Magnetic Liver Retraction Decreases Postoperative Pain and Length of Stay in Bariatric Surgery Compared to Nathanson Device

Abstract

Objective:

Retrospective case-matched comparison of magnetic liver retraction to a bedrail-mounted liver retractor in bariatric surgery specifically targeting short-term postoperative outcomes, including pain and resource utilization.

Background:

Retraction of the liver is essential to ensure appropriate visualization of the hiatus in bariatric surgery. Externally mounted retractors require a dedicated port or an additional incision. Magnetic devices provide effective liver retraction without the need of an incision.

Methods:

The sample consisted of primary and revisional bariatric surgery patients, including Roux-en-Y gastric bypass (RYGB), sleeve gastrectomy (SG), and biliopancreatic diversion with duodenal switch (BPD-DS) operations. Propensity score analysis was used to match patients with magnetic retraction to patients with a bedrail-mounted retractor with a 1:2 ratio using preoperative characteristics. Baseline characteristics and postprocedure outcomes were compared using two-sample t-tests or Wilcoxon rank sum tests and chi-square or Fisher's exact test as appropriate.

Results:

One hundred patients met inclusion criteria for the use of magnetic liver retraction (45 RYGB, 35 SG, 20 BPD-DS) with 196 suitable matched external retractor patients identified. Patients were matched and comparable for all preoperative characteristics except for transversus abdominus plane block (27% versus 47%). Patients in the magnet cohort had significantly decreased mean 12-hour postoperative pain scores (2.9 versus 4.2, P = .004) and decreased hospital length of stay (LOS) (1.5 versus 1.9 days, P = .005) while operating room supply were higher in the magnet cohort ($4600 versus $4213, P = .0001).

Conclusions:

Magnetic liver retraction in bariatric surgery is associated with decreased postoperative pain scores, decreased hospital LOS, and increased operating supply costs.

Introduction

Obesity is the most prevalent chronic disease and a leading cause of morbidity and mortality in the United States. As of 2016, more than one-third of the U.S. population meets criteria for a diagnosis of obesity.¹ Bariatric surgery remains the most effective treatment for moderate to severe obesity and resolution of obesity-related comorbidities.^2–6 Bariatric operations are quickly becoming more common and are among the most performed operations in the United States.⁷

Retraction of the liver is essential for appropriate visualization of the gastroesophageal junction in bariatric surgery and many devices of various designs have been developed and used for this purpose.⁸ Generally, these devices are mounted to the bedrail of the operative table and require a dedicated port or incision. Less common methods such as suturing, suction devices, and slings have been described with varying degrees of effectiveness.^9,10

Magnetic retractors provide effective liver retraction without the need of a dedicated port or incision.¹¹ The Magnetic Surgical System (Levita Magnetics, San Mateo, CA) is the first magnetic retractor to receive Food and Drug Administration approval and recent studies have demonstrated safe and effective use in laparoscopic surgery.^12–14 Additional reports of application in bariatric surgery demonstrate safe and effective liver retraction.¹¹ These previous pilot studies provide promising information regarding the use of magnetic retractors; however, they lack head-to-head comparison to more common methods. This study aims to assess the difference in intra- and postoperative outcomes between a magnetic surgical system and an externally mounted Nathanson liver retractor.

Methods

A retrospective review of laparoscopic bariatric cases, including Roux-en-Y gastric bypass (RYGB), sleeve gastrectomy (SG), and biliopancreatic diversion with duodenal switch (BPD-DS) performed at our center where the Magnetic Surgical System was used for liver retraction (Fig. 1). All cases were consecutive and the type of retraction was selected by surgeon preference. Patients in the magnet group were matched to control patients from the same time period encompassing 23 months from October 2016 to August 2018. All patients were treated with the same perioperative enhanced recovery analgesia protocol (Table 1).

FIG. 1.

Magnetic retractor system. (A) Grasper handle. (B) Detachable tip. (C) External controller.

Table 1.

