologist should not be asked to give “clearance” but rather act as an additional source of information as to whether IDDS therapy is appropriate, the patient has rational expectations, and this therapy is likely to succeed for an individual patient.
Trialing
Some type of therapy trial prior to permanent system implantation has been commonly performed and is often required by insurance (Medicare). Trial options include a single bolus intrathecal injection, multiple bolus injections, or a continuous infusion via percutaneous intrathecal or epidural catheter. Unfortunately, there are no commonly accepted guidelines regarding how to perform a trial. Further, the data are limited, suggesting that trialing can accurately predict efficacy after implantation. An expert consensus statement regarding guidelines for selection of patients for noncancer pain recently recommended against the requirement of a pre-implant trial because of the trial’s unsubstantiated predictive value.
The selection of a trialing method for noncancer pain will be made largely based on physician preference along with payer requirements. The majority of trials are performed in a hospital inpatient setting via epidural or intrathecal infusion. The advantage of epidural infusion is simplicity of catheter placement and a much lower risk of postdural puncture headache (PDPH). The development of a PDPH during trial may impair the patient’s ability to determine the efficacy of the trial. An intrathecal trial more closely mimics the permanent implant conditions. Both of these options are more costly than single or multiple intrathecal injections. Although injections, whether single or multiple, provide simplicity with a low incidence of PDPH, they fail to mimic the chronic infusion conditions of a permanent implant.
Trials for cancer pain are performed in a similar fashion. However, the lack of prognostic evidence for trials calls into question their value in patients with rapidly progressive disease near the end of life. According to one consensus guide to intrathecal therapy for cancer pain, the risks associated with the trial and delay of therapy may not outweigh the alleged prognostic benefits. However, other experts have recommended the use of a trial in the event of cancer pain in order to give the patient and physician the opportunity to evaluate potential responses as well as possible side effects.
Trials of baclofen therapy for spasticity are considerably more straightforward than those for cancer and noncancer pain. There are clear, objective outcome measures of spasticity that can be assessed to determine response. The commonly used criterion for a successful trial is a 2-point reduction on the Modified Ashworth Scale for grading spasticity ( Fig. 70.8 ). Patients may also demonstrate improvements in joint range of motion.
The baclofen trial consists of a bolus intrathecal injection of 25 to 50 mcg of baclofen and is performed in an outpatient setting. The onset of effect occurs in 1 to 3 hours and may last 6 to 8 hours. The patient is then evaluated with serial examinations to determine response. If the patient fails to respond, the trial may be repeated on a subsequent day at a higher dose.
Most patients that reach the level of intrathecal therapy to treat chronic pain are on some amount of systemic opioids. It is controversial whether to detoxify patients off of the opioids prior to trialing. There are many advantages to detoxifying the patient prior to a trial. First, the drug holiday will provide the opportunity to observe the patient without the systemic opioids and allows a better assessment of addiction. There is often a gray area between addiction and treating chronic pain, as there is frequently some overlap between treating valid chronic pain complaints and the presence of addiction that needs to be assessed prior to implant. Second, the drug holiday will reduce tolerance, allowing for a better assessment of response to intrathecal therapy during the trial. Third, weaning patients off of the opioid prior to implant will likely result in lower doses of intrathecal opioids in the long term. Patients that are titrated up on intrathecal opioids as they are being weaned from systemic opioids often experience withdrawal symptoms resulting in overly aggressive increases in the intrathecal dose. One disadvantage to weaning the systemic opioid prior to trialing includes the pain and suffering that the patient may experience during the drug holiday. There is currently no clear consensus on whether to wean or maintain the systemic opioid prior to trialing. The decision should be based on a discussion between the physician and patient. An addictionology consult may be considered for patients who are unable to wean from the opioids because of severe withdrawal symptoms.
Implantation
Location of Pump Placement
There are several considerations when choosing where to locate the pump. Patient comfort is of primary importance. The pump must be located away from bony prominences, including the iliac crest and ribs, to avoid rubbing. The amount of subcutaneous fat is also important. In the average-sized patient there may be too little space and tissue in the flank, whereas in the morbidly obese patient there may be too much subcutaneous fat in the abdomen. Have the patient sit up to mark the site to ensure that the pump will not lie below the belt line. Discuss the patient’s sleeping habits to determine if the patient is a side sleeper. If so, then place the pump in the contralateral abdomen. In the majority of cases, the most appropriate location for pump placement is in either the left or right lower quadrant of the abdomen. Unlike spinal cord stimulator pulse generators, the size of intrathecal pumps makes placement in the flank or upper gluteal region uncomfortable for average-sized patients. Placement of the pump pocket in the abdomen necessitates positioning the patient in lateral decubitus for the surgical procedure. Proper attention to positioning is necessary to ensure that both the pump site and catheter insertion site are accessible without breaking scrub and repositioning.
Patients who have had multiple abdominal surgeries may pose a challenge in locating the pump pocket. If an abdominal site is impossible, a smaller 20-mL pump may be placed in the superior gluteal region. In larger patients, there may also be room to place the pump in the flank midway between the 12th rib and the iliac crest. A consultation with a plastic surgeon may be helpful in patients with limited implant site options.
