Perioperative Multimodal Analgesia Reduces Opioid Use Following Skin Grafting in Nonintubated Burn Patients

Abstract

Hundreds of patients are treated for burn injuries each year at University of Wisconsin School of Medicine and Public Health. Pain management is particularly challenging during dressing changes and following skin grafting procedures. We performed a retrospective chart review from January 2011 through June 2018 to evaluate the effect of nonopioid analgesic medications on opioid use in nonintubated patients. Our primary outcome was the change in opioid use following the procedure. We found that most patients (69%) report severe pain (Numeric Rating Scale ≥7) immediately after autologous skin grafting. On average, patients required an additional 52 mg of oral morphine equivalents (ME) in the 24 h after the procedure compared with the 24 h before. The use of perioperative nonopioid analgesia varied between patients (acetaminophen 29%, gabapentin 29%, ketamine 35%, and all three 8%). Patients who received either gabapentin or a combination of acetaminophen, gabapentin, and ketamine had a smaller increase in their opioid use than patients who did not receive the medications (−25 ME, 95% confidence interval [−46, −4]; P = .018 and −47 ME, [−81, −11]; P = .010, respectively). These results support using a combination of acetaminophen, gabapentin, and ketamine for perioperative analgesia in burn patients undergoing autologous skin grafting.

Burn injuries cause severe pain and most patients require high doses of opioid medications. Pain management is particularly challenging during dressing changes and following skin grafting procedures. Perioperative protocols often include nonopioid medications such as ketamine to improve pain control and reduce reliance on opioids. An emphasis on incorporating nonopioid medications stems from a growing appreciation of the adverse effects of opioids¹ and the risk of developing chronic opioid use following surgery.^2,3

Many studies have assessed the analgesic efficacy of nonopioid medications. A recent Cochrane meta-analysis examining the use of perioperative ketamine supports a modest reduction in opioid use.⁴ On the other hand, a similar analysis of perioperative lidocaine infusion found no effect on pain or opioid use.⁵ Meta-analyses investigating the use of perioperative gabapentin have yielded mixed results, with some suggesting modest reductions in pain or opioid use⁶ and others finding little effect.^7,8 Some have raised concerns about respiratory depression and cognitive impairment associated with gabapentin.^7,9 Thus, it is important to weigh the risks and benefits of both opioid and nonopioid medications when considering pain management. The decision to use a medication may depend on the procedure and patient population.

Few studies assess the perioperative use of nonopioid medications specifically in burn patients. Since most burn patients have significant pain and opioid requirements prior to surgery, their pain management may require additional consideration compared with other surgical patients. Ketamine has been shown to reduce background and procedural pain in burn patients.¹⁰ Pregabalin has been shown to reduce neuropathic pain in burn patients.¹¹ However, various doses of gabapentin failed to demonstrate any analgesic benefit or reduction in opioid use.¹² Another study reported no benefit from perioperative propranolol.¹³ While some have used nonsteroidal anti-inflammatory drugs without adverse events, the analgesic benefit of nonsteroidal anti-inflammatory drugs in burn patients was unclear¹⁴ and may be outweighed by the risk of side effects.^10,15 Nonetheless, nonopioid analgesic medications are routinely used in the management of burn patients.^16–19 Our standard pain control regimen has been oxycodone orally every 4 h as needed. When maximum dosages of oxycodone are inadequate, extended-release oxycodone every 8 to 12 h is added. Intravenous (IV) fentanyl and low dose ketamine are used for breakthrough and procedural pain control. Acetaminophen and gabapentin have been added as adjuncts to pain control in recent years.

In this study, we describe the current use of multimodal medications for burn patients undergoing skin grafting procedures at our institution. Further, we assess the effects of nonopioid medications on opioid use and pain scores following the procedure.