Institutional Perioperative Enhanced Recovery Analgesia Protocol

Preoperative	Gabapentin 800 mg
	Acetaminophen 1 g IV
	Ketorolac 30 mg IV if GFR >60 (or 15 mg if age >60)
Intraoperative	Goal of minimal to zero narcotics
	Multimodal analgesia
	Ultrasound guided TAP block with liposomal bupivacaine OR
	Port-site infiltration with liposomal bupivacaine
	Ketamine after induction
Postoperative	Acetaminophen 650 mg every 6 hours
	Ketorolac 15 mg IV every 6 hours
	Gabapentin 300 mg every 6 hours
	Oxycodone 5–15 every 4–6 hours as needed
	Limited additional narcotics only as needed

GFR, glomerular filtration rate; IV, intravenous; OR, operating room; TAP, transverses abdominis plane.

Patients that were <18 years of age and that had previous surgery within 30 days of the target procedure were excluded from the analysis. Opioid medication administration was converted to morphine milligram equivalents (MME). Liver retraction in the control group was performed using an externally mounted Nathanson retractor placed through a separate subxiphoid incision.

The magnet cohort patients were matched to control patients in whom externally mounted liver retraction was applied using propensity matching to reduce the amount of confounding from the selected variables. The likelihood of patients being in the magnetic retraction arm (versus the control) is predicted in a logistic regression model using the characteristics: gender, age, race, baseline body mass index (BMI), procedure type, revision status, technique, and comorbidities (hypertension, diabetes, gastroesophageal reflux, and hyperlipidemia). The cases (subjects with magnetic retraction) and controls were matched in the ratio 1:2, respectively, on the proximity of the (propensity) scores from the above logistic regression with a caliper width of 0.1.

Tests of significance of baseline characteristics and postprocedure outcomes were compared using Wilcoxon rank sum tests for continuous variables and chi-square test for categorical variables. A P-value <.05 was used to consider statistical significance of baseline characteristics.

Results

A total of 113 patients in the magnet cohort were identified from the medical records. One hundred had sufficient data and met inclusion criteria, resulting in a cohort of 45 RYGB, 35 SG, and 20 BPD-DS patients. One hundred ninety-six matched control patients were identified. Complete preoperative characteristics are described in Table 2. Patients were appropriately matched and comparable for all preoperative characteristics except for transversus abdominus plane (TAP) block, as fewer patients in the magnet cohort received treatment with a TAP block (27% versus 47%, P < .001). The mean preoperative BMI was 46.1 kg/m² in the magnet group and 46.0 kg/m² in the control group.

Table 2.

Baseline Characteristics for Matched Patients—by Magnet Use Versus Controls

	Control (n = 196)	Magnet (n = 100)	Total (N = 296)	P
Procedure				.6643^a
RYGB	92 (46.9%)	45 (45.0%)	137 (46.3%)
Sleeve	73 (37.2%)	35 (35.0%)	108 (36.5%)
Switch	31 (15.8%)	20 (20.0%)	51 (17.2%)
Technique				.9769^a
Lap-Robotic	6 (3.1%)	3 (3.0%)	9 (3.0%)
Laparoscopic	190 (96.9%)	97 (97.0%)	287 (97.0%)
Revision				.6880^a
Yes	15 (7.7%)	9 (9.0%)	24 (8.1%)
Age				.9911^a
18 to <30	29 (14.8%)	14 (14.0%)	43 (14.5%)
30 to <40	42 (21.4%)	20 (20.0%)	62 (20.9%)
40 to <50	52 (26.5%)	29 (29.0%)	81 (27.4%)
50 to <60	54 (27.6%)	28 (28.0%)	82 (27.7%)
60 to <70	19 (9.7%)	9 (9.0%)	28 (9.5%)
Gender				.5074^a
Female	172 (87.8%)	85 (85.0%)	257 (86.8%)
Race				.8938^a
White	111 (56.6%)	56 (56.0%)	167 (56.4%)
Black or African American	74 (37.8%)	37 (37.0%)	111 (37.5%)
Other	11 (5.6%)	7 (7.0%)	18 (6.1%)
BMI at baseline				.8672^b
Mean (SD)	46.0 (7.4)	46.1 (7.9)	46.0 (7.6)
Range	32.8–68.5	34.3–75.5	32.8–75.5
Diabetes				.9613^a
Yes	70 (35.7%)	36 (36.0%)	106 (35.8%)
GERD				.4777^a
Yes	40 (20.4%)	24 (24.0%)	64 (21.6%)
Hypertension				.6662^a
Yes	111 (56.6%)	54 (54.0%)	165 (55.7%)
Hyperlipidemia				.8512^a
Yes	47 (24.0%)	23 (23.0%)	70 (23.6%)
COPD				.1072^a
Yes	5 (2.6%)	0 (0.0%)	5 (1.7%)
Renal insufficiency				.8256^a
Yes	11 (5.6%)	5 (5.0%)	16 (5.4%)
Previous foregut surgery				.3426^a
Yes	18 (9.2%)	6 (6.0%)	24 (8.1%)
Current smoker				.7084^a
Yes	3 (1.5%)	1 (1.0%)	4 (1.4%)
Steroid/immunosuppression				.5556^a
Yes	7 (3.6%)	5 (5.0%)	12 (4.1%)
ASA class				.7482^a
2	70 (35.7%)	39 (39.0%)	109 (36.8%)
3	125 (63.8%)	60 (60.0%)	185 (62.5%)
4	1 (0.5%)	1 (1.0%)	2 (0.7%)
TAP block				.0007^a
Yes	93 (47.4%)	27 (27.0%)	120 (40.5%)

