Sugammadex Versus Neostigmine for Reversal of Residual Neuromuscular Blocks After Surgery

A Retrospective Cohort Analysis of Postoperative Side Effects

Kurt Ruetzler, MD, FAHA; Kai Li, MD; Surendrasingh Chhabada, MD; Kamal Maheshwari, MD, MPH; Praveen Chahar, MD, FCARCSI; Sandeep Khanna, MD; Marc T. Schmidt; Dongsheng Yang, MS; Alparslan Turan, MD; Daniel I. Sessler, MD

Disclosures

Anesth Analg. 2022;134(5):1043-1053.

Results

We identified 89,753 surgeries on 70,690 patients who met inclusion and exclusion criteria, including 16,480 (18%) surgeries during which sugammadex was given at the end of surgery and 73,273 (82%) during which neostigmine was given (Figure 1). Table 1 lists the patients who met screening criteria for adjudication of clinically important adverse effects of reversal agents.

(Enlarge Image)

Figure 1.

Flow diagram. ASA indicates American Society of Anesthesiologists; ICD, International Classification of Diseases.

Patients were given a mean (SD) of 3.9 (2.5) mg kg^–1 sugammadex and a mean of 279 (100) μg kg^–1 neostigmine. The average age was 55 (19) years, and 51% were women. Baseline characteristics and surgical factors, after weighting, were well balanced except type of surgery (ASD = 0.23; Table 2; Figure 2). Consequently, the type of surgery was adjusted for in all of the analyses comparing patients who received sugammadex and neostigmine. The histograms of estimated propensity scores showed a good overlap and similar distribution across 2 groups (Supplemental Digital Content 1, Figure 1, https://links.lww.com/AA/D764). Supplemental Digital Content 2, Table 1, https://links.lww.com/AA/D765, summarized postoperative events by types of surgeries. Supplemental Digital Content 3, Table 2, https://links.lww.com/AA/D766, reported the balance of baseline characteristics and surgical factors after excluding pediatric and cardiac surgeries, and the type of surgery was also imbalanced (ASD = 0.112).

(Enlarge Image)

Figure 2.

Plot of ASD between sugammadex group (N = 16,480) and neostigmine group (N = 73,273) on covariables before and after the inverse propensity score weighting. First, we estimated the probability of receiving sugammadex (ie, the propensity score) for each patient using logistic regression with receiving sugammadex as the outcome using the following prespecified potential confounding variables (listed in Table 1). Then, we calculated the stabilized weights as follows: for the sugammadex group, weight was the proportion of sugammadex patients divided by the propensity score; for the neostigmine group, weight was the proportion of neostigmine patients divided by (1 – propensity score). To be conservative, we also truncated the stabilized weights at 1st and 99th percentiles. ASD is absolute difference in means or proportions divided by the pooled standard deviation. Any covariable with an ASD >0.10 was considered to be imbalanced. ASA indicates American Society of Anesthesiologists; ASD, absolute standardized difference; IPTW, inverse probability of treatment weighting.

During the screening process, a total of 2238 potential episodes of clinically meaningful bradycardia, 180 cases of cardiac arrests, 391 cases with anaphylaxis, and 849 cases with bronchospasm were identified. The adjudicators individually assessed all patients who were given supplemental IV atropine, along with patients who were given epinephrine, methylprednisolone, or diphenhydramine via any route. Patients with documented chest compression, defibrillation, or cardioversion were also evaluated, as were patients in whom a diagnosis of anaphylactic reaction was documented.

A total of 459 cases were individually reviewed by the adjudicators. Assessments of the 3 initial adjudicators were consistent in 416 cases. There were also 41 cases that were reviewed by the chief adjudicator who agreed with the diagnosis of the 2 matching assessments. Just 2 cases required discussion by the adjudication committee, with consensus diagnoses being unanimously assigned. Ultimately, we identified 2004 cases of bradycardia, 14 cases of cardiac arrests, 6 cases of anaphylaxis, and 731 cases of bronchospasm.

The observed incidences of the composite outcome and individual side effects are summarized in Table 3. The incidence of the composite outcome was 3.4% in the sugammadex group and 3.0% in the neostigmine group. The most common individual postoperative side effect was bradycardia (2.4% in the sugammadex group versus 2.2% neostigmine). Noninferiority was not found, with the estimated OR of 1.21 (sugammadex versus neostigmine; 95% CI, 1.09–1.34), noninferiority P = .57; (Table 3; Figure 3). However, we found that neostigmine was superior to sugammadex with an estimated OR of 0.83 (0.74–0.92), 1-side superiority P < .001. The results of the sensitivity analysis are consistent with these findings.

(Enlarge Image)

Figure 3.

Forest plot for noninferiority testing. There were significantly more side effects with sugammadex than with neostigmine. CI indicates confidence interval.

