Reducing Surgical Complications in Spine Patients Through a Medication Management Support Program

Study Design.

A hospital-wide medication management program was implemented to ensure that high-risk patients would systematically pause antiplatelet and anticoagulant medications. We analyzed complications before and during the implementation of this program.

Objective.

The goal of the study was to determine if a medication management support program was effective for reducing perioperative complications, including hemorrhage, myocardial infarction, stroke, pulmonary embolism, and deep vein thrombosis.

Data and Methods.

Using data from the National Surgical Quality Improvement Program database, we examined the presence of 5 complications before and during the implementation of a medication management support program. There were 9732 patients in the clinic population who underwent elective spine surgery between 2011 and 2020 and were included in this analysis. Of those 9732 patients, 7205 had surgery before the introduction of the program, whereas 2527 had surgery at some point after the program was introduced. We conducted a series of Pearson’s χ² tests to determine the relative frequencies of the complications before and during the program.

Results.

Results showed that during the implementation of the program, patients were relatively less likely to experience hemorrhage (3.16% vs. 1.11%; P<0.001). The reductions in thrombotic complications were clinically significant: myocardial infarction (0.12% vs. 0.00%), stroke (0.10% vs. 0.04%), pulmonary embolism (0.33% vs. 0.28%), and deep vein thrombosis (0.36% vs. 0.28%). These P values ranged from P=0.08 for myocardial infarction to P=0.67 for pulmonary embolism.

Conclusions.

The use of this medication management support program appears effective for reducing the need for blood transfusions and thrombotic complications. While promising, the results should be interpreted with caution as we do not know whether this type of program will be effective for other hospital systems.

Spine surgeons are under increasing pressure to incorporate findings from the ever-expanding literature about their field while providing competent and contemporary care for their patients. Research shows that standardizing practice can facilitate effective care and mitigate patient risk.1,2 However, in emerging areas of spine research, care can be unstandardized, especially among surgeons working in different contexts and at different stages of their careers.

One key goal of spine surgery is to reduce the risk of intraoperative hemorrhage that can result from patients taking anticoagulants or antiplatelets.3 These medications can be paused to reduce the risk of hemorrhage during surgery,3 but pausing them presents new risks—risks for thrombotic complications after surgery.4 Thus, surgeons are faced with having to delicately balance the risk of hemorrhage during surgery with the risk of clotting complications after surgery.4 Data to support best practices in this area are still emerging, but evidence suggests that a pause in these medications preoperatively may reduce the likelihood of hemorrhage, and the resumption of those medications postoperatively may reduce the likelihood of a clotting complication.5

In 2010, little was known about how to balance these 2 sets of risks. In fact, a systematic review of the literature6 concluded, “We found no studies that attempted to define a safe perioperative window of nonanticoagulation in spine surgery patients…” (pp. S121). By 2022, however, growing evidence suggested possible medical benefits to pausing certain anticoagulants and antiplatelets perioperatively.5 Incorporating these types of new findings into practice can be challenging. One solution for a hospital system is to create a medication management support program to help implement emerging standards of care and to ensure that patients adhere to the most current recommendations. However, to our knowledge, no study has evaluated whether or not this type of medication support program reduces patient complications in a spine surgery practice. Our health care system implemented such a program as it relates to this emerging literature of perioperative anticoagulants and antiplatelets. We describe the medication management program in the Method.

METHOD

Participants

All patient data came from the National Surgical Quality Improvement Program database.7 We included 9732 patients who underwent elective spine surgery between 2011 and 2020 at a level 1 mid-Atlantic trauma center with both academic and private practitioners. These patients had elective cervical, thoracic, or lumbar spine surgeries. We excluded patients who had nonelective surgeries. Of those included, 7205 had surgery before the implementation of the medication management support program, whereas 2527 had surgery at some point after the program was introduced to the practice in February of 2018. Across the 2 groups (pre- and during the medication management program), participants differed somewhat demographically. Though patients in both groups consisted of the same percentages of males and females (both 49% male, 51% female), the pre-program group was slightly younger (M Age=56.72 y, SD=13.66) than the program group (M Age=59.40 y, SD=13.25; t(9730)=−8.54, P<0.001). Also the pre-program group had lower BMI scores (M=31.23, SD=7.87) than the program group (M BMI=31.94, SD=8.23, t(9730)=−3.88, P<0.001). Compared with the program group, the pre-program group was more likely to be White (83% vs. 80%), less likely to be Black or African American (15% vs. 21%), less likely to be Asian (0.1% vs. 1%), less likely to be unknown/not reported (0% vs. 2%), similarly unlikely to be American Indian or Alaskan Native (0% vs. 0%), and similarly unlikely to be another race (0% vs. 0%). Also, 96% were non-Hispanic in both the pre-program and program groups.

