- Email: [email protected]
Investigating the use of preoperative nasogastric tubes and postoperative outcomes for infants with pyloric stenosis: a retrospective cohort study
Journal of Pediatric Surgery (2010) 45, 1020–1023
www.elsevier.com/locate/jpedsurg
Investigating the use of preoperative nasogastric tubes and postoperative outcomes for infants with pyloric stenosis: a retrospective cohort study Ahmad Elanahas, Julia Pemberton, Yasmin Yousef, Helene Flageole ⁎ McMaster University-Hamilton, Ontario, Canada L8N 3Z5 Received 1 February 2010; accepted 3 February 2010
Key words: Pyloric stenosis; Nasogastric tube; Emesis; Length of stay
Abstract Purpose: Literature is lacking regarding the role of nasogastric tubes in patients with pyloric stenosis. There is also no consensus among surgeons. Some believe that pyloric stenosis is a form of gastric outlet obstruction, and the stomach should be drained until the obstruction is relieved. Others claim that infants can handle their secretions, and draining the stomach may further exacerbate the alkalosis. This chart review examines the use of preoperative nasogastric tubes in a single pediatric institution and its effect on vomiting rates and length of stay. Methods: After research ethics board approval, a retrospective review was performed on 109 patients admitted between January 1, 2007, and December 31, 2008, with pyloric stenosis who underwent pyloromyotomy. Data were collected on presence of a preoperative nasogastric tube, preoperative electrolyte levels, ultrasound characteristics, episodes of postoperative vomiting, and length of stay. Results: One hundred six patients were used in the final analysis. A nasogastric tube was placed in 77 patients (73%). Patients with a preoperative nasogastric tube had significantly higher episodes of postoperative vomiting (P = .015; 95% confidence interval [CI] 0.29-2.63) and length of stay (P = .017; 95% CI, 2.49-25.01). Bicarbonate levels were also significantly higher in patients with a nasogastric tube. There was no difference in the duration of symptoms, ultrasound characteristics, or type of operation between the 2 cohorts. Conclusion: The data strongly suggest that preoperative nasogastric tube placement adversely affects postoperative vomiting and consequently increases length of stay. The lack of consensus about the use of preoperative nasogastric tubes coupled with these findings indicates the need to evaluate this practice with a prospective randomized controlled trial. © 2010 Elsevier Inc. All rights reserved.
One of the common postoperative complications in infantile hypertrophic pyloric stenosis (IHPS) is prolonged Presented at the 41st Annual Meeting of the Canadian Association of Paediatric Surgeons, Halifax, Nova Scotia, Canada, October 1-3, 2009. ⁎ Corresponding author. McMaster Children's Hospital, Hamilton, Ontario, Canada L8N 3Z5. E-mail address: [email protected] (H. Flageole). 0022-3468/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2010.02.026
vomiting, with rates occurring between 25% and 90% of admissions [1,2]. This problem will inevitably delay feeding and increase the length of hospital stay. Many studies have looked into the factors that potentially influence or predict the occurrence of postoperative emesis after pyloromyotomy. However, factors such as surgical technique or feeding regimen have not been shown to affect the time to full feeds or the length of stay [3-17].
Preoperative NGT and postoperative outcomes for pyloric stenosis One potential determinant of postoperative emesis and feeding tolerance that has not been sufficiently addressed in the literature is the use of preoperative nasogastric tubes (NGTs). Some pediatric surgeons believe that IHPS is a form of gastric outlet obstruction, and the stomach should therefore be decompressed as much as possible before surgery. Others contend that patients with IHPS can usually handle their gastric secretions, and the presence of an NGT will only exacerbate the patients underlying metabolic abnormality [18]. Their position is based on the principle that preoperative resuscitation and correction of metabolic abnormalities is known to ensure good outcomes in IHPS [19]. The purpose of this study was to investigate the use and effect of NGT preoperatively on rates of postoperative emesis and ultimately length of stay. This study will also serve as a primer toward a prospective randomized trial, which will evaluate the effect of preoperative NGT on rates of postoperative emesis and length of postoperative stay in children with pyloric stenosis.