Chi-square.

Wilcoxon.

ASA, American Society of Anesthesiologists; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); COPD, chronic obstructive pulmonary disease; GERD, gastroesophageal reflux disease; RYGB, Roux-en-Y gastric bypass; SD, standard deviation; TAP, transversus abdominis plane.

The magnet cohort demonstrated decreased mean 12-hour postoperative pain scores compared to control (mean: 2.9, standard deviation [SD]: 2.8 versus mean: 4.2, SD: 3.1, P = .004). The magnet group also had a shorter hospital length of stay (LOS) compared to control (mean: 1.5 days, SD: 0.9 versus mean: 1.9, SD: 1.3, P = .005). Operating room (OR) supply costs were higher in the magnet cohort (mean: $4600, SD: $916.6 versus mean: $4213, SD: $845.2, P = .0001). No significant differences in 30-day postoperative complications such as intensive care unit admission, blood transfusion, readmission, or mortality were found. Complete outcome results of the included variables are listed in Table 3.

Table 3.

Outcomes for All Patients by Magnet Use Versus Controls

	Control (n = 196)	Magnet (n = 100)	Total (N = 296)	P
1 hour pain score				.6837^a
Mean (SD)	3.7 (3.1)	3.5 (3.1)	3.7 (3.1)
6 hours pain score				.8464^a
Mean (SD)	4.1 (3.2)	4.0 (3.2)	4.1 (3.2)
12 hours pain score				.0040^a
Mean (SD)	4.2 (3.1)	2.9 (2.8)	3.8 (3.1)
24 hours opioid use (MME)				.0719^a
Mean (SD)	6.8 (7.8)	4.5 (5.1)	6.1 (7.1)
Blood Loss				.7583^a
Mean (SD)	19.5 (19.7)	23.3 (30.5)	20.8 (23.9)
Blood transfusion				.3108^b
Yes	2 (1.0%)	0 (0.0%)	2 (0.7%)
Myocardial infarction				NA
Yes	0 (0.0%)	0 (0.0%)	(0.0%)
Pneumonia				.6266^b
Yes	1 (0.5%)	1 (1.0%)	2 (0.7%)
Postop ventilator use (48 hours)				NA
Yes	0 (0.0%)	0 (0.0%)	(0.0%)
Unplanned intubation				NA
Yes	0 (0.0%)	0 (0.0%)	(0.0%)
ICU admission				.1503^b
Yes	4 (2.0%)	0 (0.0%)	4 (1.4%)
Surgical site infection				.8217^b
Yes	5 (2.6%)	3 (3.0%)	8 (2.7%)
Readmission				.3166^b
Yes	11 (5.6%)	3 (3.0%)	14 (4.7%)
Mortality				NA
Yes	0 (0.0%)	0 (0.0%)	(0.0%)
Operative time (minutes)				.4813^a
Mean (SD)	173.5 (65.7)	179.1 (67.9)	175.4 (66.4)
Hospital LOS (days)				.0051^a
Mean (SD)	1.9 (1.3)	1.5 (0.9)	1.8 (1.2)
OR supply costs				.0001^a
Mean (SD)	4212.5 (845.2)	4599.9 (916.6)	4340.8 (886.9)
Total hospital charges				.0877^a
Mean (SD)	48676.9 (10151.0)	49712.7 (9444.9)	49026.8 (9914.2)

Wilcoxon.

