Lumbar Drains After Cardiac Surgery: Evidence-Based Solutions for Safe Management

Paraplegia, one of the most significant postoperative complications for patients undergoing thoracic and thoracoabdominal aortic aneurysm surgery, results from too little perfusion to the spinal nerves.1,2 Time with an aortic cross-clamp in place, hypothermia, and reperfusion injury are a few of the contributing factors. Wan et al3 reported that the risk of paraplegia with reconstructive surgery ranges between 5% and 40%, and potential improvements are achieved with distal aortic perfusion techniques and cerebrospinal fluid (CSF) drainage. Safi et al4 proposed a 16% to 31% risk of paraplegia with high-risk thoracoabdominal aortic aneurysm repair. This risk can increase to 50% with a cross-clamp in place longer than 45 minutes.4 The risk with thoracic endovascular repair is much lower at 1.5%, although that risk increases to 2.88% when protocols are not used, and to 5.6% when protocols are applied inconsistently.5,6Lumbar drainage catheters are placed in select patients to manage CSF and to reduce the incidence of paraplegia. Removal of CSF decreases intracranial pressure (ICP), improving perfusion to the spinal nerves. Nevertheless, risks are associated with the placement and use of lumbar drainage catheters; such risks include perforation of the external catheter with subsequent unintentional drainage, internal catheter fracture, infection, and too much or too little manual drainage because of human error. In addition, consistent identification and escalation of postoperative deficits, and appropriate intervention, are essential for ensuring successful patient outcomes.The cardiac surgery intensive care unit (CSICU) at the Virginia Commonwealth University Health System, in Richmond, Virginia, provides technologically advanced care to vulnerable adult patients undergoing cardiac surgery, including high-risk thoracic and thoracoabdominal aortic procedures. The CSICU collaborative practice team engaged in a process improvement project, creating evidence-based algorithms to ensure that patients requiring lumbar drainage catheters received safe, consistent postoperative care. These algorithms address communication pathways for escalating immediate and delayed neurological deficits (DNDs) and standardize CSF drainage rates. The collaborative practice team sought out new technology to mitigate risks related to human error and subsequently implemented an automated CSF management system. Nursing introduced a standardized dressing to prevent damage to the catheter. We also revised nursing competency processes related to the assembly of external drainage system components and system management. Because this was a performance improvement project, it did not require institutional review board approval.Thoracic and thoracoabdominal aneurysms are corrected through open surgical procedures, thoracic endovascular repair, and endovascular repair. The anesthesia team place the lumbar drainage catheter perioperatively or postoperatively; the practice varies among surgical teams. Recommendations differ in the literature as to the necessity and best timing of catheter placement.5,7,8 Drains placed perioperatively allow CSF to drain during the procedure, with the goal of preventing spinal cord ischemia.1,9-11 Catheter use is considered routine during open surgical procedures. With endovascular procedures, however, the risk of infection should be weighed against the risk of spinal cord ischemia. Cheung et al9 reported a 1.2% incidence of meningitis in patients whose surgery had been performed during cardiopulmonary bypass. Hussein et al8 reported an incidence of lumbar drain–related infections between 3% and 10% in neurosurgical patients. Therefore, surgical teams may avoid placing a catheter until the patient presents with a DND after the procedure. In a systematic review, Wong et al12 concluded that additional research is needed to determine recommendations for lumbar catheter placement in patients undergoing endovascular repair. In our organization, lumbar drain catheters are routinely placed perioperatively, with special attention focused on high-risk patients such as those who are likely to require ligation of collateral aortic vessels.The terms delayed neurological deficit and perioperative spinal cord ischemia describe the nature and timing of paraplegia symptoms. Deficits noted during the initial postoperative assessment are associated with perioperative spinal cord ischemia.3 A DND is defined as reduced patient movement, strength, sensation, ability to lift against gravity, or a combination of these. Such a deficit occurs after a normal initial postoperative assessment and is associated with microembolism, reperfusion injury, and elements of postoperative management.1,3,9,13 Postoperative management for the prevention of DNDs includes maintaining a mean arterial blood pressure between 90 and 100 mm Hg, and a goal ICP less than 10 mm Hg1,5,10 (a normal ICP ranges from 5 to 15 mm Hg14). Patients may develop a DND up to 48 hours after surgery— and sometimes much later—making detailed patient assessment imperative. Failure to recognize such delays, escalate them, and intervene may result in the patient permanently losing lower-extremity function. Immediate interventions for resolving a DND include increasing the mean arterial pressure goal to 110 mm Hg and the hemoglobin threshold to 10 g/dL, and hourly CSF drainage.1,4,11,15,16Lumbar drainage catheters are managed by using either manual or automated external CSF drainage systems (EDSs). A traditional manual EDS requires nurses to manipulate stopcocks and leveling components, and to carefully visually monitor drainage to ensure that the proper amount of CSF drains from the patient each hour. Technology such as the LiquoGuard automated CSF management system (Möller Medical) drains CSF on the basis of a programmed ICP goal.17-19 The system drains CSF when the patient’s ICP rises above the programmed pressure goal and ceases draining when the patient’s ICP falls below the goal.Risks related to management of the external lumbar drain components include unintentional overdrainage, underdrainage, and infection.15 Manual systems may not prevent accidental overdrainage, which may occur when a sudden change in patient position or intrathecal pressure increases the fluid drainage rate, or when components of the system become