The Effect of Scheduled Intravenous Acetaminophen in an Enhanced Recovery Protocol Pathway in Patients Undergoing Major Abdominal Procedures: A Prospective, Randomized, and Placebo-Controlled Clinical Trial

Abstract

Objective

Enhanced recovery protocols (ERPs) utilize multi-modal analgesia regimens. Individual regimen components should be evaluated for their analgesic efficacy. We evaluated the effect of scheduled intravenous (IV) acetaminophen within an ERP on analgesia and recovery after a major abdominal surgery.

Design

This study is a prospective, randomized, double-blinded clinical trial.

Setting

The study setting was a tertiary care, academic medical center.

Subjects

Adult patients scheduled for elective major abdominal surgical procedures.

Methods

Patients in group A received 1 g IV acetaminophen, while patients in group P received IV placebo every six hours for 48 hours postoperatively within an ERP. Pain scores, opioid requirements, nausea and vomiting, time to oral intake and mobilization, length of stay, and patient satisfaction scores were measured and compared.

Results

From 412 patients screened, 154 patients completed the study (Group A: 76, Group P: 78). Primary outcome was the number of patients with unsatisfactory pain relief, defined as a composite of average Numeric Rating Scale (NRS) scores above 5 and requirement of IV patient-controlled analgesia for pain relief during the first 48 hours postoperatively, and was not significantly different between the two groups (33 (43.4%) in group A versus 42 (53.8%) patients in group P, P = .20). Opioid consumption was comparable between two groups. Group A utilized significantly less postoperative rescue antiemetics compared to group P (41% vs. 58%, P = .02).

Conclusions

Scheduled administration of IV acetaminophen did not improve postoperative analgesia or characteristics of postoperative recovery in patients undergoing major abdominal surgery within an ERP pathway.

Introduction

In response to an increased focus on improving patient outcomes and satisfaction with surgical care, a growing body of clinical evidence has recently been dedicated to enhanced recovery protocols (ERPs). These evidence-based perioperative pathways aim to optimize patients undergoing surgery in the preoperative, intraoperative, and postoperative periods. ERPs have incorporated the use of multimodal analgesia to minimize the use of intra- and postoperative opioid analgesics. Lidocaine, ketamine, magnesium, gabapentin, acetaminophen, and nonsteroidal anti-inflammatory drugs are some of the adjuvant analgesics used in combination with regional blocks to optimize analgesia and recovery. Multimodal analgesia has dependably been shown to significantly reduce postoperative opioid requirements as well as opioid-related side effects such as postoperative nausea and vomiting [1, 2]. However, the incremental roles of individual ERP components have not yet been fully elucidated.

Both oral and intravenous (IV) acetaminophen preparations have been shown to be useful adjuvants in multimodal analgesia [3]. IV acetaminophen has been of interest for its utility in postsurgical patients who have not yet been cleared for or are intolerant of oral intake. IV acetaminophen is also preferred over oral acetaminophen early after major abdominal surgery when absorption of medications given via the oral route may be erratic. Oral acetaminophen has been shown to be less effective than IV acetaminophen in patients undergoing colorectal surgery [3]. However, a large claims database study evaluating the opioid-sparing efficacy of acetaminophen questioned the utility of IV acetaminophen for patients undergoing open colectomy procedures [4]. Although IV acetaminophen is often utilized as a postoperative pain adjunct in most ERPs, few randomized clinical trials have evaluated its exact role in ERPs and non-opioid multimodal analgesic regimens for major abdominal surgery [5, 6]. The primary goal of this study was to assess the effect of a scheduled perioperative IV acetaminophen dosing regimen on postoperative pain, opioid consumption, and opioid-related side effects within an ERP. We also aimed to study overall selected characteristics of postoperative recovery of this regimen as part of an ERP at a large tertiary medical center.

Methods

Our Institutional Review Board approved this study (Study 19040241) and informed consent was obtained from the patients. The trial was registered at www.clinicaltrials.gov (NCT03198871) before recruitment and was conducted as a single-center, prospective, randomized, double-blinded, placebo-controlled trial at an academic, tertiary care hospital.