Chi-square.

ICU, intensive care unit; LOS, length of stay; MME, morphine milligram equivalents; NA, not applicable; OR, operating room; SD, standard deviation.

Subset analyses

Analyses by specific procedure type found that the increased OR supply costs and hospital LOS were maintained in RYGB and SG, but not in BPD-DS. Complete details are available in Table 4. In addition, the difference in 12-hour postoperative pain score was not demonstrated in the RYGB group, unlike the SG and BPD-DS groups. Two additional subanalyses were performed; (1) removal of patients who received TAP block and (2) removal of patients who had a revision or received a robotic procedure. No difference in pain scores or postoperative opioid pain medication use was found when patients who received a TAP block were excluded (Table 5). Among patients who received only primary laparoscopic procedures, a more pronounced reduction in postoperative opioid medication use for cases compared to controls was observed (mean: 4.4 MME, SD: 4.6 versus mean: 6.9 MME, SD: 7.4, P = .051). Complete outcome results for this subset are listed in Table 6.

Table 4.

Select Outcomes by Procedure Type for Magnet Use Versus Controls

	RYGB			SG			BPD-DS
	Control (n = 92)	Magnet (n = 45)	P	Control (n = 73)	Magnet (n = 35)	P	Control (n = 31)	Magnet (n = 20)	P
1 hour pain score
Mean (SD)	3.8 (3.1)	3.2 (3.0)	.4287^a	3.7 (2.9)	3.8 (3.1)	.8627^a	3.7 (3.5)	3.8 (3.1)	.9008^a
6 hours pain score
Mean (SD)	4.0 (3.0)	3.6 (3.1)	.4855^a	4.2 (3.2)	4.3 (3.2)	.9914^a	4.2 (3.7)	4.6 (3.2)	.6709^a
12 hours pain score
Mean (SD)	4.2 (3.2)	3.2 (2.7)	.1730^a	3.9 (2.8)	2.5 (2.7)	.0474^a	5.8 (3.4)	3.1 (3.0)	.0390^a
24 hours opioid use (MME)
Mean (SD)	7.0 (7.6)	4.6 (6.0)	.0855^a	6.3 (8.0)	4.6 (4.2)	.6632^a	7.5 (8.1)	4.4 (4.2)	.4423^a
Blood loss (mL)
Mean (SD)	18.0 (14.8)	20.0 (22.2)	.8992^a	17.7 (17.4)	12.1 (11.3)	.1280^a	28.1 (32.2)	50.3 (49.5)	.0509^a
Blood transfusion
Yes	2 (2.2%)	0 (0.0%)	.3191^b	0 (0.0%)	0 (0.0%)	NA	0 (0.0%)	0 (0.0%)	NA
Pneumonia
Yes	1 (1.1%)	0 (0.0%)	.4827^b	0 (0.0%)	0 (0.0%)	NA	0 (0.0%)	1 (5.0%)	.2086^b
Surgical site infection
Yes	3 (3.3%)	1 (2.2%)	.7345^b	1 (1.4%)	1 (2.9%)	.5916^b	1 (3.2%)	1 (5.0%)	.7500^b
Readmission
Yes	6 (6.5%)	1 (2.2%)	.2831^b	2 (2.7%)	0 (0.0%)	.3229^b	3 (9.7%)	2 (10.0%)	.9698^b
Operative time (minutes)
Mean (SD)	179 (48)	182 (42)	.5361^a	126 (38)	123 (37)	.5768^a	269 (54)	270 (56)	.8849^a
Hospital LOS (days)
Mean (SD)	1.5 (0.8)	1.2 (0.4)	.0022^a	1.6 (0.7)	1.3 (0.6)	.0244^a	3.7 (1.9)	2.8 (1.0)	.0610^a
OR supply costs (dollars)
Mean (SD)	3981 (583)	4276 (621)	.0067^a	4094 (647)	4314 (655)	.0480^a	5179.9 (1200.1)	5782.6 (889.5)	.0654^a
Hospital charges (dollars)
Mean (SD)	46,266 (6698)	46,954 (4873)	.2151^a	45,305 (5437)	45,526 (5196)	.5859^a	63,774 (13,481)	63,248 (10,878)	.9923^a

Wilcoxon.