1 2 3 4

Next Section

Table 1. Screening Criteria for Adjudication of Clinically Significant Adverse Effects of Reversal Agents
Adverse event	Screening criteria necessitating adjudication^a	Cases found	Cases reviewed	Events after adjudication
Bradycardia^a	Administration of IV atropine	28	Yes	2004
	Administration of IV glycopyrrolate	630	No
	Administration of epinephrine	140	Yes
	Administration of ephedrine	1440	No
Cardiac arrest	Documented chest compressions/defibrillation/cardioversion	40	Yes	14
Cardiac arrest	Administration of IV epinephrine	140	Yes	14
Anaphylaxis	Administration of epinephrine	140	Yes	6
	Administration of methylprednisolone	12	Yes
	Administration of diphenhydramine	238	Yes
	Documented diagnosis of anaphylactic/anaphylactoid reaction	1	Yes
Bronchospasm	Administration of albuterol	709	No	731
Bronchospasm	Administration of epinephrine	140	Yes	731

Abbreviation: IV, intravenous.
^aOnly considered, if neostigmine was coadministered with either glycopyrrolate or atropine, and an additional dosage of any medication was required afterward.

Table 2. Demographic and Baseline Characteristics Before and After the IPTW ^a
Variables	Sugammadex, n = 16,480	Neostigmine, n = 73,273	Unadjusted ASD^b	ASD after IPTW^b
Age (y)	57 ± 19	55 ± 19	0.093	0.015
Sex (female) (%)	50	52	0.031	0.024
Race (White) (%)	81	82	0.045	0.006
ASA status (%)
I	2	3	0.266	0.051
II	14	21
III	63	62
IV	21	13
V	0.11	0.09
Emergency (%)	4.2	3.1	0.058	0.011
Smoking (%)	52	48	0.082	0.013
Medical history (%)
Congestive heart failure	8	6	0.076	0.012
Vascular disease	7	5	0.082	0.008
Pulmonary circulation disease	1	1	0.029	0.009
Peripheral vascular disease	7	5	0.077	0.017
Hypertension	53	48	0.097	0.024
Paralysis	2	2	0.031	0.008
Other neurological disorders	7	6	0.052	0.016
Chronic pulmonary disease	16	11	0.135	0.022
Diabetes	20	18	0.068	0.014
Hypothyroidism	14	13	0.021	0.004
Renal failure	11	10	0.036	0.015
Liver disease	5	4	0.018	0.005
Peptic ulcer disease (bleeding)	1	1	0.008	0.008
AIDS	0	0	0.010	0.002
Lymphoma	1	1	0.007	0.000
Metastatic cancer	8	7	0.041	0.006
Solid tumor without metastasis	13	12	0.042	0.017
Rheumatoid arthritis/collagen vas	4	3	0.028	0.001
Coagulopathy	5	4	0.050	0.018
Obesity	20	19	0.015	0.015
Weight loss	12	10	0.073	0.027
Fluid and electrolyte disorders	16	13	0.062	0.028
Chronic blood loss anemia	1	1	0.001	0.011
Deficiency anemia	9	8	0.021	0.004
Alcohol abuse	1	1	0.014	0.008
Drug abuse	1	1	0.004	0.002
Psychoses	2	2	0.002	0.003
Depression	14	14	0.001	0.007
Duration of surgery (h)	2.3 (1.3–3.8)	3.1 (2.0–4.4)
Type of surgery: top 10 (column %)
Diagnostic bronchoscopy and biopsy of bronchus	8	4	0.772	0.227
Colorectal resection	3	5
Other OR therapeutic procedures on respiratory system	9	3
Laminectomy, excision intervertebral disk	3	4
Other OR lower GI therapeutic procedures	3	4
Conversion of cardiac rhythm	3	4
Other hernia repair	3	3
Hysterectomy, abdominal, and vaginal	2	3
Other OR therapeutic procedures on skin and breast	1	3
Spinal fusion	1	3

Summary statistics (presented as percentage of patients, mean ± SD, or median [Q1–Q3], respectively, for factors, symmetric continuous variables, and skewed continuous variables; duration of surgery was summarized by median [10th, 90th percentiles]).
Abbreviations: AIDS, acquired immune deficiency syndrome; ASA, American Society of Anesthesiologists; ASD, absolute standardized difference; GI, gastrointestinal; IPTW, inverse probability of treatment weighting; OR, odds ratio; SD, standard deviation.
^aIPTW: first, we estimated the probability of receiving sugammadex (ie, the propensity score) for each patient using logistic regression with receiving sugammadex as the outcome using the following prespecified potential confounding variables (listed in ). Then, we calculated the stabilized weights as follows: for the sugammadex group, weight was the proportion of sugammadex patients divided by the propensity score; for the neostigmine group, weight was the proportion of neostigmine patients divided by (1 – propensity score). To be conservative, we also truncated the stabilized weights at the 1st and 99th percentiles.
^bASD: absolute difference in means or proportions divided by the pooled standard deviation after weighting each observation with IPTW. Any covariable with an ASD >0.10 was considered to be imbalanced and would be adjusted for in the analysis.