Study Population

The study population included patients scheduled for elective colorectal, pancreatic, and other major abdominal procedures (gastric, abdominal wall ventral hernias, etc.) via our institution’s ERP. Patients who refused to participate in the study or part of the ERP, as well as those with a documented allergy to acetaminophen, pregnancy, chronic alcoholism, chronic malnutrition, preoperative renal insufficiency (creatinine clearance ≥30 mL/minute or hemodialysis), history of hepatic impairment or active hepatic disease, severe chronic pain condition requiring daily preoperative opioid medications, or pre-existing dementia and/or other neuropsychiatric conditions impeding accurate assessment of pain scores or other study measures were excluded. Patients with stable psychiatric conditions who were taking psychiatric medications were not excluded, but the conditions and medications were recorded for documentation. Patients with a positive urine pregnancy test were excluded. Patients taking isoniazid and warfarin at the time of surgery were excluded.

Interventions

After obtaining informed consent, patients were randomly assigned to one of the two groups: IV acetaminophen (group A) or placebo/control (group P). Our research pharmacy prepared the study medications according to random, computer-generated numbers assigning the enrolled patients to either the placebo or acetaminophen group via a 1:1 non-block randomization. Treatments were assigned in numerical order. Tylenol and saline were dispensed in identically packaged IV bags to ensure blinding. Patients, anesthesiologists, nurses, and research staff were all be blinded to the medications prepared by the research pharmacy. In addition to receiving multimodal analgesia in accordance with our institution’s current ERP (Supplementary Data  Table  1), patients in group A patients received 1 gram of IV acetaminophen at the start of wound closure to be repeated every 6 hours for 48 hours postoperatively. The maximum recommended dosage of acetaminophen is 3 grams per day, especially if administered for longer periods (more than seven days). We used the maximum recommended dose of 4 grams IV acetaminophen per day, as we used only for 48 hours. Patients in group P received IV saline placebo at wound closure and every 6 hours for 48 hours postoperatively. Oral acetaminophen is typically a standard component of our institutional ERP perioperative medication protocol, but was omitted for the purpose of the study. Our institutional ERP protocol allowed for variations in analgesic and anesthesia drug administration based on the patient’s age, risk for postoperative nausea and vomiting, presence of contraindications to spinal morphine administration, and intraoperative hemodynamic response to nociceptive stimulation (e.g., persistent tachycardia) (Supplementary Data  Table  1).

Table 1.

Preoperative patient characteristics and surgical variables

	Group P (N = 78)	Group A (N = 76)	P value
Age (years)*	64 (53.75, 74.00)	62 (55, 70.75)	.67
Male	34 (43.6%)	36 (47.4%)	.64
BMI (kg/m2)*	28.25 (23.80, 32.40)	27.70 (24.40, 32.55)	.84
CCI
0	41(52.6%)	44 (57.9%)	.35
1	10 (12.8%)	4 (5.3%)
2	24 (30.8%)	23 (30.3%)
> 2	3 (3.9%)	5 (6.6%)
ASA status			.52
II	18 (23.1%)	17 (22.4%)
III	55 (70.5%)	57 (75%)
IV	5 (6.4%)	2 (2.6%)
Race
Asian	1 (1.2%)	0	.64
African American	2 (2.5%)	4 (5.3%)
Caucasian	73 (93.6%)	70 (92.1%)
Other/NA	2 (2.5%)	2 (2.6%)
Medications for mental health	27 (34.6%)	15 (19.7%)	.04
Diabetes	3 (3.8%)	4 (5.3%)	1.00
Hypertension	16 (20.5%)	10 (13.2%)	.19
Coronary artery disease	4 (5.1%)	2 (2.6%)	.68
Congestive failure	3 (3.8%)	1(1.3%)	.62
Atrial fibrillation	2 (2.5%)	2 (2.6%)	1.00
Pulmonary disease	6 (7.7%)	3(3.9%)	.49
Malignancy	23 (29.5%)	28 (36.8%)	.38
Preoperative serum creatinine (mg/dL)*	0.82 (0.7, 1.05)	0.8 (0.7,1.0)	.45
Open	26 (33.3%)	36 (47.4%)	.19
Laparoscopic	41 (52.6%)	33 (43.4%)
Robotic	11 (14.1%)	7 (9.2%)
Colorectal	44 (56.4%)	48 (63.2%)	.48
Pancreatic	15 (19.2%)	14 (18.4%)
Other major abdominal	19 (24.4%)	14 (18.4%)
Duration of surgery (min)*	189.50 (117, 326)	183 (127.50, 284)	.98
Blood loss (mL)*	50 (25, 150)	50 (20, 200)	.87
Fluids (mL)*	2500 (1700, 3500)	2500 (1800, 3000)	.81
Transfusion required	2 (2.5%)	0	.50

^*Data listed as median (25%, 75% quartile). All others described as number of patients (%).