METHODS

This study was approved by the University of Wisconsin—Madison Institutional Review Board. Using research electronic data capture (REDCap),²⁰ a retrospective chart review was performed for burn patients who underwent autologous skin grafting procedures between January 2011 and June 2018. We included patients 18 to 80 years old and who were inpatients for at least 24 h before and after the procedure. We excluded patients who were intubated before or after surgery, had surgery the day prior or had a history of chronic opioid use prior to their burn injury. These factors may have affected pain assessments and opioid requirements perioperatively. We were unable to accurately quantify home opioid use from the medical record. This would be important to gauge opioid tolerance and accurately characterize perioperative opioid requirements. Further, we did not identify an adequate number of patients to perform subgroup analyses for these different factors. When patients underwent multiple skin grafting procedures, only data for the first skin grafting procedure were collected. We collected data on patient demographics, BSA burned, opioid use, pain scores (Numeric Rating Scale [NRS] 1–10), and nonopioid analgesic medications used in the perioperative period. All opioid use was converted to morphine equivalents (ME), where 1 ME equals 1 mg oral morphine using the University of Iowa Equianalgesic Chart.²¹ Patients who received ≥500 mg acetaminophen, ≥200 mg gabapentin, ≥200 mg ibuprofen, ≥10 mg ketorolac, >1 mg/kg/hr lidocaine, ≥0.4 mg/kg/hr dexmedetomidine, ≥25 mg of ketamine, or ≥10 mg/hr ketamine infusion were counted as having received that medication. Preoperative medications (acetaminophen and gabapentin) were counted if administered up to 4 h before surgery. Ketamine could be administered during surgery or up to 4 h after surgery. Due to the low number of patients who received ketamine infusions and the variable duration of the infusion, patients who received ketamine infusions were grouped with those who received ketamine boluses for multivariate analysis.

The primary outcome measure of this study was the paired difference between opioid use 24 h before skin grafting and 24 h after skin grafting. Data were summarized using mean (SD), n (%), or median (interquartile range, IQR). BSA burned was determined to be nonnormally distributed, therefore BSA burned was log transformed to achieve a normal distribution and the log-transformation was used for analysis. Multivariable linear regression was used to assess possible associations of baseline variables with daily opioid use prior to skin grafting. Multivariable analysis of variance was used to assess the effect of different nonopioid analgesic medications on the difference in 24-h opioid use while controlling for a set of prespecified covariates. Results were considered statistically significant at P < .05. Analysis was performed using R version 3.5.3.

RESULTS

Two hundred and fourteen patients who underwent autologous skin grafting for a burn injury met all study inclusion and exclusion criteria. Demographic information is summarized in Table 1. Study patients tended to be male (76%) and broadly represented the age range recruited for the study (range 18–80 years). Median BSA burned was 8.1% (4.0%–15.0%; range 0.5%–58%). Median grafted area was 685 cm² (314–1383 cm²; range 20–6570 cm²). Most skin grafts were harvested from the lower extremity. Of the demographic information, age was the strongest predictor of opioid requirements in the 24 h before surgery (Table 2; −2.7 ME, 95% confidence interval [CI; −3.4, −2.0], P < .001). For every year of age, patients required 2.7 ME less opioid on average prior to surgery. Larger BSA burns also correlated with higher opioid requirements (1.8 ME per 10% increase in BSA percentage [0.7, 2.9], P = .002).

Table 1.

Patient demographics summarized as mean (SD), N (%), or median (IQR)

Demographics	Summary
Age	44.5 (15.4)
Gender (male)	162 (75.7%)
BMI (kg/m2)	28.9 (6.5)
BSA burned (%)	8.1 (4.0–15.0)
Grafted area (cm2)	685 (314–1383)

BMI, body mass index.

Table 2.

Predictors of daily opioid requirements prior to skin grafting

Predictor	Coefficient	(95% CI)	P
Age—per year	−2.7	(−3.5, −2.0)	<.001
Gender—male	16.2	(−10.0, 42.5)	.225
BMI—per 1 kg/m2	−0.8	(−2.6, 0.9)	.343
BSA burned*—per 10% increase	1.8	(0.7, 2.9)	.002

BMI, body mass index; CI, confidence interval; ME, morphine equivalents.

*Log-transformed for analysis with coefficient reported on raw ME units, but increase is based on percent increase and not linear increase.

The use of nonopioid medications for analgesia varied and is summarized in Table 3. While many patients received acetaminophen (n = 63), gabapentin (n = 62), or ketamine (n = 71), relatively few patients received all three medications (n = 17). Some patients (n = 7) received ketamine only during their postoperative recovery and were not included in the ketamine bolus or infusion groups. Most ketamine infusions were run intraoperatively and were not continued in recovery (n = 11). Over the study period, patients did not receive ibuprofen, ketorolac, lidocaine infusions, or dexamethasone infusions (n = 0). Therefore, these interventions were not included in the regression analysis.