1. Methods After obtaining approval from the research ethics board, a review of all medical records was conducted on patients admitted to the McMaster Children's Hospital (Hamilton, Ontario, Canada) with a primary diagnosis of IHPS and who received a pyloromyotomy between January 1, 2007, and December 31, 2008. Data were collected on patient variables, such as age, sex, duration of symptoms, electrolytes on admission, use of NGT, operating time, time to first feed, time to full feeds, episodes of postoperative emesis, and total hospital stay. A total of 109 patients were identified during the study period. Two patients were excluded because they were preterm infants who developed IHPS as part of their complicated neonatal course, and one was excluded because it was unclear from the chart whether the patient had an NGT. The remaining 106 patients were divided into 2 cohorts based on the presence or absence of a preoperative NGT. For statistical analyses, descriptive statistics were used where appropriate and a 2-tailed, Student's t test for unequal sample sizes was used to determine significant differences between the 2 cohorts. All analyses were performed using SPSS V17.0 (SPSS Inc, Chicago, Ill), and a P value less than .05 was used to establish statistical significance.
Table 1
Descriptive statistics
Age (d) Sex (male/female) Duration of symptoms (d) Total length of stay (h) ⁎ Postoperative length of stay (h) Admission bicarbonate level (mmol/L) ⁎ U/S length U/S thickness Operating time (min) Postoperative emesis (n) ⁎ Time to first feeds (h) Time to full feeds (h) ⁎
NGT+ (n = 77), mean (SD)
NGT− (n = 29), mean (SD)
48.56 (31.47) 63:14 10.16 (10.45)
35.52 (15.10) 22:7 9.14 (7.04)
75.74 (28.07) 48.56 (24.98)
61.99 (19.59) 39.82 (17.37)
27.29 (5.71)
26.32 (4.50)
20.4 (3.49) 4.8 (1.10) 74.7 (17.09) 3.32 (2.93)
20.2 (2.84) 4.8 (1.30) 73.1 (13.87) 1.86 (2.01)
8.90 (4.83) 36.35 (21.93)
7.97 (4.00) 28.93 (16.21)
⁎ P b .05.
omy. Only 2 patients were readmitted after discharge because of vomiting and both improved spontaneously. The analysis revealed that 77 patients (82%) had an NGT inserted preoperatively (NGT+), either on admission to McMaster Children's Hospital or at the referring center. Both cohorts were well matched in demographics, weight on admission, duration of symptoms, and ultrasound dimensions. Operating times along with time to first and full feeds were also comparable between the 2 groups (Table 1). There was a statistically significant difference in the number of postoperative emesis episodes. Patients in the NGT+ group vomited a mean of 3.32 (SD, 2.93) times postoperatively vs a mean of 1.86 (SD, 2.01) in the cohort who did not receive a preoperative nasogastric tube (NGT−) group (P = .015; 95% confidence interval [CI], 0.29-2.63). The total length of stay was significantly shorter (61.99 vs 75.74 hours) in the NGT− group compared to the NGT+ group (P b .02; 95% CI, 2.4925.00). The postoperative length of stay was also shorter in the NGT− group although not reaching statistical significance (P = .087; 95% CI, 1.27-18.25). In addition, the admission bicarbonate level in the NGT+ group was significantly higher than the NGT− group (P = .004; 95% CI, −1.77 to −0.33). Table 2
2. Results A total of 88 males (80.73%) and 21 females (19.27%) were included in the study analysis. All patients had a preoperative ultrasound (even when the pyloric mass was palpable) and underwent an open periumbilical pyloromyot-
1021
NGT+ NGT− Total
The effect of NGT on postoperative emesis N2 episodes of postoperative emesis
b2 episodes of postoperative emesis
Total
36 7 43
41 22 63
77 29
1022
A. Elanahas et al.
From a clinical perspective, a significant reduction in postoperative emesis was considered to be an event rate of 2 or less. Therefore, according to the data analysis, NGT+ patients are 1.94 times more likely to experience greater than 2 episodes of postoperative emesis than NGT− patients (Table 2).