Chi-square.

BPD-DS, biliopancreatic diversion with duodenal switch; LOS, length of stay; MME, morphine milligram equivalents; NA, not applicable; OR, operating room; RYGB, Roux-en-Y gastric bypass; SD, standard deviation; SG, sleeve gastrectomy.

Table 5.

Outcomes for No Transversus Abdominus Plane Sample by Magnet Use Versus Controls

	Control (n = 103)	Magnet (n = 73)	Total (N = 176)	P
1 hour pain score				.9523^a
Mean (SD)	3.7 (3.1)	3.6 (2.9)	3.7 (3.1)
Median	4.0	4.0	4.0
Range	0.0–10.0	0.0–9.0	0.0–10.0
6 hours pain score				.9286^a
Mean (SD)	4.4 (3.2)	4.4 (3.2)	4.4 (3.2)
Median	5.0	4.0	5.0
Range	0.0–10.0	0.0–10.0	0.0–10.0
12 hours pain score				.0048^a
Mean (SD)	4.6 (3.3)	3.0 (2.8)	3.9 (3.2)
Median	5.0	3.0	4.0
Range	0.0–10.0	0.0–8.0	0.0–10.0
24 hours opioid use (MME)				.1263^a
Mean (SD)	7.8 (8.1)	5.3 (5.6)	6.7 (7.3)
Median	5.2	3.5	5.0
Range	0.0–36.0	0.0–27.9	0.0–36.0

Wilcoxon.

MME, morphine milligram equivalents; SD, standard deviation.

Table 6.

Outcomes by Primary Laparoscopic Cases Only

	Control (n = 177)	Magnet (n = 89)	Total (n = 266)	P
1 hour pain score				.9847^a
N	172	85	257
Mean (SD)	3.6 (3.0)	3.6 (3.1)	3.6 (3.0)
6 hours pain score				.4005^a
N	128	70	198
Mean (SD)	4.3 (3.1)	4.0 (3.2)	4.2 (3.1)
12 hours pain score				.0034^a
N	126	69	195
Mean (SD)	4.2 (3.2)	2.8 (2.8)	3.7 (3.1)
24 hours opioid use (MME)				.0512^a
N	177	89	266
Mean (SD)	6.9 (7.4)	4.4 (4.6)	6.0 (6.7)
Blood Loss				.9023^a
Mean (SD)	19.2 (18.4)	21.6 (28.7)	20.0 (22.4)
Blood transfusion				.3141^b
Yes	2 (1.1%)	0 (0.0%)	2 (0.8%)
Myocardial infarction				NA
Yes	0 (0.0%)	0 (0.0%)	0 (0.0%)
Pneumonia				.6187^b
Yes	1 (0.6%)	1 (1.1%)	2 (0.8%)
Post-op ventilator use (48 hours)				NA
Yes	0 (0.0%)	0 (0.0%)	0 (0.0%)
Unplanned intubation				NA
Yes	0 (0.0%)	0 (0.0%)	0 (0.0%)
ICU admission				.2168^b
Yes	3 (1.7%)	0 (0.0%)	3 (1.1%)
Surgical site infection				.5933^b
Yes	4 (2.3%)	3 (3.4%)	7 (2.6%)
Readmission				.6570^b
Yes	8 (4.5%)	3 (3.4%)	11 (4.1%)
Mortality				NA
Yes	0 (0.0%)	0 (0.0%)	0 (0.0%)
Operative time (minutes)				.1882^a
Mean (SD)	165.7 (61.0)	178.0 (68.9)	169.8 (63.9)
Hospital LOS (days)				.0260^a
Mean (SD)	1.8 (1.2)	1.5 (0.9)	1.7 (1.1)
OR supply costs				.0003^a
Mean (SD)	4195.6 (763.2)	4596.6 (936.1)	4328.8 (844.3)
Total hospital charges				.0265^a
Mean (SD)	47846.6 (9004.4)	49872.7 (9791.9)	48524.5 (9306.2)

Wilcoxon.