Table 3.

Medications and the frequency of their use as perioperative multimodal analgesia therapy

Medication	Patients Who Received
Acetaminophen	63 (29.4%)
Gabapentin	62 (29.0%)
Ketamine bolus	71 (33.2%)
Ketamine infusion	12 (5.6%)
Either ketamine	75 (35.0%)
All three medications	17 (7.9%)

Average opioid use increased from 166 ± 94 ME in the 24 h before skin grafting to 218 ± 117 ME in the 24 h after skin grafting. Average pain rating also increased from 3.8 ± 2.7 NRS (1–10) before skin grafting to 7.3 ± 2.7 NRS after. Patients often reported severe pain and required significant amounts of IV opioid medications (44.5, 20–75 ME) during their postoperative recovery. Figure 1 depicts the relationship between opioid use in the 24 h before skin grafting and opioid use after surgery. Vertical distance from the equal use line (y = x, solid line) represents the difference in 24-h opioid use for each patient. Most data points fall above this line, indicating an increase in opioid use. A multivariable analysis of variance was performed controlling for age, gender, log-transformed BSA burned, log-transformed grafted area, and intraoperative opioid use in order to assess the effect of medications on the difference in 24-h opioid use (Table 4). Patients who received either gabapentin or a combination of acetaminophen, gabapentin, and ketamine had a smaller increase in their opioid use than patients who did not receive the medications (−25 ME, 95% CI [−46, −4]; P = .018 and −47 ME, [−82, −11]; P = .010, respectively). This is illustrated for patients who received all three medications (Figure 1, filled circles). While we accounted for intraoperative opioid use in our analysis, a Pearson’s correlation did not demonstrate a significant effect on 24-h opioid use (ρ = 0.10, 95% CI [−0.04, 0.23]) or IV opioid use in postoperative recovery (ρ = 0.04, [−0.09, 0.18]).

Figure 1.

Scatterplot of 24-h opioid use (ME) after skin grafting compared with before. Circles represent data from individual patients. Filled circles represent data from patients who received acetaminophen, gabapentin, and ketamine*. Solid line demarcates equal use before and after skin grafting (y = x). Dashed line depicts the regression line through the data (y = 0.99x + 54). *Ketamine group includes patients who received bolus or infusions.

Table 4.

Multivariable ANOVA of the effect of different analgesic medications on opioid use while controlling for age, gender, log-transformed BSA burned, log-transformed grafted area, and intraoperative opioid dose

Medication	n	No	Yes	Difference	P
Tylenol	63	56.0 (43.5, 68.5)	42.8 (24.2, 61.4)	−13.2 (−34.1, 7.7)	.215
Gabapentin	62	59.2 (46.9, 71.5)	33.9 (15.3, 52.6)	−25.3 (−46.2, −4.4)	.018
Ketamine*	75	55.9 (42.8, 69.1)	45.3 (27.8, 62.9)	−10.6 (−31.5, 10.2)	.316
All three medications	17	55.6 (44.5, 66.7)	9.1 (−25.3, 43.4)	−46.5 (−81.6, −11.4)	.010

ANOVA, analysis of variance; CI, confidence interval; ME, morphine equivalents.

Data, reported as estimated mean (95% CI), represent change in 24-h opioid use after surgery compared with 24 h before (ME) for those that received and did not receive each medication.

*Ketamine group includes patients who received bolus or infusions.

DISCUSSION

Pain management for burn injuries continues to be challenging, especially around the time of skin grafting procedures.^16,22 A multimodal approach to analgesia is generally recommended. However, there are limited studies addressing the efficacy of these approaches in the burn patient population. This retrospective study focuses on pain management at the time of skin grafting. Within-subject opioid use 24 h after surgery compared with 24 h before surgery is the primary outcome measure and serves two purposes. It attempts to reduce intersubject variability, including pain tolerance and differences in burn injury. Further, it focuses on the impact of autologous skin grafting and the medications used for perioperative analgesia. Interpretation of results is limited by the retrospective nature of the study. In order to maximize the power of the study, the analysis included a limited number of controlling factors. Age and gender are known predictors of pain and opioid use. TBSA burned and grafted area were included because of the anticipated association between wound size and opioid use. However, neither the donor site nor graft site was included in the analysis.