3. Discussion In the last 2 decades, studies with respect to IHPS have been concentrating mainly on comparing surgical approaches (eg, laparoscopic vs open) [4-8] or investigating nonoperative treatment strategies (eg, feeding regimens and atropine) [1,9-16]. Studies have been unable to determine a causal relationship between operative approach and outcomes such as episodes of postoperative emesis, time to full feeds, and length of postoperative stay [4-8]. Nonetheless, it is encouraging to see high-quality evidence, such as randomized controlled trials, used to address this area of interest. Several other studies have also failed to prove any predictive value to early vs late feeding regimens [1,9-16] and the use of atropine [17] with respect to postoperative emesis. However, little attention has been placed in evaluating the use and potential benefit (or harm) of NGTs as an adjunct to the usual preoperative management for IHPS. Golladay et al (1987) [20] published the only study, which reviewed 90 infant charts and concluded that preoperative NGT drainage resulted in a better acceptance of graduated feeding, fewer episodes of emesis, earlier completion of full feeding, and shortened hospital stay. Discussions among clinicians reveal that there is still no consensus on the use of preoperative NGT. Advocates for NGT insertion rationalize its use as a means to reduce the risk of aspiration during the induction of anesthesia and relentless vomiting. They argue that a stomach chronically distended with milk curds and excess mucous would slow the recovery of gastric motility causing a delay in time to full feeds and subsequently a prolonged hospital stay postpyloromyotomy. On the other hand, proponents for not using NGT state that patients with IHPS do not have a complete obstruction and can usually handle their gastric secretions. Furthermore, continuous suctioning of gastric contents without intravenous replacement is well-known to cause electrolyte disturbances, and this can theoretically compound the characteristic hypokalemic, hypochloremic metabolic alkalosis seen in IHPS [21-23]. A delay in achieving normal acidbase balance and electrolyte levels will ultimately delay the time a patient would be ready for surgery. The lack of understanding with respect to the factors that influence postoperative emesis and the possible association with NGTs was the main motive behind this study. Our results reveal that the NGT+ patient population had significantly more episodes of postoperative emesis (Fig. 1). They also had a significantly longer total length of
Fig. 1 Episodes of postoperative emesis depending on presence of NGT.
stay and took longer to reach full feeds although this latter difference was not statistically significant (Fig. 2). One of the limitations in the interpretation of our data was realizing that in our institution, NGT insertion was based on the health care provider's preference. Consequently, the reason(s) patients' received a preoperative NGT are unclear. The patient may have had more protracted vomiting or could have been perceived to be sicker. The analysis revealed a higher admission bicarbonate level in the NGT+ group, which suggests that the decision to insert an NGT may have been prompted by the perception that these patients were more ill. This inference is based on the theory that patients with severe alkalosis are at a more advanced stage of their disease and therefore require more aggressive therapy including an NGT. In addition, there was no consistency in the use of low intermittent suction or straight drainage of NGT, and therefore, major discrepancies in gastric return existed within the NGT+ group. Another source of ambiguity in the analysis would be defining the term emesis. Recording a minor «spit up» rather than an actual episode of vomiting depends on the experience of both the caregiver and the nurse. Furthermore, although no
Fig. 2 Comparing time to key outcomes based on presence of NGT. LOS indicates length of stay.
Preoperative NGT and postoperative outcomes for pyloric stenosis significant difference was found between the 2 cohorts in the time to first feeds, there was no existing standardized postoperative feeding protocol in our institution. First feeding was introduced after 6 hours or later (12 or 18 hours) and full feeds were attained when patients tolerated smaller, more frequent feeds or went back to their premorbid feeding quantity. In this study analysis, full feeds was defined as no vomiting after 2 consecutive feeds, regardless of amount. However, this created some inconsistencies as some patients would start vomiting even after more than 2 continuous tolerated feeds. Therefore, using time to full feeds is not a reliable measurement of outcome and a more appropriate substitute would be postoperative length of stay. The results of this study challenges the only previous literature addressing this issue, a chart review of 90 patients published in 1987 that found a reduction in postoperative emesis events in patients where a preoperative NGT was placed [20]. With the paucity of literature focused on NGT use in the IHPS population and the low level of available evidence, it becomes clear that additional research of higher methodological quality is needed to resolve this quandary and standardize clinical practice.