BMI=body mass index; CCI=Charlson comorbidity index; ASA Class=American Society of Anesthesiologists.

Pain Assessment and Management

Postoperative pain intensity was measured using the Numeric Rating Scale (NRS) with 0 representing no pain and 10 representing maximum pain. The analgesic efficacy in both groups was also evaluated by the amount of total opioid consumption. Pain scores and opioid consumption were assessed both in the post anesthesia care unit (PACU) as well as on the surgical floor following PACU discharge. In the PACU, total IV hydromorphone was primarily utilized as a rescue analgesic and its total consumption was noted. Pain scores were recorded every 30 minutes until discharge to the floor. Readiness to discharge from the PACU was noted. After PACU discharge, a pain score was obtained upon admission to the floor and thereafter every four hours for the first 24 hours postsurgery, then every six hours until 48 hours postoperatively and then every 12 hours until 72 hours postoperatively. Rescue analgesia was given if NRS score was higher than 4. If the patient could take oral medications, oxycodone 5–10 mg orally was given. If the patient could not take oral medications or if the pain was not responsive to oral oxycodone, an “acute pain management protocol” was initiated (Supplementary Data Figure 1). Hydromorphone 0.5 mg IV was administered as rescue analgesic. Ketorolac 15–30 mg IV could also be given for pain relief after clearance from the surgical team. If the pain was refractory (NRS > 4) to three doses of hydromorphone IV in eight hours, ketorolac IV, and an increase in oral oxycodone dose, then patient controlled analgesia (PCA) with hydromorphone was offered. Time to PCA initiation on the floor, total duration of PCA use and total, PCA hydromorphone consumption were noted. A standard opioid conversion table (Supplementary Data Table 2) was used to convert the oral and IV opioids utilized by the patients to oral morphine equivalent doses (OMEs) for analysis.

Outcome Measures

The primary outcome for the study was the number of patients with unsatisfactory pain relief defined as composite of average NRS scores higher than 5 and requirement of IV PCA for pain relief during the first 48-hour postoperative period. Secondary outcome measures included NRS scores, total opioid consumption, PCA requirement, time to readiness for discharge from PACU, time to bowel movement, time to oral intake (liquid and regular diet), time to ambulation, time to hospital discharge, intensive care unit (ICU) admission, 30-day readmission to the hospital, percentage of patients readmitted because of pain-related issues, overall patient satisfaction, and patient satisfaction related to pain management. Overall patient satisfaction and satisfaction of pain management during hospitalization were measured at hospital discharge using a NRS, with 0 being the lowest satisfaction and 10 being the highest satisfaction.

Additionally, several potential adverse events of interest were recorded. Investigators used the Intensive Care Delirium Screening Checklist (ICDSC) at baseline to assess delirium and then postoperative delirium every 12 hours for 72 hours after surgery. Postoperative nausea and/or vomiting (PONV) was evaluated by nausea score (on a scale of 0–10, 0—no nausea and 10—worst nausea) measured every 12 hours for the first 3 postoperative days, the incidence and frequency of emesis (0–72 hours), and rescue antiemetic requirement (0–72 hours). Patient’s overall health at 30 days after discharge was evaluated using the SF-12 health survey.

Statistics

Sample size was calculated based on the percentage of ERP population with unsatisfactory pain relief defined as a composite of average visual analog scale scores higher than 5 and requirement of IV PCA for pain relief during the first 48 hours postoperatively. Based on a retrospective review of our own institution’s medical records, the incidence of unsatisfactory pain relief was observed in 50% of ERP abdominal surgical patients. The investigators defined an absolute reduction in the incidence of patients having unsatisfactory pain relief by 25% to be clinically meaningful for any intervention. In order to detect this difference with 90% power (assuming α = 0.05), a total of 154 patients randomized to two groups were required. With 15% screen failures and drop-outs, we planned to enroll 182 patients total into the study, with the option to terminate the study after enrollment of 154 patients with complete data.

