Received: 17 May 2001Accepted: 3 April 2002Published online: 10 May 2002© Springer-Verlag 2002

Abstract Objective: To test air insuf-flation as an adjunct to placement ofenteral feeding tubes and the effec-tiveness of using a smaller insuffla-tion volume in pediatric patients. Design and setting: A randomized,controlled study in a pediatric inten-sive care unit in two tertiary hospitals.Patients: A total of 78 children withindication for transpyloric tube feed-ing were studied. Interventions: Anunweighted feeding tube was placedinto the stomach through the nares; a20-ml syringe was used to insufflate10 ml/kg air into the stomach. Thetube was advanced an estimated dis-tance into the pylorus or beyond. Thecontrol group received the same pro-cedure except for air insufflation.Resident physicians performed allprocedures. Abdominal radiography

was performed 3 h later. Results: Of38 tubes in the study group 33(86.8%) were successfully placed in asingle attempt, compared to 18 of 40tubes (45%) in the control group.Compared with the technique of us-ing 20 ml/kg air for insufflation, nostatistically significant difference wasobserved. No significant complicationwas observed. Conclusions: The gas-tric insufflation technique required noexpensive equipment, minimal train-ing, and consistently allowed transpy-loric passage of feeding tubes. Theuse of 10 ml/kg air may significantlyimprove the rate of success withoutincreasing risks.

Keywords Insufflation · Enteral nutrition · Intubation · Critical care ·Intensive care units · Child

Intensive Care Med (2002) 28:943–946DOI 10.1007/s00134-002-1326-2 N E O N ATA L A N D P E D I AT R I C I N T E N S I V E C A R E

Paulo Sérgio Lucas da SilvaChiu Seing Tsok PauloSimone Brasil de Oliveira IglesiasWerther Brunow de CarvalhoFlávia Santana e Meneses

Bedside transpyloric tube placement in the pediatric intensive care unit: a modified insufflation air technique

Introduction

Since 1997 there has existed a consensus to recommendenteral nutrition as routine therapy practice in the manage-ment of critically ill patients [1]. The major limiting factorfor early enteral feeding is the impairment of gastric mo-tility following injury, severe illness, or drugs (gastropare-sis) [2, 3]. Several techniques [4, 5, 6, 7, 8, 9, 10, 11, 12]have been developed to introduce a feeding tube throughthe pylorus to overcome the impaired gastric emptying.Among these techniques, the bedside placement per-formed blindly is the least invasive, safest, and cheapestmethod. Several investigators have applied the gastric in-sufflation technique in adults to improve insertion of tubes[13, 14]. There is only one clinical study in children com-paring this technique with the standard insertion method

[15], which employs a total volume of 20 ml/kg air for in-sufflation, with a success rate of 92% on the first attempt.

We conducted a randomized, controlled study to eval-uate whether the use of lower air volume improves thesuccess rate of enteral tube placement and to determinepossible complications involved in such procedure.Moreover, this study was undertaken to establish a tech-nique which would enable any member of the health careteam to successfully place a nasoenteral feeding tubewithout the aid of costly equipment.

Patients and methods

This randomized clinical trial included all 78 children admitted tothe pediatric intensive care unit of two tertiary hospitals betweenOctober 1998 and October 2000 in whom there was an indication

P.S.L. da Silva (✉ ) · C.S.T. Paulo S.B. de Oliveira Iglesias · W.B. de Carvalho F. Santana e MenesesRua das Aroeiras, 30 ap 22. Bairro Jardim,Santo André, São Paulo, 09090-000, Brazile-mail: psls.nat@terra.com.brTel.: +55-11-49946952Fax: +55-11-2889928

of enteral tube feeding according to the ICU attending physician.Exclusion criteria were facial trauma, gastrostomy, and use of gas-tric prokinetic drugs. The study group included 38 children (age1–168 months, median 11; weight 2.0–56.0 kg, median 6.9) andthe control group 40 children (age 1–144 months median 14.5;weight 2.0–45.0 kg, median 8.8). Their diagnoses and demographiccharacteristics are presented in Tables 1 and 2, respectively. Therewere no significant difference between groups in demographic da-ta, Pediatric Risk of Mortality score [16], or requirement for me-chanical ventilation. The study was approved by our institutionalreview board; informed written consent was obtained from parentsor guardians.

Technique

Each insertion procedure was performed by a resident physicianunder supervision of one of the researchers; the procedure was ex-plained to the resident prior to performance. Bedside placementwas a modified procedure based on air insufflation as described bySpalding et al. [15]. An additional air bolus (10 ml/kg) was notutilized.

Feeding tubes

All patients received an unweighted polyurethane feeding tube (6, 8, or 10 Fr) with a flexible wire stylet (Flocare, Cow & GATE,Trowbridge, Wiltshire). The tube was chosen according to eachpatient’s size.

