Systematic Nasogastric Tube in Aortic Surgery: Is It Necessary? Manuela M. Hern andez Mateo, Isaac Martı´nez-Lopez, Manuel Hernando-Rydings, Sergio Revuelta-Suero, Pablo Marques de Marino, and Francisco J. Serrano Hernando, Madrid, Spain
Background: Routine nasogastric tube (NGT) decompression has been traditionally used to prevent nausea and vomiting after abdominal surgery. Besides, many studies having demonstrated no benefits derived from this practice after an elective laparotomy, little evidence has been published regarding its use in aortic surgery. In this study, we analyze the effects of the selective use of the NGT in patients undergoing infrarenal aortic surgery in our center. Methods: Prospective cohort study including patients who underwent elective infrarenal aortic surgery between January 2011 and December 2012. Patients were prospectively included in group A (systematic NGT placement) and group B (nonsystematic NGT). The main end point was the occurrence of postoperative nausea and vomiting (PONV). Secondary end points were postoperative complications, time to first oral intake, and hospital stay. Results: One hundred patients were finally included in the study, 50 patients per group. Preoperative and intraoperative data were similar between both groups. Higher incidence of PONV (48% vs. 10%; relative risk, 2.4; 95% confidence interval [CI], 1.3e4.5; P ¼ 0.003) was observed in group A. Selective NGT behaved as a protective factor regarding earlier first oral intake in first postoperative 48 hours (hazard ratio, 0.67; 95% CI, 0.45e0.99; P ¼ 0.05). There were no differences in other adverse events although a trend toward fewer respiratory complications was observed in patients with nonsystematic NGT. Conclusions: This study demonstrates higher incidence of PONV and longer time to first oral intake in patients with systematic NGT with no benefits derived from this practice. Based on these results, selective NGT decompression should be encouraged in patients undergoing infrarenal aortic surgery.
INTRODUCTION Over the last 100 years, nasogastric tube (NGT) decompression after abdominal surgery has been used to prevent postoperative nausea and vomiting (PONV), abdominal distension, pulmonary complications, and wound dehiscence.1 However, despite this longstanding surgical tradition, many authors have published better results for the selective use
Department of Angiology and Vascular Surgery, Hospital Clı´nico San Carlos, Madrid, Spain. Correspondence to: Manuela M. Hernandez Mateo, MD, Department of Angiology and Vascular Surgery, Hospital Clı´nico San Carlos, Martı´n Lagos, St. Madrid, Spain; E-mail: [email protected] Ann Vasc Surg 2015; 29: 786–791 http://dx.doi.org/10.1016/j.avsg.2015.01.002 Ó 2015 Elsevier Inc. All rights reserved. Manuscript received: April 2, 2014; manuscript accepted: January 1, 2015; published online: March 4, 2015.
786
of the NGT in terms of postoperative complications after an elective laparotomy compared with its systematic placement.1e5 Many of these studies have been published according to results obtained on colorectal, gynecologic, and urologic interventions.1e3 Nevertheless, little evidence has been published regarding the use of the NGT in aortic surgery. Only 2 studies have evaluated NGT use in this field.4,5 Both of them state that routine NGT use is not beneficial even detrimental regarding postoperative lung complications and longer hospital stay. Thus, based on these results, the Society of Vascular Surgery questioned this practice on its latest practice guidelines.6 Based on the reported literature, the purpose of this study was to analyze the results derived from the selective use of NGT in open aortic surgery in our center.
Vol. 29, No. 4, May 2015
PATIENTS AND METHODS Study Design This is a prospective cohort study. One hundred and fifty-three consecutive patients referred to our institution underwent open surgery to treat infrarenal aortic aneurysm or aortoiliac occlusive disease between January 2011 and December 2012. Excluded from this study were patients with emergent or urgent surgery (ruptured or symptomatic aortic aneurysms), previous esophagogastric or pancreatoduodenal surgery, gastroesophageal reflux or hiatal hernia, retroperitoneal approach, accidental intraoperative bowel injury, adhesiolisis for more than 1 hr, postoperative mechanical ventilation for more than 12 hr, or postoperative consciousness disorders (Glasgow score 38 C) were observed in patients with routine NGT although these did not reach statistical significance (RR, 5; 95% CI, 0.6e41.3; P ¼ 0.2 and RR, 6; 95% CI, 0.7e48.05; P ¼ 0.1, respectively).