References [1] Leahy A, Fitzgerald RJ. The influence of delayed feeding on postoperative vomiting in hypertrophic pyloric stenosis. Br J Surg 1982; 69:658-9. [2] Spitz L. Vomiting after pyloromyotomy for infantile hypertrophic pyloric stenosis. Arch Dis Child 1979;54:886-9. [3] Sommerfield T, Chalmers J, Youngson G, et al. The changing epidemiology of infantile hypertrophic pyloric stenosis in Scotland. Arch Dis Child 2008:1-12. [4] St Peter SD, Tsao K, Sharp SW, et al. Predictors of emesis and time to goal intake after pyloromyotomy: analysis from a prospective trial. J Pediatr Surg 2008;43:2038-41. [5] St Peter SD, Holcomb CM, Calkins GW, et al. Open versus laparoscopic pyloromyotomy for pyloric stenosis. A prospective, randomized trial. Ann Surg 2006;244:363-70.
1023
[6] Leclair MD, Plattner V, Mirallie E, et al. Laparoscopic pyloromyotomy for hypertrophic pyloric stenosis: a prospective, randomized controlled trial. J Pediatr Surg 2007;42:692-8. [7] Hall NJ, Van Der Zee J, Tan HL, et al. Meta-analysis of laparoscopic versus open pyloromyotomy. Ann Surg 2004;240:774-8. [8] Kim SS, Lau ST, Lee SL, et al. Pyloromyotomy: a comparison of laparoscopic circumumbilical and right upper quadrant operative techniques. J Am Coll Surg 2005;201:66-70. [9] Van der Bilt JDW, Kramer WL, van der Zee DC, et al. Laparoscopic pyloromyotomy for hypertrophic pyloric stenosis. Surg Endosc 2004; 18:907-9. [10] Carpenter RO, Schaffer RL, Maeso CE, et al. Postoperative ad lib feeding for hypertrophic pyloric stenosis. J Pediatr Surg 1999;34: 959-61. [11] Foster ME, Lewis WG. Early postoperative feeding-a continuing controversy in pyloric stenosis. J Royal Soc Med 1989;82:532-3. [12] Garza JJ, Morash D, Dzakovic A, et al. Ad libitum feeding decreases hospital stay for neonates after pyloromyotomy. J Pediatr Surg 2002; 37:493-5. [13] Georgeson KE, Corbin TJ, Griffen JW, et al. An analysis of feeding regimens after pyloromyotomy for hypertrophic pyloric stenosis. J Pediatr Surg 1993;28:1478-80. [14] Gollin G, Doslouglu H, Flummerfeldt P, et al. Rapid advancement of feedings after pyloromyotomy for pyloric stenosis. Clin Pediatr (Phila) 2000;39:187-90. [15] Lee AC, Munro FD, MacKinlay GA. An audit of post-pyloromyotomy feeding regimens. Eur J Pediatr Surg 2001;11:12-4. [16] Adibe OO, Nichol PF, Lim FY, et al. Ad libitum feeds after laparoscopic pyloromyotomy: a retrospective comparison with a standardized feeding regimen in 227 infants. J Laparoendosc Surg Tech A 2007;17:235-7. [17] Singh UK, Kumar R, Prasad R. Oral atropine sulfate for infantile hypertrophic pyloric stenosis. Indian Pediatr 2005;42:473-6. [18] O'Neill JA, Coran AG, Fonkalsrud, et al. Pediatric surgery. St Louis, MO: Mosby Year-Book; 2006. [19] Allan C. Determinants of good outcome in pyloric stenosis. J Paediatr Child Health 2006;42:86-8. [20] Golladay ES, Broadwater JR, Mollitt DL. Pyloric stenosis—a timed perspective. Arch Surg 1987;122:825-6. [21] Wilikes-Holmes C. Safe placement of nasogastric tubes in children. Paediatr Nursing 2006;18:9. [22] Richardson DS, Branowicki PA, Zeidman-Rogers L, et al. An evidence-based approach to nasogastric tube management: special considerations. J Pediatr Nursing 2006;21:388-93. [23] Ellett MC. What is known about methods of correctly placing gastric tubes in adults and children. Gastroenterol Nursing 2004;27:253-9.