All continuous variables were compared between the two groups using the Wilcoxon rank sum test and dichotomous variables were compared with the Chi-square test. The trend of pain scores over time between both groups was evaluated using mixed model regression after adjusting for variables affecting pain intensity. SAS version 9.4 (SAS Institute Inc., Cary, NC) was utilized for statistical analysis. P < .05 was considered statistically significant.

Results

Of 180 patients enrolled, 154 completed the study and were included in the final analysis (Figure  1): 78 patients received saline placebo (group P) and 76 received IV acetaminophen (group A). Demographic variables were comparable between groups (Table  1). Patient comorbidities were similar, except the proportion of patients with mental illnesses taking antidepressants and other medications was significantly higher in the placebo group (group A, 20% vs group P, 35%; P = .04) (Table  1). Type of surgical procedure, approach to surgery (open/laparoscopic/robotic), surgical duration, blood loss, and fluids given were statistically comparable (P > .05) (Table  1). All intraoperative anesthetic and analgesic medications were given in similar frequency and comparable doses (Table  2).

Figure 1.

Consort flow diagram for the study population.

Table 2.

Intraoperative analgesics/anesthetics administered as a part of ERP

	Group P (N = 78)	Group A (N = 76)	P value
Anesthesia medications
Spinal opioids	66 (84.6%)	60 (79.0%)	.56
Oral morphine premedication	7 (9%)	13 (17%)	.16
Intraoperative opioids	15 (19.2%)	24 (31.6%)	.10
Total pre and intraoperative opioids (oral morphine equivalents, mg)*	6.45 (17.10)	11.35 (25.81)	.09
IV Midazolam premedication	47 (60.2%)	48 (63.2%)	.71
Patient at risk for PONV received Scopolamine patch/Aprepitant	6 (7.7%)	2 (2.7%)	.28
Lidocaine, mg**	100 (80–100)	100 (60–100)	.41
Ketamine, mg**	117.80 (82.89, 170)	109.15 (83.60, 144.73)	.50
Intraoperative IV dexmedetomidine	19 (24.4%)	22 (28.9%)	.52
Intraoperative ketorolac	8 (10.3%)	8 (10.5%)	.95
Total intravenous anesthesia with propofol	9 (11.5%)	9 (11.8%)	.92

^*Data listed as mean (SD) and

^**median (25%, 75% quartile).

All others described as number of patients (%).

The primary outcome measure of the number of patients with unsatisfactory pain relief, defined as a composite of average NRS scores higher than 5 and IV PCA requirement for pain relief during the first 48 hours postoperatively, was not significantly different between the two groups (42 (53.8%) patients in group P versus 33 (43.4%) in group A, P = 0.20). Mixed model regression analysis did not show any significant difference in pain scores over time between the two groups (Figure  2). Total opioid consumption measured by OMEs, PCA requirement, and other non-opioid analgesic requirement were not different between the two groups (Table  3).

Figure 2.

After adjusting for age, gender, mental health medications, and intraoperative narcotic use, mixed model analysis showed no significant difference in pain scores at rest (P = .38) or with movement (P = 0.49) over time. ((A) Those treated with acetaminophen reported lower pain scores at rest (mean = 5.01, SE = 0.70) compared to the control group (mean = 5.86, SE 0.42) four hours post operatively. However, the acetaminophen group reported a higher mean increase in pain (mean = 0.03, SE 0.01) over time compared to the control group (mean = 0.02, SE = 0.01). ((B). Similar results were found with movement. The acetaminophen group reported a lower score at four hours post operatively (mean = 6.22, SE (0.91) compared to the control group (mean = 6.89, SE 0.54). Over time, the treatment group again reported a higher mean increase in pain (mean = 0.02, SE = 0.01) compared to the control group (mean = 0.01, SE = 0.01).

Table 3.