Annals of Vascular Surgery
Three patients with no systematic NGT needed the implantation of an NGT during the postoperative course because of vomiting in the first 24 hr in 1 patient and because of the presence of postoperative ileus in the remaining 2 patients. Reintroduction of the NGT was also needed in 1 patient with previous routine NGT because of postoperative ileus. Only 1 patient died, in group A, due to multiorgan failure after colonic ischemia, events not related to the use of NGT. Also, compared with patients with no NGT, those with routine insertion of an NGT had a longer mean hospital stay (10.72 days vs. 7.36 days, P ¼ 0.02) despite no differences were found in median hospital stay, adjusted due to longer stay of the previous mentioned patient who died due to multiorgan failure. In a subgroup analysis based on the treated pathology, differences obtained regarding PONV and first oral intake were only present in patients treated for aneurysm. The incidence of PONV was also greater in patients treated for aneurysm who belonged to group A (50% vs. 6.3%; RR, 8; 95% CI, 2e31.9; P < 0.001) and cumulative incidence of first oral intake on those belonging to group B (96.9% vs. 61.8%; HR, 0.49; 95% CI, 0.28e0.84; P ¼ 0.009). However, no differences were found regarding PONV and resumption of oral intake between groups in patients treated for aortic occlusive disease (Table IV).
DISCUSSION This was a prospective cohort study in which systematic NGT use was compared with its selective use in patients undergoing open aortic repair. Higher PONV incidence and longer resumption of oral intake were observed in patients with a systematic use of NGT. No statistical significance was obtained regarding postoperative complications, but a trend toward fewer respiratory complications was also observed in patients with selective NGT use. No differences, however, were found in this sample regarding PONV and resumption of first oral intake between groups in patients treated for aortic occlusive disease. This result may be explained because of the higher population with abdominal aortic aneurysm in our sample compared with the presence of occlusive disease with no differences observed in baseline and intraoperative characteristics between patients with occlusive or aneurysmal aortic disease. In 2005, a meta-analysis was published to assess selective versus routine nasogastric decompression after elective laparotomy2 including 26 trials with
Vol. 29, No. 4, May 2015
Systematic NGT in aortic surgery
789
Table I. Baseline clinical parameters Variable
Age, years, mean ± SD Sex (M/F) Hypertension Dyslipidemia Diabetes BMI, kg/m2, mean ± SD Nutritional parameters, mean ± SD Cholesterol HDL cholesterol LDL cholesterol Triglycerides Total protein Prealbumin Albumin Tobacco use Alcohol abuse COPD Coronary artery disease Stroke PAD Laparotomy history Aortic pathology Aneurysm Aortic occlusive disease
Group A, systematic NGT (n ¼ 50), n (%)
Group B, nonsystematic NGT (n ¼ 50), n (%)
P
64.02 ± 9 48 (96)/2 (4) 38 (76) 37 (74) 15 (30) 28.49 ± 2.84
64.96 ± 7.58 49 (98)/1 (2) 32 (64) 34 (68) 5 (10) 28.81 ± 2.81
0.57 0.56 0.19 0.51 0.01a 0.57
164.28 ± 32.65 42.28 ± 9.76 106.24 ± 37.64 147.2 ± 57.75 6.33 ± 1.06 27.72 ± 6.30 4 ± 0.84 15 (30) 7 (14) 3 (6) 13 (26) 3 (6) 18 (32) 6 (12)
184.68 ± 35.38 39.78 ± 11.67 130.68 ± 42.06 147.68 ± 53.02 6.57 ± 0.89 26.82 ± 6.83 3.4 ± 0.64 21 (42) 1 (2) 6 (12) 9 (18) 4 (8) 19 (38) 10 (20)
0.08 0.17 0.09 0.91 0.27 0.39 0.67 0.44 0.06 0.29 0.33 0.69 0.83 0.27
34 (68) 18 (36)
32 (64) 19 (38)
0.67 0.83
SD, standard deviation; M, male; F, female; BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; COPD, chronic obstructive pulmonary disease; PAD, peripheral artery disease. a Significant value.