It was not possible to account for all combinations of medications used for perioperative analgesia, their dose and their timing of administration. Medications such as gabapentin may have been administered for a variable period prior to the day of surgery. Dose was not evaluated as each medication was accounted for in a logistic fashion (either received or not received on the morning of surgery). This approach was taken because nonopioid medications may not have a linear effect on opioid requirements. Assuming a linear (or nonlinear) relationship for each medication could significantly confound the results of a regression analysis.

The primary outcome of this study supports the use of multimodal analgesia for perioperative analgesia in burn patients. Patients who received a combination of acetaminophen, gabapentin, and ketamine had a smaller increase in opioid use after their skin grafting procedure. In fact, many patients who received all three medications barely increased their opioid use at all. The estimated effect size of −46 ME per day seems clinically meaningful considering the adverse effects of opioids¹ and the increasing risk of chronic use with high doses of opioids.²³ This finding agrees with studies supporting the use of multimodal analgesia in a broad patient population.²⁴ However, the primary outcome of this study does not support the individual use of acetaminophen or ketamine for perioperative analgesia in burn patients. This may reflect the power of the study or the synergistic effect of combining medications for pain management. The time and dose of ketamine administration in the perioperative period were quite variable. Thus, it was not possible to determine from the medical record whether some patients received ketamine as part of a preemptive multimodal strategy or as a reactive effort to control severe pain. One should interpret the efficacy of ketamine in context. Many randomized controlled studies have addressed the use of ketamine for perioperative analgesia and demonstrated a beneficial effect.^4,10

The use of gabapentin for perioperative analgesia is controversial.²⁵ Animal models of burn injury have demonstrated an opioid-sparing effect with gabapentin.^26,27 Many clinical studies of have also reported opioid sparing with gabapentin.^28–31 Similarly, this study found that preoperative gabapentin was associated with a smaller increase in opioid use after surgery. However, recent meta-analyses of randomized controlled trials have found no effect on opioid consumption or pain scores.^7,8 Given this evidence, it is uncertain whether gabapentin reduces opioid requirements after surgery. It is possible that burn patients do benefit from gabapentin and are not well-represented in recent meta-analyses. High preoperative opioid use and preexisting acute injury are not typical of most surgical populations. Although one randomized controlled trial found no benefit to administering gabapentin to patients with burn injury during their hospital stay, it did not address perioperative analgesia.¹² Alternatively, some patients in this study were taking scheduled gabapentin but did not receive the medication on the morning of surgery. Thus, the results of this study may represent the harm of holding scheduled gabapentin prior to surgery.

An unintended benefit from this analysis was the recognition that there was significant variation in the use of nonopioid medications for analgesia. Systematic issues that led to holding prescribed medications, such as acetaminophen and gabapentin, on the morning of surgery were apparent from review of the medical record. Perioperative workflow and communication between teams can be improved to address this issue and will best be accomplished by a collaborative approach between burn surgeons and anesthesiologists specializing in the care of burn patients. At our institution, we are refining protocols for the perioperative care of burn patients to improve pain control for our burn population. Additionally, limited or conflicting evidence supporting the use of nonopioid medications for analgesia (alone or in combination) has also contributed to a diversity of clinical practice. Ongoing research and meta-analyses continue to improve the evidence for perioperative pain management.

In contrast to a previous study in burn patients,³² intraoperative opioid use was not correlated with an increase in postoperative opioid use. Thus, we did not observe opioid induced hyperalgesia in response to intraoperative opioid administration. Similar to previous studies, opioid use was correlated with TBSA of the burn injury.³³ Previous studies have also shown an inverse correlation between age and burn pain and between age and opioid use.^32,34

In conclusion, this study supports the use of multimodal analgesia for perioperative analgesia in burn patients who undergo autologous skin grafting. A combination of acetaminophen, gabapentin, and ketamine was most effective in reducing opioid use. Pain management in this patient population continues to be challenging as many patients report severe postoperative pain despite multimodal analgesia.

Funding: This project was supported by the Clinical and Translational Science Award (CTSA) program, through the National Institutes of Health (NIH) National Center for Advancing Translational Sciences (NCATS), grant UL1TR002373. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Conflict of interest statement. The authors have no conflicts to disclose.