Chi-square.

ICU, intensive care unit; LOS, length of stay; MME, morphine milligram equivalents; NA, not applicable; OR, operating room; SD, standard deviation.

Discussion

Liver retraction is essential for bariatric surgery and other foregut operations for proper visualization of the diaphragmatic hiatus. This comparative study demonstrates that incisionless magnetic liver retraction is not only safe but also superior to a traditional mounted retractor in several aspects. Nathanson retractors require an additional incision, typically just below the xiphoid process, and are statically mounted to the bedrail. This orientation often leads to torque and additional force on the abdominal wall. The magnetic retractor is introduced through an existing 12 mm port and attached directly to the liver edge (Fig. 2). The magnetic tip is coupled to the external controller to provide dynamic liver retraction and good exposure of the gastroesophageal junction without the need of an extra incision. It can be easily adjusted without any stress on the abdominal wall, effectively eliminating pain at that site. The quality of retraction from the magnetic retractor was always adequate and did not require combined use with a table-mounted retractor.

FIG. 2.

Magnetic liver retraction. Retraction of left lobe exposing gastroesophageal hiatus (A). Removal from liver demonstrating minimal tissue trauma (B).

If visualization is not adequate, a second magnetic tip can be placed to provide additional retraction; however, this was never required in this cohort. Of note, there were no significant differences in intraoperative or postoperative complications supporting the safe use of the magnet retractor. Even though the magnetic retractor has potential to tear a small portion of tissue when clamped, in our experience, we did not have any significant tear or damage to the liver in our bariatric patients where nonalcoholic fatty liver disease is more prevalent.¹¹

Postoperative pain is a major issue with reports of inadequate control in >80% of surgical patients in the United States.¹⁵ Numerical Rating Scales have been validated in several studies and are easy for patients to quickly and accurately express their current pain.^16–19 Our study found that postoperative pain scores using a standard Numerical Rating Scale of 0–10 were decreased 12 hours after surgery with patients in the magnet group reporting a mean pain score of 2.9 compared to 4.2 in the control group. This represents 31% lower scores and a 20% change in score is widely regarded as being clinically significant.^{19, 20} The absence of difference in 1- and 6-hour scores is likely related to residual effects of local and general anesthetics and possibly too early in the recovery period for patients to have ambulated and reestablished activity.

The increased interest in decreasing the use of opioid analgesics in postoperative patients is largely driven by the outcomes of abuse and death. Several studies suggest that surgery is associated with an increased risk of long-term opioid use leading to a phenomenon known as persistent postoperative opioid use.^21,22 While 90% of surgical patients are opioid-naive, estimates suggest that 49%–95% are discharged home with an opioid prescription^23–25 and 6% continue to use the medication beyond the normal recovery period.²¹

Efforts to reduce postoperative opioid use are to be utilized as they can have a direct effect on opioid use at the population level in addition to individual effects on the patient.^26,27 Mean postoperative opioid use for all patients in the magnet group was 4.5 MME compared to 6.8 MME in the control, representing a 34% decrease (P = .0719). In a subanalysis excluding robot-assisted and revision cases, the difference increased to 36% (P = .0512). These findings maintain clinical import, as the majority of bariatric operations performed are primary cases with a laparoscopic approach.

The control group had a significantly larger proportion of patients who received a TAP block, which one would expect to decrease pain scores and subsequently opioid use, but this was not demonstrated in the findings. When pain scores and opioid use were compared among only those patients without TAP blocks, no significant differences were observed. This may be explained by the preoperative enhanced recovery protocol requiring that all patients be given either a TAP block or local port site infiltration with 266 mg liposomal bupivacaine. The benefits of a TAP block are well known,^28,29 but may have been lost when compared to infiltration of port site only.^30–33 Nonetheless, the effects of a TAP block on postoperative pain scores and opioid use appear to be insignificant in this cohort.

As expected, the total OR supply costs were increased in the magnet group compared to the control group, likely due to the costs of the single-use magnet device. This small difference is likely offset by cost savings related to decreased LOS, which can range from $2000 to $2500 per day depending on organization type.³⁴

Hospital LOS was decreased in the combined magnet group compared to the control. Decreased LOS allows for improved resource utilization and not only results in improved cost savings but also frees up resources for other patients in need of care. Additional savings from ancillary services such as nutrition and housekeeping play a part, but are difficult to measure. Similar decreases in LOS were found in the procedure specific subanalysis except with BPD-DS (Table 4).