Table II. Intraoperative data Variable
Surgery duration, min, mean ± SD Cross-clamping type, n (%) Infrarenal Suprarenal Cross-clamping duration, min, mean ± SD Temperature at the end of the surgery, C, mean ± SD Intraoperative blood loss, mL, mean ± SD Intraoperative morphine (fentanyl vial 0.05 mg/mL), vial
almost 4,000 patients who met the inclusion criteria. Fewer pulmonary complications and fewer days to first oral intake were assessed for patients with no nasogastric decompression. Therefore, authors concluded that routine NGT use was not supported in the published literature. Following the same objective, the Cochrane Database published a review including 5,700 patients from clinical trials published between 1996 and 2009.1 Results were similar to the previous study, with less pulmonary postoperative complications and earlier return of
Group A, systematic NGT (n ¼ 50)
Group B, nonsystematic NGT (n ¼ 50)
158.28 ± 36.26
164.44 ± 36.13
43 (86) 7 (14) 29.10 ± 35.85 ± 887.22 ± 4.80 ±
41 (82) 9 (18) 29.90 ± 35.98 ± 781 ± 4.84 ±
9.16 0.41 622.42 0.93
12.60 0.26 397.69 0.79
P
0.36 0.59
0.59 0.16 0.39 0.73
bowel function in those patients with selective NGT. Regarding the use of systematic NGT in aortic surgery, there are only 2 studies published. First of all, Friedman et al.4 published their results derived from a prospective randomized study including 80 patients. No benefit from NGT maintenance after surgery was demonstrated compared to early removal. Another study including 40 patients5 also stated that routine NGT use was not beneficial even detrimental regarding postoperative lung complications and longer hospital stay.
790 Hern andez Mateo et al.
Annals of Vascular Surgery
Table III. Postoperative data Variable
PONV, n (%) Morphine consumption, mg, mean ± SD 24 hr 48 hr 72 hr Type of ICU analgesia, n (%) Intravenous Epidural catheter Combined Mortality, n (%) Postoperative complications, n (%) Respiratory (pneumonia-atelectasis) Ileus Unknown fever Wound infection Evisceration Arrhythmia Renal failure Time to first oral intake, n (%) 96 hr Reintervention, n (%) Length of hospital stay, days Median, range Mean ± SD
Group A, systematic NGT, (n ¼ 50)
Group B, nonsystematic NGT, (n ¼ 50)
P
24 (48)
10 (20)
0.003a
0.38 ± 1.19 0.48 ± 1.69 0.24 ± 1.69
0.12 ± 0.48 0.56 ± 2.13 0
0.16 0.84 0.32
2 40 8 1
(4) (80) (16) (2)
1 (2) 42 (84) 7 (14) d
0.8
5 1 6 1 2 1 1
(10) (2) (12) (2) (4) (2) (2)
1 2 1 d 1 2 2
(2) (4) (2)
0.20 1 0.11 0.31 1 0.10 0.10
7 22 17 2 1 3
(14) (44) (34) (4) (2) (6)
21 21 3 1 4 2
(42) (42) (6) (2) (8) (4)
0.008a
6 (1.25) 7.36 ± 3.38
0.19 0.02a
7 (4.5) 10.72 ± 9.5
0.31
(2) (4) (4)
0.65
SD, standard deviation; ICU, intensive care unit. a Significant value.