disconnected. Distraction of the end user during the drainage process is another common cause. Underdrainage may occur if the end user does not understand goals related to established drainage protocols. An automated CSF management system minimizes the risk of overdrainage and underdrainage using intraspinal pressure readings with redundant pressure sensors.17 Most importantly, it removes the need for hourly manual interventions by the end user. Risks with an automated system include incorrect programming of the hourly CSF removal rate. Catheter malfunctions, such as punctures in the external portion of the lumbar drainage catheter itself or separation of the catheter from the Luer spike, can also result in unintentional overdrainage,20 which may occur regardless of which EDS is used. Overdrainage can lead to intracranial hypotension. Catastrophic results from intracranial hypotension include subdural intracranial hemorrhage, cerebral herniation, and death.10 Current literature supports the use of protocols to prevent CSF overdrainage.5,10,11The CSICU clinical nurse specialist (CNS) and provider team synthesized our postoperative CSF drainage practices into evidence-based algorithms. The first algorithm focused on the manual EDS (Figure 1, available online only at www.ccnonline.org). After the CSICU adopted the external automated CSF management system, the team created another algorithm that became our primary method for managing lumbar drains (Figure 2, available online only at www.ccnonline.org). Both evidence-based algorithms guide cerebrospinal pressure management to achieve and maintain pressures less than 10 mm Hg, and they provide instructions for escalating the drainage rate should lumbar pressures increase.3-5,9,15,16,19 In addition, both algorithms define and address the development of DNDs, along with appropriate clinical interventions and provider notification.We created a small group of bedside nurse experts to provide didactic peer-to-peer education using a Microsoft PowerPoint presentation and hands-on return demonstration of the CSF management system. All newly hired nursing staff receive education through a unit onboarding class taught by unit experts. In practice, the surgeon communicates with the CNS when a surgical drain is planned, and the CNS then organizes additional just-in-time training provided at the point of care by the CNS, nurse experts, or a clinical representative of the automated CSF management system distributor. Hands-on return demonstration of equipment assembly and application of the algorithm are included in annual ongoing competencies for the unit nurses.The team created electronic orders that were added to the unit’s cardiac surgery order sets, providing nurses with instructions to follow the ordered algorithm. The algorithms are also imbedded into the cardiac surgery order sets in the form of Word documents; this inclusion ensures their presence in the patient’s electronic medical record. In collaboration with a pediatric CNS, the CNS created a documentation band in the nursing charting system that allows a registered nurse to input into the electronic medical record all data related to the automated CSF management system.Three episodes of accidental overdrainage related to catheter compromise prompted the CNS and her nursing student to standardize a method for securing the drainage catheter to prevent future incidents. This standardized method was implemented in the CSICU in October 2016. The securement method was designed to achieve 3 goals: (1) to support the rubber catheter with gauze and a soft plastic stent where the catheter connects to the Luer spike, thereby preventing bending, puncture, and separation; (2) to prevent skin injury from EDS components by using a foam insert and a silicone foam dressing; and (3) to protect the entire external portion of the rubber catheter from tears and punctures by using a transparent dressing. In February 2019 this securement method was merged with the use of a prefabricated transducer mold developed by the automated CSF management system manufacturer (Figure 3). This mold has further simplified our process for preventing unintentional drainage.Catheters remain in place for 96 hours with open repairs and for 72 hours with endovascular repairs, although they might be removed earlier at the discretion of the surgeon. Between January 2016 and October 2019, 22 patients required perioperative placement of a lumbar drainage catheter. All were managed postoperatively with the automated CSF management system and our CSF drainage algorithm. Two patients presented with preoperative neurological deficits and a third patient died from a cause unrelated to drain placement. All 3 were excluded from this analysis. All data are related to the inpatient time frame only. Three other patients experienced a postoperative neurological deficit. One of these deficits was identified during the immediate postoperative assessment. The deficits resolved in all 3 patients after the bedside nurse escalated the deficit and implemented interventions outlined in the algorithm. Of the remaining 16 patients, none developed DNDs. Consistent use of the method to secure lumbar drainage catheters since its implementation has allowed our unit to avoid any other episodes of unintentional drainage related to catheter tubing perforation. None of the patients developed infection.The overarching goal of this performance improvement intervention was to ensure that paraplegia events are minimized and interventions are congruent with data from current research. Of the patients included in this analysis, 100% experienced positive outcomes: in 1 an immediate postoperative deficit was resolved, in 2 a DND was resolved, and 16 never developed a DND. Since implementation of the automated CSF management system, no episodes of postoperative CFS overdrainage or underdrainage have occurred.The author thanks the dedicated team members of the cardiac surgery intensive care unit, who recognized this project as a priority for patient safety; Melissa Scott-Doney, MSN, RN, ACNP-BC, FNP-BC, CCRN-CSC, who initially helped the author spearhead this process; and Jill McGehee, MS, RN, CCRN-K, who collaborated on building the electronic medical record, and Maureen Flattery, MS, RN, ANP, for her guidance. The author also recognizes former nursing student Shipra Shah, RN, who helped develop a meaningful solution to a complicated subject.