www.elsevier.com/locate/jpedsurg
Investigating the use of preoperative nasogastric tubes and postoperative outcomes for infants with pyloric stenosis: a retrospective cohort study Ahmad Elanahas, Julia Pemberton, Yasmin Yousef, Helene Flageole ⁎ McMaster University-Hamilton, Ontario, Canada L8N 3Z5 Received 1 February 2010; accepted 3 February 2010
Key words: Pyloric stenosis; Nasogastric tube; Emesis; Length of stay
Abstract Purpose: Literature is lacking regarding the role of nasogastric tubes in patients with pyloric stenosis. There is also no consensus among surgeons. Some believe that pyloric stenosis is a form of gastric outlet obstruction, and the stomach should be drained until the obstruction is relieved. Others claim that infants can handle their secretions, and draining the stomach may further exacerbate the alkalosis. This chart review examines the use of preoperative nasogastric tubes in a single pediatric institution and its effect on vomiting rates and length of stay. Methods: After research ethics board approval, a retrospective review was performed on 109 patients admitted between January 1, 2007, and December 31, 2008, with pyloric stenosis who underwent pyloromyotomy. Data were collected on presence of a preoperative nasogastric tube, preoperative electrolyte levels, ultrasound characteristics, episodes of postoperative vomiting, and length of stay. Results: One hundred six patients were used in the final analysis. A nasogastric tube was placed in 77 patients (73%). Patients with a preoperative nasogastric tube had significantly higher episodes of postoperative vomiting (P = .015; 95% confidence interval [CI] 0.29-2.63) and length of stay (P = .017; 95% CI, 2.49-25.01). Bicarbonate levels were also significantly higher in patients with a nasogastric tube. There was no difference in the duration of symptoms, ultrasound characteristics, or type of operation between the 2 cohorts. Conclusion: The data strongly suggest that preoperative nasogastric tube placement adversely affects postoperative vomiting and consequently increases length of stay. The lack of consensus about the use of preoperative nasogastric tubes coupled with these findings indicates the need to evaluate this practice with a prospective randomized controlled trial. © 2010 Elsevier Inc. All rights reserved.
One of the common postoperative complications in infantile hypertrophic pyloric stenosis (IHPS) is prolonged Presented at the 41st Annual Meeting of the Canadian Association of Paediatric Surgeons, Halifax, Nova Scotia, Canada, October 1-3, 2009. ⁎ Corresponding author. McMaster Children's Hospital, Hamilton, Ontario, Canada L8N 3Z5. E-mail address: [email protected] (H. Flageole). 0022-3468/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2010.02.026
vomiting, with rates occurring between 25% and 90% of admissions [1,2]. This problem will inevitably delay feeding and increase the length of hospital stay. Many studies have looked into the factors that potentially influence or predict the occurrence of postoperative emesis after pyloromyotomy. However, factors such as surgical technique or feeding regimen have not been shown to affect the time to full feeds or the length of stay [3-17].
Preoperative NGT and postoperative outcomes for pyloric stenosis One potential determinant of postoperative emesis and feeding tolerance that has not been sufficiently addressed in the literature is the use of preoperative nasogastric tubes (NGTs). Some pediatric surgeons believe that IHPS is a form of gastric outlet obstruction, and the stomach should therefore be decompressed as much as possible before surgery. Others contend that patients with IHPS can usually handle their gastric secretions, and the presence of an NGT will only exacerbate the patients underlying metabolic abnormality [18]. Their position is based on the principle that preoperative resuscitation and correction of metabolic abnormalities is known to ensure good outcomes in IHPS [19]. The purpose of this study was to investigate the use and effect of NGT preoperatively on rates of postoperative emesis and ultimately length of stay. This study will also serve as a primer toward a prospective randomized trial, which will evaluate the effect of preoperative NGT on rates of postoperative emesis and length of postoperative stay in children with pyloric stenosis.