Procedures

We analyzed records that contained diagnoses of each of the 5 complications: hemorrhage, myocardial infarction, stroke, pulmonary embolism, and deep vein thrombosis.

Each diagnosis was made by a board-certified physician applying standard American Medical Association assessment procedures. Hemorrhage was operationalized as the number of blood transfusions up to 72 hours post-incision. Myocardial infarction was identified by elevated troponin levels that fell within the 99% percentile whether the patient was symptomatic or not. Stroke was diagnosed through brain imaging. Pulmonary embolism was diagnosed perioperatively using a physical exam, blood tests, medical imaging such as computed tomography pulmonary angiography, or a pulmonary ventilation/perfusion scan. Deep vein thrombosis was diagnosed using ultrasound, venography, or magnetic resonance imaging/venography. The Pearson’s χ² tests were conducted using the Statistical Package for Social Sciences (SPSS) Version 26.8 These tests consider whether observed frequencies differ from what would be expected by chance in a contingency table. To confirm that the 2 groups were similar in terms of their surgical complexity, we investigated the number of nonprimary procedures performed during surgery through a t-test.

Medication Management Support Program

The purpose of the medication management support program was to optimize care and minimize complications for spine surgery patients by better managing medication protocols. As part of this program, before surgery, patients were evaluated for surgical risk by Perioperative Evaluation and Preparation nurses, and all patient medications were documented. Patients on certain anticoagulants (Warfarin, Dabigatran Etexilate, Rivaroxaban, Apixaban, and Edoxaban) or antiplatelet medications (Aspirin, Clopidogrel, Ticagrelor, and Prasugrel), were considered high-risk patients and referred to the PeriOperative Evaluation & Treatment Service (POETS) team before surgery. The POETS team visited with the patient either in person or virtually, depending on the patient’s preference, to inform the patient on how to minimize risk. POETS advised the patients to stop taking Prasugrel and Ticagrelor 7 days before their operation, Clopidogrel and Warfarin 5 days before their operation, and Rivaroxaban, Apixaban, Edoxaban, and Dabigatran Etexilate 2 days before surgery. POETS also advised patients to reduce their dose of Aspirin to 81 mg daily. Finally, surgeons were reminded through email that patients with kidney disease or higher thrombotic risk should seek a POETS consultation. Postoperatively, surgeons consulted with POETS through PowerChart if they wanted assistance with any medical or medication management. Surgeons were provided the flexibility to reinstitute anticoagulant medication, depending upon complications that may have arisen during surgery or recovery. If a patient refused the POETS visit, the POETS team called the surgeon’s office and alerted the office of the patient’s decision to help determine whether follow-up with the patient was required. To help implement this program of care, a member of the POETS team sent an email to surgeons about the available support. That email detailed all of the information above. In that same email, surgeons were asked to eliminate any existing anticoagulation handouts in their offices and replace them with a handout that detailed all of the medical information described above. Finally, surgeons were reminded that this type of standard practice helps to mitigate risk and promote successful outcomes for patients. Surgeons were given an opportunity to contact the POETS office with questions. See the flowchart in Figure 1 for parsimony.

Figure 1.

Medication management program flowchart. POETS advised patients to pause Prasugrel and Ticagrelor 7 days before their operation, Clopidogrel and Warfarin 5 days before their operation, and Rivaroxaban, Apixaban, Edoxaban, and Dabigatran Etexilate 2 days before surgery. POETS also advised patients to reduce their dose of Aspirin to 81 mg daily. POETS indicates PeriOperative Evaluation and Treatment Service.

RESULTS

Findings indicated that the number of patient blood transfusions within 72 hours from the incision was lower after the implementation of this medication management support program (see Fig. 2). Before the implementation of the program, 3.16% (n=228/7205) of patients required transfusions. During the implementation of the program, 1.11% (n=28/2527) required transfusions. This difference was statistically significant, χ²(1)=30.89, P<0.001.

Figure 2.

Standardizing perioperative management of antiplatelets and anticoagulants reduces the percentage of patients needing transfusions within 72 hours of incision. Note. Pearson χ2 (1)=30.89, P<0.001.

Findings further indicated that thrombotic complications were lower during the program, as indicated by a lower percentage of patients with myocardial infarction, stroke, pulmonary embolism, and deep vein thrombosis (see Fig. 3). Before the introduction of the program, myocardial infarction occurred at a rate of 0.12% (9/7205) versus 0.00% (0/2527) during the program. Stroke occurred at a rate of 0.10% (7/7205) before the program versus 0.04% (1/2527) during the program. Pulmonary embolism occurred at a rate of 0.33% (24/7205) before the program versus 0.28% (7/2527) during the program. Also, deep vein thrombosis occurred at a rate of 0.36% (26/7205) before the program versus 0.28% (7/2527) during the program. We present these latter findings descriptively. The reductions in thrombotic complications were clinically significant, and all trended in the same positive direction, but Pearson’s χ² tests did not reach statistical significance; the P values ranged from P=0.08 for myocardial infarction to P=0.67 for pulmonary embolism.