Table IV. Outcomes regarding treated aortic pathology
AAA (n ¼ 66) PONV Time to first oral intake (48 hr)
Group A, systematic NGT, n ¼ 34
Group B, nonsystematic NGT, n ¼ 32
Impact measures
P
17 (50%) 21 (61.8%)
2 (6.3%) 31 (96.9%)
RR 8 (CI 95% 2e31.9) HR 0.49 (CI 95% 0.28e0.84)
Background: Routine nasogastric tube (NGT) decompression has been traditionally used to prevent nausea and vomiting after abdominal surgery. Besides, many studies having demonstrated no benefits derived from this practice after an elective laparotomy, little evidence has been published regarding its use in aortic surgery. In this study, we analyze the effects of the selective use of the NGT in patients undergoing infrarenal aortic surgery in our center. Methods: Prospective cohort study including patients who underwent elective infrarenal aortic surgery between January 2011 and December 2012. Patients were prospectively included in group A (systematic NGT placement) and group B (nonsystematic NGT). The main end point was the occurrence of postoperative nausea and vomiting (PONV). Secondary end points were postoperative complications, time to first oral intake, and hospital stay. Results: One hundred patients were finally included in the study, 50 patients per group. Preoperative and intraoperative data were similar between both groups. Higher incidence of PONV (48% vs. 10%; relative risk, 2.4; 95% confidence interval [CI], 1.3e4.5; P ¼ 0.003) was observed in group A. Selective NGT behaved as a protective factor regarding earlier first oral intake in first postoperative 48 hours (hazard ratio, 0.67; 95% CI, 0.45e0.99; P ¼ 0.05). There were no differences in other adverse events although a trend toward fewer respiratory complications was observed in patients with nonsystematic NGT. Conclusions: This study demonstrates higher incidence of PONV and longer time to first oral intake in patients with systematic NGT with no benefits derived from this practice. Based on these results, selective NGT decompression should be encouraged in patients undergoing infrarenal aortic surgery.
INTRODUCTION Over the last 100 years, nasogastric tube (NGT) decompression after abdominal surgery has been used to prevent postoperative nausea and vomiting (PONV), abdominal distension, pulmonary complications, and wound dehiscence.1 However, despite this longstanding surgical tradition, many authors have published better results for the selective use
Department of Angiology and Vascular Surgery, Hospital Clı´nico San Carlos, Madrid, Spain. Correspondence to: Manuela M. Hernandez Mateo, MD, Department of Angiology and Vascular Surgery, Hospital Clı´nico San Carlos, Martı´n Lagos, St. Madrid, Spain; E-mail: [email protected] Ann Vasc Surg 2015; 29: 786–791 http://dx.doi.org/10.1016/j.avsg.2015.01.002 Ó 2015 Elsevier Inc. All rights reserved. Manuscript received: April 2, 2014; manuscript accepted: January 1, 2015; published online: March 4, 2015.
786
of the NGT in terms of postoperative complications after an elective laparotomy compared with its systematic placement.1e5 Many of these studies have been published according to results obtained on colorectal, gynecologic, and urologic interventions.1e3 Nevertheless, little evidence has been published regarding the use of the NGT in aortic surgery. Only 2 studies have evaluated NGT use in this field.4,5 Both of them state that routine NGT use is not beneficial even detrimental regarding postoperative lung complications and longer hospital stay. Thus, based on these results, the Society of Vascular Surgery questioned this practice on its latest practice guidelines.6 Based on the reported literature, the purpose of this study was to analyze the results derived from the selective use of NGT in open aortic surgery in our center.
Vol. 29, No. 4, May 2015
PATIENTS AND METHODS Study Design This is a prospective cohort study. One hundred and fifty-three consecutive patients referred to our institution underwent open surgery to treat infrarenal aortic aneurysm or aortoiliac occlusive disease between January 2011 and December 2012. Excluded from this study were patients with emergent or urgent surgery (ruptured or symptomatic aortic aneurysms), previous esophagogastric or pancreatoduodenal surgery, gastroesophageal reflux or hiatal hernia, retroperitoneal approach, accidental intraoperative bowel injury, adhesiolisis for more than 1 hr, postoperative mechanical ventilation for more than 12 hr, or postoperative consciousness disorders (Glasgow score 38 C) were observed in patients with routine NGT although these did not reach statistical significance (RR, 5; 95% CI, 0.6e41.3; P ¼ 0.2 and RR, 6; 95% CI, 0.7e48.05; P ¼ 0.1, respectively).