1. Methods After obtaining approval from the research ethics board, a review of all medical records was conducted on patients admitted to the McMaster Children's Hospital (Hamilton, Ontario, Canada) with a primary diagnosis of IHPS and who received a pyloromyotomy between January 1, 2007, and December 31, 2008. Data were collected on patient variables, such as age, sex, duration of symptoms, electrolytes on admission, use of NGT, operating time, time to first feed, time to full feeds, episodes of postoperative emesis, and total hospital stay. A total of 109 patients were identified during the study period. Two patients were excluded because they were preterm infants who developed IHPS as part of their complicated neonatal course, and one was excluded because it was unclear from the chart whether the patient had an NGT. The remaining 106 patients were divided into 2 cohorts based on the presence or absence of a preoperative NGT. For statistical analyses, descriptive statistics were used where appropriate and a 2-tailed, Student's t test for unequal sample sizes was used to determine significant differences between the 2 cohorts. All analyses were performed using SPSS V17.0 (SPSS Inc, Chicago, Ill), and a P value less than .05 was used to establish statistical significance.
Table 1
Descriptive statistics
Age (d) Sex (male/female) Duration of symptoms (d) Total length of stay (h) ⁎ Postoperative length of stay (h) Admission bicarbonate level (mmol/L) ⁎ U/S length U/S thickness Operating time (min) Postoperative emesis (n) ⁎ Time to first feeds (h) Time to full feeds (h) ⁎
NGT+ (n = 77), mean (SD)
NGT− (n = 29), mean (SD)
48.56 (31.47) 63:14 10.16 (10.45)
35.52 (15.10) 22:7 9.14 (7.04)
75.74 (28.07) 48.56 (24.98)
61.99 (19.59) 39.82 (17.37)
27.29 (5.71)
26.32 (4.50)
20.4 (3.49) 4.8 (1.10) 74.7 (17.09) 3.32 (2.93)
20.2 (2.84) 4.8 (1.30) 73.1 (13.87) 1.86 (2.01)
8.90 (4.83) 36.35 (21.93)
7.97 (4.00) 28.93 (16.21)
⁎ P b .05.
omy. Only 2 patients were readmitted after discharge because of vomiting and both improved spontaneously. The analysis revealed that 77 patients (82%) had an NGT inserted preoperatively (NGT+), either on admission to McMaster Children's Hospital or at the referring center. Both cohorts were well matched in demographics, weight on admission, duration of symptoms, and ultrasound dimensions. Operating times along with time to first and full feeds were also comparable between the 2 groups (Table 1). There was a statistically significant difference in the number of postoperative emesis episodes. Patients in the NGT+ group vomited a mean of 3.32 (SD, 2.93) times postoperatively vs a mean of 1.86 (SD, 2.01) in the cohort who did not receive a preoperative nasogastric tube (NGT−) group (P = .015; 95% confidence interval [CI], 0.29-2.63). The total length of stay was significantly shorter (61.99 vs 75.74 hours) in the NGT− group compared to the NGT+ group (P b .02; 95% CI, 2.4925.00). The postoperative length of stay was also shorter in the NGT− group although not reaching statistical significance (P = .087; 95% CI, 1.27-18.25). In addition, the admission bicarbonate level in the NGT+ group was significantly higher than the NGT− group (P = .004; 95% CI, −1.77 to −0.33). Table 2
2. Results A total of 88 males (80.73%) and 21 females (19.27%) were included in the study analysis. All patients had a preoperative ultrasound (even when the pyloric mass was palpable) and underwent an open periumbilical pyloromyot-
1021
NGT+ NGT− Total
The effect of NGT on postoperative emesis N2 episodes of postoperative emesis
b2 episodes of postoperative emesis
Total
36 7 43
41 22 63
77 29
1022
A. Elanahas et al.
From a clinical perspective, a significant reduction in postoperative emesis was considered to be an event rate of 2 or less. Therefore, according to the data analysis, NGT+ patients are 1.94 times more likely to experience greater than 2 episodes of postoperative emesis than NGT− patients (Table 2).