Postoperative pain outcomes

	Group P (N = 78)	Group A (N = 76)	P value
PACU opioid consumption (OME mg)*	8 (0, 20)	8 (0, 20)	.78
PACU Duration (minutes)*	152.50 (98.50, 208.50)	129.00 (106, 186)	.41
Number of patients with
Average pain > 5 and requirement of IVPCA (composite) for 48 hours postoperative period	42 (53.8%)	33 (43.4%)	.20
Postoperative opioid consumption (OME, mg)*
0–24 hours	36 (17.5,85)	27.4 (10,49)	.09
24–48 hours	45 (15,95)	31.3 (7.5,73.3)	.12
48–72 hours	37.5 (0–70)	30 (0,70)	.50
0–48 hours	88.8 (42.5,167.5)	74 (25.5,113)	.09
0–72 hours	127.1 (49.3, 252.00)	105.00 (41.3, 186.3)	.20
PCA required	17 (21.8%)	12 (15.8%)	.34
NSAID (IV ketorolac, oral ibuprofen) 0–72 hours	38 (48.7%)	43 (56.6%)	.33
Gabapentin 0–72 hours	5 (6.4%)	3 (3.9%)	.72
Acetaminophen, 48–72 hours	32 (41%)	30 (39.5%)	.90
Lidocaine infusion postoperatively	3 (3.8%)	1 (1.3%)	.62
Ketamine infusion postoperatively	3 (3.8%)	0	.25

^*Data listed as median (25%,75% quartile). All others described as n (%).

PACU=post anesthesia care unit; NSAID=Nonsteroidal anti-inflammatory drug.

Incidence of emesis, number of emetic episodes, and nausea scores measured every 12 hours for the first three postoperative days were not significantly different between the groups. However, group A had a significantly decreased postoperative rescue antiemetic requirement compared to group P (58% vs 41%, P = .02). Time to ambulation, oral intake, and bowel movement were not significantly different, although a trend towards earlier discharge was seen in acetaminophen-treated patients. Reported frequency of ICU admission and urinary retention were similar for both groups. Patient satisfaction scores and SF-12 scores after discharge were also not different between the two groups (Table  4). None of the patients in either group scored higher than 4 on the ICDSC scale, which suggests that none of the patients developed postoperative delirium at any point of the study.

Table 4.

Postoperative adverse effects and other secondary outcome measures

	Group P (N = 78)	Group A (N = 76)	P value
ICU admission	4 (5.1%)	5 (6.6%)	.70
Emesis incidence 0–72 hours	14 (17.9%)	12 (15.8%)	.72
Episodes of emesis in 72 hours*	0.07 (0.26)	0.03(0.16)	.62
Antiemetic use	45 (57.7%)	31(40.8%)	.02
Nausea score POD 1 (am visit)*	0.60 (1.76)	1.13 (2.82)	.61
Nausea score POD 1 (pm visit)*	0.92 (2.53)	0.49 (1.54)	.51
Nausea score POD 2 (am visit)*	0.75 (2.39)	0.71 (1.97)	.65
Nausea score POD 2 (pm visit)*	0.69 (2.36)	0.55 (1.70)	.45
Nausea score POD 3 (am visit)*	0.66 (2.18)	0.32 (1.37)	.98
Nausea score POD 3 (pm visit)*	0.78 (2.42)	0.56 (1.77)	.83
Urinary retention	3 (3.8%)	5 (6.5%)	.49
Time to bowel movement (hours)**	64.66 (29.27, 115.33)	46.30 (25.38, 90.85)	.11
Time to oral intake (hours)**	9.40 (3.87, 43.38)	12.15 (4.53, 33.17)	.51
Time to ambulation (hours)**	17.38 (4.28, 22.65)	18.82 (4.42, 37.77)	.13
Time to hospital discharge (day)**	4.94 (3.94, 7.12)	4.08 (3.07, 6.04)	.06
Patient satisfaction score, overall**	10 (9, 10)	10 (9, 10)	.68
Patient satisfaction score, pain management**	10 (8, 10)	10 (9, 10)	.34
SF 12 Physical score**	38.07 (32.56, 45.85)	38.72 (30.73, 43.28)	.60
SF12 Mental score**	52.08 (42.66, 57.88)	54.02 (48.33, 57.42)	.53
Readmissions (within 30 days)	10 (12.8%)	9 (11.8%)	1.00

^*Data listed as mean (SD) and median (25%,75% quartile).