When re-implanting a pump after previous pump site infection, it is best to use a new location such as the contralateral abdomen to minimize re-infection risk. Placement of a pump into scar tissue is not recommended, as the blood supply can be erratic, limiting the flow of antibiotics to the area during implantation. This is evident with pump re-implantations for battery failure in which the capsule surrounding the pump has minimal bleeding when opened. In these cases, meticulous care must be taken with sterile technique to minimize the risk of infection with the re-implant.
Anesthesia for Pump Implantation
Intrathecal pump implantation can be performed under local anesthesia with sedation, spinal anesthesia, or general anesthesia. The anesthesia method of choice is to place the catheter under local anesthesia followed by either a spinal anesthetic or local anesthetic infiltration for catheter anchoring, tunneling, and pump placement. This method allows for communication with the patient during catheter placement to minimize the risk of nervous system injury. It is also acceptable to use general anesthesia, but the risks must be disclosed to the patient. The choice of anesthesia will depend on surgeon and patient preference. Nonsurgically trained pain physicians may be more comfortable with an awake patient under local anesthesia who can communicate with the implanter during needle and catheter placement and allow for the assessment of neurologic status. Orthopedic and neurosurgeons may be more accustomed to a patient under general anesthesia. In addition, general anesthesia may be more comfortable and better tolerated by some patients. If implanting under spinal anesthesia, the implanter will first place the Tuohy needle into the intrathecal space under local anesthesia. The implanter can then thread the intrathecal catheter through the Tuohy, communicating with the patient to make sure there is no pain or paresthesias. Once the catheter is confirmed to be in the intrathecal space radiographically and by free flow of cerebrospinal fluid (CSF), injection of 20 to 30 mg of preservative-free lidocaine will achieve adequate spinal anesthesia for both the posterior midline incision and abdominal pump pocket incision.
Patient Positioning
Most commonly the pump will be located in the abdomen, which necessitates either a lateral decubitus position or repositioning of the patient from prone to supine midway through the surgery. Most implanters use the lateral decubitus position to avoid the time delay and potential infection risk of repositioning. Careful positioning with pressure points padded and an axillary roll in place is necessary for patient comfort and to avoid intraoperative nerve compression injuries. The use of sticky rolls to support the upper abdomen or posterior thorax may assist in positioning the patient in a true lateral decubitus. Taking the time to ensure the patient is well supported and the sagittal axis is perpendicular to the operating room table will ensure a true anterior-posterior (AP) view with fluoroscopy. Position the patient’s arms to avoid interfering with fluoroscopic visualization. As stated before, it is critical to mark the pump implantation site with the patient in the sitting position because tissue will shift when in the lateral position. Marking the patient while in the lateral position risks having the pump end up in an uncomfortable location.
Surgical Technique
Sterile Technique
Good sterile technique and pre-operative antibiotics are the two most important controllable factors in preventing wound infections. In a study of primary implantations and revisions of intrathecal baclofen pumps in pediatric and adult patients, an infection rate of 4% to 5% was found. In more than half of the infections, the primary organism was Staphylococcus aureus. In a multicenter prospective study of intrathecal pump implantations for both pain and spasticity, Follett found that the most common complication was infection, with an incidence of 7%. Most surgical site infections in clean surgeries are due to skin microorganisms. Careful cleansing of both the surgeon’s hands and the patient’s skin at the operative site with either a chlorhexidine-alcohol or povidone-iodine scrub reduces skin bacterial counts. Scrubs of 2 to 3 minutes are superior to shorter scrub times. Several studies have suggested that chlorhexidine-alcohol scrubs may be superior to povidone-iodine scrubs. Body hair should be removed only if excessive, and removal should be performed with clippers instead of shaving. Shaving results in microtrauma of the skin, resulting in the inability of the antiseptics to reach the cracks and crevices that may be hiding bacteria. Use of chlorhexidine skin preparation of the surgical site followed by the use of povidone-iodine-impregnated adhesive skin drapes is recommended, as this produces another barrier between the skin flora and wound. If using these drapes, allow extra time for the skin prep to dry so that the adhesive will stick firmly to the skin. Adherence to the patient’s skin is necessary for the antiseptic effect. Surgeons and assistants should double glove to reduce the infection risk associated with glove puncture. Prophylactic systemic antibiotics against the most likely pathogens (skin flora) should be administered within 60 minutes of skin incision. If patient allergy to cephalosporins and penicillin precludes their use, one recommended alternative is clindamycin. There is no evidence supporting the postoperative administration of oral antibiotics to prevent wound infections.
Additional practices that may reduce the risk of infection include (1) soaking the pump in a mixture of saline and povidone-iodine solution, (2) packing the wounds with povidone-iodine-soaked sponges for a few minutes, (3) painting the edge of the wound with povidone-iodine and redraping the edge of the wound prior to placing the pump into the wound, and (4) regloving after surgical dissection is complete and before handling the pump. These practices may reduce the risk of skin flora colonization of the pump and pocket.
Catheter Placement and Anchoring
Prior to needle insertion, the spinous processes of the levels of insertion should be marked on the skin. This will identify the midline incision to be used for dissection and anchoring. Insert the introducer needle using a shallow-angle, paramedian oblique technique ( Fig. 70.9 ). An angle of approximately 30 degrees off of the spine is ideal. For an average-sized patient, the skin entry point will be 1 to 2 cm lateral to the midline on the side of the intended pump pocket, and 1 to 1.5 vertebral levels caudad to the targeted interlaminar entry site. The typical entry site will range from the L2 to L4 interlaminar spaces unless anatomic considerations or previous surgery dictates otherwise.