3. Discussion In the last 2 decades, studies with respect to IHPS have been concentrating mainly on comparing surgical approaches (eg, laparoscopic vs open) [4-8] or investigating nonoperative treatment strategies (eg, feeding regimens and atropine) [1,9-16]. Studies have been unable to determine a causal relationship between operative approach and outcomes such as episodes of postoperative emesis, time to full feeds, and length of postoperative stay [4-8]. Nonetheless, it is encouraging to see high-quality evidence, such as randomized controlled trials, used to address this area of interest. Several other studies have also failed to prove any predictive value to early vs late feeding regimens [1,9-16] and the use of atropine [17] with respect to postoperative emesis. However, little attention has been placed in evaluating the use and potential benefit (or harm) of NGTs as an adjunct to the usual preoperative management for IHPS. Golladay et al (1987) [20] published the only study, which reviewed 90 infant charts and concluded that preoperative NGT drainage resulted in a better acceptance of graduated feeding, fewer episodes of emesis, earlier completion of full feeding, and shortened hospital stay. Discussions among clinicians reveal that there is still no consensus on the use of preoperative NGT. Advocates for NGT insertion rationalize its use as a means to reduce the risk of aspiration during the induction of anesthesia and relentless vomiting. They argue that a stomach chronically distended with milk curds and excess mucous would slow the recovery of gastric motility causing a delay in time to full feeds and subsequently a prolonged hospital stay postpyloromyotomy. On the other hand, proponents for not using NGT state that patients with IHPS do not have a complete obstruction and can usually handle their gastric secretions. Furthermore, continuous suctioning of gastric contents without intravenous replacement is well-known to cause electrolyte disturbances, and this can theoretically compound the characteristic hypokalemic, hypochloremic metabolic alkalosis seen in IHPS [21-23]. A delay in achieving normal acidbase balance and electrolyte levels will ultimately delay the time a patient would be ready for surgery. The lack of understanding with respect to the factors that influence postoperative emesis and the possible association with NGTs was the main motive behind this study. Our results reveal that the NGT+ patient population had significantly more episodes of postoperative emesis (Fig. 1). They also had a significantly longer total length of
Fig. 1 Episodes of postoperative emesis depending on presence of NGT.
stay and took longer to reach full feeds although this latter difference was not statistically significant (Fig. 2). One of the limitations in the interpretation of our data was realizing that in our institution, NGT insertion was based on the health care provider's preference. Consequently, the reason(s) patients' received a preoperative NGT are unclear. The patient may have had more protracted vomiting or could have been perceived to be sicker. The analysis revealed a higher admission bicarbonate level in the NGT+ group, which suggests that the decision to insert an NGT may have been prompted by the perception that these patients were more ill. This inference is based on the theory that patients with severe alkalosis are at a more advanced stage of their disease and therefore require more aggressive therapy including an NGT. In addition, there was no consistency in the use of low intermittent suction or straight drainage of NGT, and therefore, major discrepancies in gastric return existed within the NGT+ group. Another source of ambiguity in the analysis would be defining the term emesis. Recording a minor «spit up» rather than an actual episode of vomiting depends on the experience of both the caregiver and the nurse. Furthermore, although no
Fig. 2 Comparing time to key outcomes based on presence of NGT. LOS indicates length of stay.
Preoperative NGT and postoperative outcomes for pyloric stenosis significant difference was found between the 2 cohorts in the time to first feeds, there was no existing standardized postoperative feeding protocol in our institution. First feeding was introduced after 6 hours or later (12 or 18 hours) and full feeds were attained when patients tolerated smaller, more frequent feeds or went back to their premorbid feeding quantity. In this study analysis, full feeds was defined as no vomiting after 2 consecutive feeds, regardless of amount. However, this created some inconsistencies as some patients would start vomiting even after more than 2 continuous tolerated feeds. Therefore, using time to full feeds is not a reliable measurement of outcome and a more appropriate substitute would be postoperative length of stay. The results of this study challenges the only previous literature addressing this issue, a chart review of 90 patients published in 1987 that found a reduction in postoperative emesis events in patients where a preoperative NGT was placed [20]. With the paucity of literature focused on NGT use in the IHPS population and the low level of available evidence, it becomes clear that additional research of higher methodological quality is needed to resolve this quandary and standardize clinical practice.