Table 3. Noninferiority Analysis: Comparison Between Patients Who Received Sugammadex Versus Neostigmine on the Primary Composite Outcome Using IPTW ^a
Primary outcome	Sugammadex (N = 16,480) event (%)	Neostigmine (N = 73,273) event (%)	OR (95% CI)^b	Noninferiority^c P value^d	Superiority P value^d
Collapsed composite	558 (3.4)	2167 (3.0)
Sugammadex versus neostigmine (reference)			1.21 (1.09–1.34)	.57	NA
Neostigmine versus sugammadex (reference)			0.83 (0.74–0.92)	<.001	<.001
Individual components
Bradycardia	401 (2.4)	1603 (2.2)
Cardiac arrest	2 (0.01)	12 (0.02)
Anaphylaxis	3 (0.02)	3 (0.004)
Bronchospasm	161 (0.98)	570 (0.78)
Sensitivity analysis (adult patients with noncardiac surgery, superiority analysis)
	N = 13,065	N = 63,558
Collapsed composite	486 (3.7)	1937 (3.1)
Sugammadex versus neostigmine (reference)			1.28 (1.15–1.44)	.88	NA
Neostigmine versus sugammadex (reference)			0.78 (0.69–0.87)	<.001	<.001
Individual component
Bradycardia	353 (2.7)	1472 (2.3)
Cardiac arrest	2 (0.02)	9 (0.01)
Anaphylaxis	3 (0.02)	2 (0.00)
Bronchospasm	135 (1.03)	470 (0.74)

^aIPTW: First, we estimated the probability of receiving sugammadex (ie, the propensity score) for each patient using logistic regression with receiving sugammadex as the outcome using the following prespecified potential confounding variables (listed in ). Then, we calculated the stabilized weights as follows: for the sugammadex group, weight was the proportion of sugammadex patients divided by the propensity score; for the neostigmine group, weight was the proportion of neostigmine patients divided by (1 – propensity score). We also truncated the stabilized weights at the 1st and 99th percentiles.
^bThe OR of the composite outcome was estimated using a GEE model with IPTW and adjusted for the type of surgery.
^cNoninferiority delta is an OR of 1.2.
^d P value was from 1–1-tailed t test in the given direction.
Abbreviations: CI, confidence interval; GEE, generalized estimating equation; IPTW, inverse probability of treatment weighting; NA, not applicable; OR, odds ratio.

Authors and Disclosures

Kurt Ruetzler, MD, FAHA*†, Kai Li, MD*‡, Surendrasingh Chhabada, MD*§, Kamal Maheshwari, MD, MPH*†, Praveen Chahar, MD, FCARCSI*†||, Sandeep Khanna, MD*†¶, Marc T. Schmidt*, Dongsheng Yang, MS#, Alparslan Turan, MD*† and Daniel I. Sessler, MD*

Departments of *Outcomes Research and †General Anesthesiology, Anesthesiology Institute, Cleveland Clinic, Cleveland, Ohio; ‡Department of Anesthesia, China-Japan Union Hospital of Jilin University, Changchun, China; and Departments of §Pediatric Anesthesiology, ||Intensive Care and Resuscitation, and ¶Cardiothoracic Anesthesia, Anesthesiology Institute, Cleveland Clinic, Cleveland, Ohio; and #Department of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio

Address correspondence to
Kurt Ruetzler, MD, FAHA, Departments of Outcomes Research and General Anesthesiology, Anesthesiology Institute, Cleveland Clinic, 9500 Euclid Ave, P-77, Cleveland, OH 44195. Address e-mail to ruetzlk@ccf.org.

The authors declare no conflicts of interest.

Disclosures
Name: Kurt Ruetzler, MD, FAHA.
Contribution: This author helped with study design, data collection and interpretation, and writing the manuscript, and read and approved the final version of the manuscript.
Name: Kai Li, MD.
Contribution: This author helped with data collection and interpretation and writing the manuscript, and read and approved the final version of the manuscript.
Name: Surendrasingh Chhabada, MD.
Contribution: This author helped with data collection and interpretation and writing the manuscript, and read and approved the final version of the manuscript.Name: Kamal Maheshwari, MD, MPH.
Contribution: This author helped with study design, data collection and interpretation, and writing the manuscript, and read and approved the final version of the manuscript.
Name: Praveen Chahar, MD, FCARCSI.
Contribution: This author helped with data collection and interpretation and writing the manuscript, and read and approved the final version of the manuscript.
Name: Sandeep Khanna, MD.
Contribution: This author helped with data collection and interpretation and writing the manuscript, and read and approved the final version of the manuscript.
Name: Marc T. Schmidt, MD.
Contribution: This author helped with data collection and interpretation and writing the manuscript, and read and approved the final version of the manuscript.
Name: Dongsheng Yang, MS.
Contribution: This author helped with study design; data collection, interpretation, and analysis; and writing the manuscript; and read and approved the final version of the manuscript.
Name: Alparslan Turan, MD.
Contribution: This author helped with study design, data collection and interpretation, and writing the manuscript, and read and approved the final version of the manuscript.
Name: Daniel I. Sessler, MD.
Contribution: This author helped with study design, data collection and data interpretation, and writing the manuscript, and read and approved the final version of the manuscript.
This manuscript was handled by: Ken B. Johnson, MD.