Figure 3.

Standardizing perioperative management of antiplatelets and anticoagulants reduces the percentage of patients with thrombotic complications postoperatively.

To determine if the 2 groups were roughly similar in terms of their surgical complexity, we investigated the number of nonprimary procedures performed during surgery. Results showed that the 2 groups did not differ in complexity. Before the implementation of the program, patients had on average of 2.47 (SD=1.98) nonprimary procedures, whereas during the program, patients had on average of 2.49 (SD=2.25) nonprimary procedures, t (3973)=−0.51, P=0.63, equal variances could not be assumed.

DISCUSSION

The present study was designed to determine whether using a medication management support program for anticoagulant and antiplatelet medications would reduce perioperative complications in spine surgery patients. We examined the incidence of 5 complications (hemorrhage, myocardial infarction, stroke, pulmonary embolism, and deep vein thrombosis) in a large number of patients who underwent spine surgery before and after the implementation of a medication management support program. Results showed that the program may have been effective in reducing hemorrhage (3.16% vs. 1.11%), myocardial infarction (0.12 vs. 0.00%), stroke (0.10% vs. 0.04%), pulmonary embolism (0.33% vs. 0.28%), and deep vein thrombosis (0.36% vs. 0.28%) in patients following surgery.

These findings should be interpreted with caution and are consistent with existing research showing that pausing anticoagulants and antiplatelets reduces hemorrhage during surgery3 and that resuming these medications decreases clotting complications after surgery3 The reduction in transfusions that occurred during the implementation of the medicine management support program was rather pronounced, cutting the need for a transfusion in about half, from just over 3% to just over 1%. The reduction in myocardial infarction was clinically very meaningful. Whereas 9 people out of 7205 suffered from myocardial infarction before the program, 0 out of 2527 suffered from myocardial infarction during the program. Though the incidence of myocardial infarction was already low before the implementation of the program, it was further reduced during the program, thus saving money and potentially saving lives.

The literature focusing on the importance of perioperative management of anticoagulant and antiplatelet medications continues to evolve 3–6 and currently suggests that it is important to correctly manage these medications before and after surgery. Yet, there are a host of challenges to implementing and ensuring adherence to evolving medication protocols. This medication management support program reduced complications from spine surgery, suggesting that this type of support program may be a valuable tool for care providers and managers of health care systems to improve patient outcomes. We describe this intervention in some detail in the Method section so that any health care system can choose to implement this program with the goal of reducing spine surgery complications and improving patient care.

Although the findings from this study may be useful and informative, this study is not without limitations. The sample size of patients who had standardized perioperative care regarding anticoagulants and antiplatelets was relatively small (n=2527), given the low incidence of complications in our health care system. Nonetheless, the findings are conservatively positive and promising, and future work should examine the efficacy of such a program using larger samples.

Also, the current study spanned 10 years. Though there are strengths to using this timeframe, there are also potential limitations, given that there might have been variations in both the hospital environment and the sample characteristics. For example, it is important to note that many of the patients who elected to have surgery after the medication management program was introduced were patients seeking care during the COVID-19 pandemic. The pandemic led to different patient demographics across our subsamples, likely because many elective surgeries were eliminated, with only the most at-risk patients having surgery during the height of the pandemic. Consistent with this hypothesis, on average, patients in the medication management subsample were older with higher BMIs than patients in the pre-program group. Although the groups differed in terms of certain demographics, one might predict that these demographic differences for the medication management group (eg, older age and higher BMIs) should have led to worse outcomes compared with patients in the pre-program group.9 Yet, despite these demographic risks (older age and higher BMI), the medication management program group still fared better than the pre-program group. Finally, in this study, we examined patient complications within just one health care system in one part of the United States. Clearly, future work is needed before the generalizability of these findings can be firmly established. For now, though, the current results suggest that the medication management support program may have been effective at reducing complications during and following spine surgery.

In conclusion, the current findings highlight the possible benefit of standardizing care for high-risk spine patients. In a dynamic field where research findings are constantly unfolding, physicians on the front lines and their patients can potentially benefit from hospital infrastructure that supports the careful implementation of new findings. Though previous work has investigated how to lower complications from spine surgery,10,11 this is the first study we know that has evaluated the effectiveness of a medication management support program like this one in a real-world hospital setting.

Key Points

• The implementation of a medication management support program led to fewer blood transfusions within 72 hours of incision.

• The use of a medication management program led to fewer thrombotic complications: myocardial infarction, stroke, pulmonary embolism, and deep vein thrombosis.

• These findings should be taken with caution and are consistent with existing research showing that standardizing health care reduces risk.