References [1] Leahy A, Fitzgerald RJ. The influence of delayed feeding on postoperative vomiting in hypertrophic pyloric stenosis. Br J Surg 1982; 69:658-9. [2] Spitz L. Vomiting after pyloromyotomy for infantile hypertrophic pyloric stenosis. Arch Dis Child 1979;54:886-9. [3] Sommerfield T, Chalmers J, Youngson G, et al. The changing epidemiology of infantile hypertrophic pyloric stenosis in Scotland. Arch Dis Child 2008:1-12. [4] St Peter SD, Tsao K, Sharp SW, et al. Predictors of emesis and time to goal intake after pyloromyotomy: analysis from a prospective trial. J Pediatr Surg 2008;43:2038-41. [5] St Peter SD, Holcomb CM, Calkins GW, et al. Open versus laparoscopic pyloromyotomy for pyloric stenosis. A prospective, randomized trial. Ann Surg 2006;244:363-70.
1023
[6] Leclair MD, Plattner V, Mirallie E, et al. Laparoscopic pyloromyotomy for hypertrophic pyloric stenosis: a prospective, randomized controlled trial. J Pediatr Surg 2007;42:692-8. [7] Hall NJ, Van Der Zee J, Tan HL, et al. Meta-analysis of laparoscopic versus open pyloromyotomy. Ann Surg 2004;240:774-8. [8] Kim SS, Lau ST, Lee SL, et al. Pyloromyotomy: a comparison of laparoscopic circumumbilical and right upper quadrant operative techniques. J Am Coll Surg 2005;201:66-70. [9] Van der Bilt JDW, Kramer WL, van der Zee DC, et al. Laparoscopic pyloromyotomy for hypertrophic pyloric stenosis. Surg Endosc 2004; 18:907-9. [10] Carpenter RO, Schaffer RL, Maeso CE, et al. Postoperative ad lib feeding for hypertrophic pyloric stenosis. J Pediatr Surg 1999;34: 959-61. [11] Foster ME, Lewis WG. Early postoperative feeding-a continuing controversy in pyloric stenosis. J Royal Soc Med 1989;82:532-3. [12] Garza JJ, Morash D, Dzakovic A, et al. Ad libitum feeding decreases hospital stay for neonates after pyloromyotomy. J Pediatr Surg 2002; 37:493-5. [13] Georgeson KE, Corbin TJ, Griffen JW, et al. An analysis of feeding regimens after pyloromyotomy for hypertrophic pyloric stenosis. J Pediatr Surg 1993;28:1478-80. [14] Gollin G, Doslouglu H, Flummerfeldt P, et al. Rapid advancement of feedings after pyloromyotomy for pyloric stenosis. Clin Pediatr (Phila) 2000;39:187-90. [15] Lee AC, Munro FD, MacKinlay GA. An audit of post-pyloromyotomy feeding regimens. Eur J Pediatr Surg 2001;11:12-4. [16] Adibe OO, Nichol PF, Lim FY, et al. Ad libitum feeds after laparoscopic pyloromyotomy: a retrospective comparison with a standardized feeding regimen in 227 infants. J Laparoendosc Surg Tech A 2007;17:235-7. [17] Singh UK, Kumar R, Prasad R. Oral atropine sulfate for infantile hypertrophic pyloric stenosis. Indian Pediatr 2005;42:473-6. [18] O'Neill JA, Coran AG, Fonkalsrud, et al. Pediatric surgery. St Louis, MO: Mosby Year-Book; 2006. [19] Allan C. Determinants of good outcome in pyloric stenosis. J Paediatr Child Health 2006;42:86-8. [20] Golladay ES, Broadwater JR, Mollitt DL. Pyloric stenosis—a timed perspective. Arch Surg 1987;122:825-6. [21] Wilikes-Holmes C. Safe placement of nasogastric tubes in children. Paediatr Nursing 2006;18:9. [22] Richardson DS, Branowicki PA, Zeidman-Rogers L, et al. An evidence-based approach to nasogastric tube management: special considerations. J Pediatr Nursing 2006;21:388-93. [23] Ellett MC. What is known about methods of correctly placing gastric tubes in adults and children. Gastroenterol Nursing 2004;27:253-9.