Our study is primarily limited by its retrospective design and the sample that was available. Possible confounding of demographic and comorbidity variables was minimized by using propensity score matching. The relatively small sample size increases the chance of Type II error (i.e., smaller samples have lower statistical power), especially in the BPD-DS subanalysis.

As experience with this novel device increases among surgeons, further benefits may emerge and more patients will be available to achieve greater statistical power for further studies, including the possibility of a randomized control trial to confirm our findings with more rigor and validity. We limited our analysis to short-term perioperative outcomes for logistic reasons and because this device should not have bearing on typical long-term outcome such as weight loss. Nonetheless, extending the follow-up period to include long-term outcomes such as port-site hernias may be worth exploring.

Conclusion

Magnetic liver retraction in bariatric surgery is safe and effective. When the device was compared in a combined all-procedure cohort, we found not only decreased 12-hour pain score and decreased LOS but also an increased OR supply costs. There is a trend toward decreased postoperative opioid use that may be better explored in a larger prospective study.

Footnotes

Disclaimer

Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NCATS or NIH.

Disclosure Statement

D.P. has received a research support from Medtronic, education grant from Levita and Gore, is a speaker Nova Nordisk, and is a consultant for Medtronic and Intuitive. A.D.G. is a consultant for Levita and Biom'Up, and a speaker for Gore and Medtronic. The other authors declare that they have no conflict of interest.

Funding Information

The Duke BERD Methods Core's support of this project was made possible (in part) by Grant No. UL1TR002553 from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research.

References

Hales

, Carroll

, Fryar

, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief, 2017; (288):1–8.

Currie

, Chetwood

, Ahmed

. Bariatric surgery and renal function. Obes Surg, 2011; 21:528–539.

Nostedt

, Switzer

, Gill

, et al. The effect of bariatric surgery on the spectrum of fatty liver disease. Can J Gastroenterol Hepatol, 2016; 2016:2059245.

Sarkhosh

, Switzer

, El-Hadi

, et al. The impact of bariatric surgery on obstructive sleep apnea: A systematic review. Obes Surg, 2013; 23:414–423.

Tan

, Carr

. The impact of bariatric surgery on obesity-related infertility and in vitro fertilization outcomes. Semin Reprod Med, 2012; 30:517–528.

Yska

, van Roon

, de Boer

, et al. Remission of type 2 diabetes mellitus in patients after different types of bariatric surgery: A population-based cohort study in the United Kingdom. JAMA Surg, 2015; 150:1126–1133.

Nguyen

, Masoomi

, Magno

, et al. Trends in use of bariatric surgery, 2003–2008. J Am Coll Surg, 2011; 213:261–266.

Vargas-Palacios

, Hulme

, Veale

, et al. Systematic review of retraction devices for laparoscopic surgery. Surg Innov, 2016; 23:90–101.

Benzing

, Weiss

, Krenzien

, et al. Intra-abdominal trocar-free vacuum liver retractor for upper-gastrointestinal surgery. Surg Innov, 2017; 24:186–191.

10.

Madnani

, Patel

, Gupta

, et al. Novel trocarless, scarless technique for left lobe liver retraction in laparoscopic upper gastrointestinal surgeries: Simple, cost-effective and with better cosmesis. Indian J Surg, 2015; 77,(Suppl 3):1441–1443.

11.

Davis

, Davalos

, Ortega

, et al. Magnetic liver retraction: An incision-less approach for less invasive bariatric surgery. Obes Surg, 2019; 29:1068–1073.

12.

Rivas

, Robles

, Riquelme

, et al. Magnetic surgery: Results from first prospective clinical trial in 50 patients. Ann Surg, 2018; 267:88–93.

13.

Haskins

, Strong

, Allemang

, et al. Magnetic surgery: First U.S. experience with a novel device. Surg Endosc, 2018; 32:895–899.

14.

Guerron

, Ortega

, Park

, et al. Magnetic robot-assisted single-incision cholecystectomy. CRSLS, 2017; e2017.00073. DOI: 10.4293/CRSLS.2017.00073.