^**All others described as n (%).

ICU = intensive care unit; POD = postoperative day.

Since this is a randomized study, 13% more open surgery occurred in group A (47%) than in group P (34%). We therefore performed a sub-group analysis to evaluate any selective analgesic efficacy of IV acetaminophen in open or minimally invasive (laparoscopy + robotic surgical procedures) surgery compared to placebo. No significant difference between the two interventions could be demonstrated in either the open or minimally invasive surgical subgroups (Supplementary Data Tables 2–5).

Discussion

In this randomized clinical trial, we strived to identify the incremental effect of a single drug (IV acetaminophen) within a complex ERP utilizing several multimodal analgesic drugs (spinal morphine, ketamine, gabapentin). Complex interrelationships between different analgesic components may be difficult to determine, as the sum of effects may be greater than the effect of individual components. Scheduled IV acetaminophen compared to placebo as part of an ERP in abdominal surgery did not significantly decrease the number of patients with either average pain scores > 5 or those who required IV PCA for pain relief during first 48 hours postoperatively. We measured this as a composite primary outcome because reported pain scores alone are a very subjective measure and opioid consumption alone as a measure of analgesic efficacy in this climate of restrictive opioid use is questionable. IV PCA was given when the pain was not manageable by other means as a last resort, indicating significant pain experienced by the patients. Hence, we felt that a composite measure would be a more reliable indicator of unsatisfactory pain relief in this patient population.

The use of IV acetaminophen has gained traction as an opioid-sparing alternative for postsurgical pain [5, 7–13]. A number of retrospective and prospective studies on IV acetaminophen use in patients undergoing abdominal surgery have been conducted but the results were not conclusive with regards to its beneficial effects on pain scores, PONV, patient satisfaction, and length of hospital stay [1, 5, 6, 14–20]. A recently published prospective trial by Aryaie et al. used 1 g IV acetaminophen every 6 hours vs placebo in 97 patients undergoing colorectal surgery for up to five days postoperatively [6]. The IV acetaminophen group had significantly decreased opioid consumption, pain scores, length of hospital stay, time to first bowel movement, time to diet ordered, and rates of postoperative ileus. Several reasons could have contributed to this group’s outcomes that we were not able to demonstrate in the current study. First, Aryaie et al. did not employ multimodal analgesia or spinal morphine in their trial. Use of these analgesics in our study could have masked any opioid-sparing effect of IV acetaminophen. Second, all patients in the Aryaie et al. study were given IV PCA from the immediate postoperative period, whereas less than 25% of our patients received PCA. In line with current guidelines for providing opioid-restrictive intra- and postoperative analgesia to patients with major surgical procedures, we restricted PCA for patients with severe refractory pain. Third, they used IV acetaminophen every six hours until discharge compared to only 48 hours of scheduled use in our study. This could have contributed to decreased length of stay and decreased postoperative ileus in their study. We allowed IV and oral acetaminophen on demand after 48 hours, but only 40% of patients required acetaminophen thereafter.

A meta-analysis of acetaminophen administration postoperatively demonstrated that oral or IV acetaminophen carried a greater opioid-sparing and pain-reducing effect in orthopedic surgery compared to abdominal surgery [21, 22]. IV acetaminophen may be more effective in treating somatic pain than the visceral pain associated with major abdominal surgery. This could explain the beneficial effects of IV acetaminophen in patients undergoing orthopedic surgery [12, 13]. IV acetaminophen use was associated with decreased opioid use and lower incidence of delirium in patients undergoing cardiac and orthopedic surgical procedures [23, 24]. We did not find any patients developing postoperative delirium in this study. Our study population’s age was comparable (median age in both groups over 60 years), but the nature of the procedure (cardiac or hip surgery) predisposing to micro-emboli may explain the differences in the study findings. The use of a more validated tool, the Confusion Assessment Method in their study, also may have played a role [25].