Annals of Vascular Surgery
Three patients with no systematic NGT needed the implantation of an NGT during the postoperative course because of vomiting in the first 24 hr in 1 patient and because of the presence of postoperative ileus in the remaining 2 patients. Reintroduction of the NGT was also needed in 1 patient with previous routine NGT because of postoperative ileus. Only 1 patient died, in group A, due to multiorgan failure after colonic ischemia, events not related to the use of NGT. Also, compared with patients with no NGT, those with routine insertion of an NGT had a longer mean hospital stay (10.72 days vs. 7.36 days, P ¼ 0.02) despite no differences were found in median hospital stay, adjusted due to longer stay of the previous mentioned patient who died due to multiorgan failure. In a subgroup analysis based on the treated pathology, differences obtained regarding PONV and first oral intake were only present in patients treated for aneurysm. The incidence of PONV was also greater in patients treated for aneurysm who belonged to group A (50% vs. 6.3%; RR, 8; 95% CI, 2e31.9; P < 0.001) and cumulative incidence of first oral intake on those belonging to group B (96.9% vs. 61.8%; HR, 0.49; 95% CI, 0.28e0.84; P ¼ 0.009). However, no differences were found regarding PONV and resumption of oral intake between groups in patients treated for aortic occlusive disease (Table IV).
DISCUSSION This was a prospective cohort study in which systematic NGT use was compared with its selective use in patients undergoing open aortic repair. Higher PONV incidence and longer resumption of oral intake were observed in patients with a systematic use of NGT. No statistical significance was obtained regarding postoperative complications, but a trend toward fewer respiratory complications was also observed in patients with selective NGT use. No differences, however, were found in this sample regarding PONV and resumption of first oral intake between groups in patients treated for aortic occlusive disease. This result may be explained because of the higher population with abdominal aortic aneurysm in our sample compared with the presence of occlusive disease with no differences observed in baseline and intraoperative characteristics between patients with occlusive or aneurysmal aortic disease. In 2005, a meta-analysis was published to assess selective versus routine nasogastric decompression after elective laparotomy2 including 26 trials with
Vol. 29, No. 4, May 2015
Systematic NGT in aortic surgery
789
Table I. Baseline clinical parameters Variable
Age, years, mean ± SD Sex (M/F) Hypertension Dyslipidemia Diabetes BMI, kg/m2, mean ± SD Nutritional parameters, mean ± SD Cholesterol HDL cholesterol LDL cholesterol Triglycerides Total protein Prealbumin Albumin Tobacco use Alcohol abuse COPD Coronary artery disease Stroke PAD Laparotomy history Aortic pathology Aneurysm Aortic occlusive disease
Group A, systematic NGT (n ¼ 50), n (%)
Group B, nonsystematic NGT (n ¼ 50), n (%)
P
64.02 ± 9 48 (96)/2 (4) 38 (76) 37 (74) 15 (30) 28.49 ± 2.84
64.96 ± 7.58 49 (98)/1 (2) 32 (64) 34 (68) 5 (10) 28.81 ± 2.81
0.57 0.56 0.19 0.51 0.01a 0.57
164.28 ± 32.65 42.28 ± 9.76 106.24 ± 37.64 147.2 ± 57.75 6.33 ± 1.06 27.72 ± 6.30 4 ± 0.84 15 (30) 7 (14) 3 (6) 13 (26) 3 (6) 18 (32) 6 (12)
184.68 ± 35.38 39.78 ± 11.67 130.68 ± 42.06 147.68 ± 53.02 6.57 ± 0.89 26.82 ± 6.83 3.4 ± 0.64 21 (42) 1 (2) 6 (12) 9 (18) 4 (8) 19 (38) 10 (20)
0.08 0.17 0.09 0.91 0.27 0.39 0.67 0.44 0.06 0.29 0.33 0.69 0.83 0.27
34 (68) 18 (36)
32 (64) 19 (38)
0.67 0.83
SD, standard deviation; M, male; F, female; BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; COPD, chronic obstructive pulmonary disease; PAD, peripheral artery disease. a Significant value.