15.

Gan

. Poorly controlled postoperative pain: Prevalence, consequences, and prevention. J Pain Res, 2017; 10:2287–2298.

16.

Jensen

, Karoly

, Braver

. The measurement of clinical pain intensity: A comparison of six methods. Pain, 1986; 27:117–126.

17.

Kremer

, Atkinson

, Ignelzi

. Measurement of pain: Patient preference does not confound pain measurement. Pain, 1981; 10:241–248.

18.

Closs

, Barr

, Briggs

, et al. A comparison of five pain assessment scales for nursing home residents with varying degrees of cognitive impairment. J Pain Symptom Manage, 2004; 27:196–205.

19.

Farrar

, Young

Jr ., LaMoreaux

, et al. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain, 2001; 94:149–158.

20.

Farrar

, Portenoy

, Berlin

, et al. Defining the clinically important difference in pain outcome measures. Pain, 2000; 88:287–294.

21.

Brummett

, Waljee

, Goesling

, et al. New persistent opioid use after minor and major surgical procedures in US adults. JAMA Surg, 2017; 152:e170504.

22.

Hah

, Bateman

, Ratliff

, et al. Chronic opioid use after surgery: Implications for perioperative management in the face of the opioid epidemic. Anesth Analg, 2017; 125:1733–1740.

23.

Clarke

, Soneji

, Ko

, et al. Rates and risk factors for prolonged opioid use after major surgery: Population based cohort study. BMJ, 2014; 348:g1251.

24.

Fujii

, Hodges

, Russell

, et al. Post-discharge opioid prescribing and use after common surgical procedure. J Am Coll Surg, 2018; 226:1004–1012.

25.

Nooromid

, Blay

Jr. , Holl

, et al. Discharge prescription patterns of opioid and nonopioid analgesics after common surgical procedures. Pain Rep, 2018; 3:e637.

26.

Bicket

, Long

, Pronovost

, et al. Prescription opioid analgesics commonly unused after surgery: A systematic review. JAMA Surg, 2017; 152:1066–1071.

27.

Bartels

, Mayes

, Dingmann

, et al. Opioid use and storage patterns by patients after hospital discharge following surgery. PLoS One, 2016; 11:e0147972.

28.

Mittal

, Dey

, Siddhartha

, et al. Efficacy of ultrasound-guided transversus abdominis plane (TAP) block for postoperative analgesia in laparoscopic gastric sleeve resection: A randomized single blinded case control study. Surg Endosc, 2018; 32:4985–4989.

29.

Johns

, O'Neill

, Ventham

, et al. Clinical effectiveness of transversus abdominis plane (TAP) block in abdominal surgery: A systematic review and meta-analysis. Colorectal Dis, 2012; 14:e635–e642.

30.

Kargar

, Minas

, Gorgin-Karaji

, et al. Transversus abdominis plane block under laparoscopic guide versus port-site local anaesthetic infiltration in laparoscopic excision of endometriosis: A double-blind randomised placebo-controlled trial. BJOG, 2019; 126:647–654.

31.

Rashid

, Gorissen

, Ris

, et al. No benefit of ultrasound-guided transversus abdominis plane blocks over wound infiltration with local anaesthetic in elective laparoscopic colonic surgery: Results of a double-blind randomized controlled trial. Colorectal Dis, 2017; 19:681–689.

32.

El Hachem

, Small

, Chung

, et al. Randomized controlled double-blind trial of transversus abdominis plane block versus trocar site infiltration in gynecologic laparoscopy. Am J Obstet Gynecol, 2015; 212:182.e1–e9.

33.

, Long

, Lujan-Hernandez

, et al. Transversus abdominis-plane block versus local anesthetic wound infiltration in lower abdominal surgery: A systematic review and meta-analysis of randomized controlled trials. BMC Anesthesiol, 2014; 14:121.

34.

Hospital Adjusted Expenses per Inpatient Day by Ownership

March 07

, 2019. https://www.kff.org/health-costs/state-indicator/expenses-per-inpatient-day-by-ownership/?currentTimeframe=0&sortModel=%7B%22colId%22:%22Location%22,%22sort%22:%22asc%22%7D (accessed July 23, 2019).