Strengths of our study include the prospective, randomized, double-blind placebo-controlled design, as well as the standardized ERP that was used for all patients. Our study has few limitations. First, this is a single-center study; thus, the findings should be interpreted in the background of an institution-specific ERP and pain management protocols. For instance, several institutions use IV PCA as the first line of therapy starting from the recovery room. At our institution, we utilize IV PCA only after the failure of oral and breakthrough nurse-administered intermittent opioid analgesia regimens (Supplementary Data Figure 1). Second, although we protocolized the administration of all multimodal analgesic drugs (Table S1), physicians caring for patients are allowed to respond to the clinical situation and act accordingly. One such example is administration of dexmedetomidine and opioids intraoperatively if the patients exhibit persistent tachycardia. We analyzed the frequency and total dose of all perioperative analgesics (Table  2) between the two groups and could not identify that the preference to use of any of these analgesic drugs accounted for the results. Third, patients with psychiatric medications were randomly assigned more often to the placebo group and open surgical procedures were assigned more often to the IV acetaminophen group. This could have affected the pain outcomes. We performed a mixed model analysis adjusted for psychiatric medications and other factors to compare pain scores over time and did not find any difference in adjusted pain scores between the two groups (Figure  2). Fourth, although patients’ baseline pain data was not collected, patients with severe chronic pain conditions that required daily preoperative opioid medication were excluded. The findings of this study would not be applicable to this patient population with chronic pain. Another limitation is the use of NRS to assess pain, as functional pain scales have shown a higher responsiveness index compared to other pain assessment tools, including the NRS [26]. We decided to use the NRS, as this is a simple, easy to use instrument utilized by the nurses on the floor for years and literature reviews have shown higher compliance rates with NRS [27, 28]. We also performed a subgroup analysis for open and minimally invasive procedures. There was a trend towards better pain outcomes for IV acetaminophen-treated patients in the open surgery subgroup, but a separate study with an adequate sample size is required to evaluate this effect. Another limitation was that we did not evaluate patient-centered outcome measures such as Quality of Recovery scores. We evaluated time to mobilization and pain on movement but did not monitor or report the frequency and effectiveness of physical therapy among the study subjects. Finally, we did not compare oral acetaminophen to IV acetaminophen. Marcotte et al. reported that IV acetaminophen use was associated with less opioid and PCA use with lower PONV compared to PO acetaminophen in patients undergoing colorectal procedures with an ERAS multimodal analgesia protocol, although patient-reported pain scores were not found to be different [3]. This was a single institutional, non-randomized, retrospective study without protocolized acetaminophen administration. Study patients belonged to two different time periods. Thus, the question of whether oral acetaminophen is as effective as IV acetaminophen remains unanswered. However, since IV acetaminophen lacks efficacy, with limited absorption of oral acetaminophen in this patient population, we suspect it is also unlikely to have superior efficacy.

Although the primary composite outcome did not favor the use of scheduled IV acetaminophen administration for major abdominal surgery, there are other study findings that deserve further investigation. We analyzed pain scores and opioid consumption separately as secondary outcomes. Postoperative opioid consumption decreased in IV acetaminophen group but not to a level of significance. A difference in median opioid consumption (OMEs) of 14 mg may not be statistically significant but can be clinically relevant. This is especially relevant with the findings of decreased rescue antiemetic requirement and trend for earlier discharge in the IV acetaminophen group. Postoperative nausea and vomiting (PONV) is one of the most common adverse events after surgery. PONV is distressing for patients, decreases patient satisfaction and increases the risk of other adverse events such as readmissions [29]. Having less PONV is financially beneficial for health care institutions by decreasing the length of PACU and hospital stay, staff burden, supply costs and readmission rates [30, 31].

Conclusion

In this study, we found that IV acetaminophen, within the context of an ERP abdominal surgery population, was not very effective at reducing postoperative pain measures. The need for postoperative rescue antiemetics was lower in the IV acetaminophen group.

 

Funding sources: We acknowledge the following grant support for this study:

1. Research grant (Grant number; 712886) Mallinckrodt Pharmaceuticals for the conduct of the study (Kathirvel Subramaniam & University of Pittsburgh).

2. The project described was also supported by the National Institutes of Health through grant number UL1TR001857 (statistical analysis—Kristine Ruppert).

Disclosure: The study was supported by a grant from Mallinckrodt Corporation.