Table II. Intraoperative data Variable
Surgery duration, min, mean ± SD Cross-clamping type, n (%) Infrarenal Suprarenal Cross-clamping duration, min, mean ± SD Temperature at the end of the surgery, C, mean ± SD Intraoperative blood loss, mL, mean ± SD Intraoperative morphine (fentanyl vial 0.05 mg/mL), vial
almost 4,000 patients who met the inclusion criteria. Fewer pulmonary complications and fewer days to first oral intake were assessed for patients with no nasogastric decompression. Therefore, authors concluded that routine NGT use was not supported in the published literature. Following the same objective, the Cochrane Database published a review including 5,700 patients from clinical trials published between 1996 and 2009.1 Results were similar to the previous study, with less pulmonary postoperative complications and earlier return of
Group A, systematic NGT (n ¼ 50)
Group B, nonsystematic NGT (n ¼ 50)
158.28 ± 36.26
164.44 ± 36.13
43 (86) 7 (14) 29.10 ± 35.85 ± 887.22 ± 4.80 ±
41 (82) 9 (18) 29.90 ± 35.98 ± 781 ± 4.84 ±
9.16 0.41 622.42 0.93
12.60 0.26 397.69 0.79
P
0.36 0.59
0.59 0.16 0.39 0.73
bowel function in those patients with selective NGT. Regarding the use of systematic NGT in aortic surgery, there are only 2 studies published. First of all, Friedman et al.4 published their results derived from a prospective randomized study including 80 patients. No benefit from NGT maintenance after surgery was demonstrated compared to early removal. Another study including 40 patients5 also stated that routine NGT use was not beneficial even detrimental regarding postoperative lung complications and longer hospital stay.
790 Hern andez Mateo et al.
Annals of Vascular Surgery
Table III. Postoperative data Variable
PONV, n (%) Morphine consumption, mg, mean ± SD 24 hr 48 hr 72 hr Type of ICU analgesia, n (%) Intravenous Epidural catheter Combined Mortality, n (%) Postoperative complications, n (%) Respiratory (pneumonia-atelectasis) Ileus Unknown fever Wound infection Evisceration Arrhythmia Renal failure Time to first oral intake, n (%) 96 hr Reintervention, n (%) Length of hospital stay, days Median, range Mean ± SD
Group A, systematic NGT, (n ¼ 50)
Group B, nonsystematic NGT, (n ¼ 50)
P
24 (48)
10 (20)
0.003a
0.38 ± 1.19 0.48 ± 1.69 0.24 ± 1.69
0.12 ± 0.48 0.56 ± 2.13 0
0.16 0.84 0.32
2 40 8 1
(4) (80) (16) (2)
1 (2) 42 (84) 7 (14) d
0.8
5 1 6 1 2 1 1
(10) (2) (12) (2) (4) (2) (2)
1 2 1 d 1 2 2
(2) (4) (2)
0.20 1 0.11 0.31 1 0.10 0.10
7 22 17 2 1 3
(14) (44) (34) (4) (2) (6)
21 21 3 1 4 2
(42) (42) (6) (2) (8) (4)
0.008a
6 (1.25) 7.36 ± 3.38
0.19 0.02a
7 (4.5) 10.72 ± 9.5
0.31
(2) (4) (4)
0.65
SD, standard deviation; ICU, intensive care unit. a Significant value.
Table IV. Outcomes regarding treated aortic pathology
AAA (n ¼ 66) PONV Time to first oral intake (48 hr)
Group A, systematic NGT, n ¼ 34
Group B, nonsystematic NGT, n ¼ 32
Impact measures
P
17 (50%) 21 (61.8%)
2 (6.3%) 31 (96.9%)
RR 8 (CI 95% 2e31.9) HR 0.49 (CI 95% 0.28e0.84)