Standard insertion technique (control group)

Two identifying marks were made on each tube according to ex-ternal anatomical sites to assist with placement. The first mark(gastric) accounted for a nasogastric location of the tube based on

the distance between the nares and the xyphoid appendix by rou-tine measurement. The second mark (pyloric) was obtained byadding the distance between the xyphoid appendix and the rightlateral bottom rib’s margin. The distance between the two marksapproximates the distance from the gastroesophageal junction tothe pylorus. Patients were placed in elevated decubitus at 45° withthe neck in anteroflexion. If a nasogastric tube was present, it wasclamped prior to the insertion. A tube with the wire stylet in placewas lubricated with distilled water and inserted through the narewith progression of the tip parallel to the palate and nasal septumuntil the first mark was reached. Criteria for withdrawal werecoughing, oxygen saturation decrease, cyanosis, and resistance toprogression. Gastric location was confirmed by auscultation of10 ml air injected by a syringe over the epigastric region. The pa-tient was kept in right lateral decubitus if allowed by clinical con-dition, and the tube was then progressed towards the second mark.The patient returned to supine position.

Gastric insufflation insertion technique (study group)

The same technique as described above was used except for the at-tachment of a 20-ml syringe to the nasoenteral tube to allow theinsufflation of 10 ml/kg air into the stomach (maximum of500 ml) as soon as the first tube mark was reached. The patientwas then kept in right lateral decubitus, and the tube progressedtowards the second mark. Eventual nasogastric tubes wereclamped during the procedure until radiography was performed.The patient returned to supine position.

In both groups the position of the tube was confirmed by radio-graphy performed 3 h after the procedure and was considered trans-pyloric if the tip was located in the first portion of duodenum or be-yond. If positioning was correct, feeding was started immediately.

Statistical analysis

Differences between patient groups were analyzed by unpaired,two-tailed t tests and assuming different variants based on F testresults (previously applied to test for equivalence of variants). Thecategorical data were analyzed using a χ2 test (Yates’ correctionwhen necessary) or Fisher’s exact test where expected frequencieswere below 5. Nonparametric data were analyzed using the Mann-Whitney test. Data are presented as mean ±standard deviation. Weconsidered p≤0.05 to be significant.

Results

Utilizing the studied technique, 33 of 38 tubes (86.8%)were successfully inserted at the first attempt, compared

944

Table 1 Patient diagnostic categories

Diagnosis Study group Control group(n=38) (n=40)

Respiratory failure 10 13Neurological diseasea 9 7Trauma 1 0Exogenous intoxication 2 1Sepsis 6 13Burn injuries 3 2Cardiac disease 7 4

a Status epilepticus, Guillain-Barré syndrome, hypoxic-ischemicencephalopathy, meningoencephalitis

Table 2 Patient characteristics in each group (median standard de-viation)

Characteristic Study group Control group p

Age (months) 43.44±52.13 23.95±35.04 0.055Weight (kg) 14.49±14.02 9.43±9.31 0.064Pediatric Risk of 12.37±6.16 11.08±4.63 0.448Mortality scoreSex: male/female 18/20 16/24 0.669Mechanical ventilation 34/38 36/40 0.614

Table 3 Tube position after 3 h

Position Study group (n=38) Control group (n=40)

StomachCardia 2 16Antrum 0 2Pylorus 3 4

DuodenumFirst portion 9 8Second portion 16 4Third portion 8 5Fourth portion 0 0Jejunum 0 1

to 18 of 40 tubes (45%) in the control group (p<0.00026;Table 3). The time required to perform this procedure didnot exceed 10 min in either group. No complicationswere observed in either groups.

Discussion

Studies have shown spontaneous progression of enteraltubes in only 5–30% of critically ill patients [17, 18].Among blind bedside techniques used, gastric insuffla-tion has proved a useful tool. The rationale of using airinsufflation is that gastric distension opens the pyloricoutlet and also stimulates gastric contraction (motility).Gastric insufflation in critically ill adults has been usedwith various volumes of air, and the reported success ofbedside transpyloric placement of feeding tubes rangesfrom 66% to 93% [14, 19, 20]. Success rates may reach92% when the corkscrew technique is associated withthe air insufflation [7]. Nonetheless, this technique istime consuming and difficult to learn.

Krafte-Jacobs et al. [10] carried out a clinical trial in pe-diatric patients in which gastric insufflation was used fortranspyloric placement of pH-assisted feeding tubes. How-ever, the volume of air varied with the patient’s age, andthe study did not compare the standard technique and gas-tric insufflation. There is only one study in children com-paring gastric insufflation to the standard technique [15]. Atotal air volume for insufflation of 20 ml/kg was used, witha success rate of 92% and 100% in the first and second at-tempts, respectively. The air volume of 10 ml/kg utilized inour study is less than the gastric loading capacity and lessthan that in the previous study. This lower the rate of pro-cedure-related complications, such as abdominal disten-sion, aspiration, and gastric/duodenal perforation. We eval-uated the success rate only at the first attempt.

Interestingly, the age distribution is comparable tothat in the study by Spalding et al. [15], that is, the aver-age age of the studied group was twice that of the controlgroup. This difference was not statistically significantdue to a large standard deviation (inclusion of two 14-year old patients). The median age value in bothgroups is the best parameter for comparison.

The success rate for transpyloric tube insertion withthe standard technique (45%) was similar to that ob-served by other authors [4, 7, 15]. No significant differ-ence was observed for the volumes of air used in the twotechniques (10 and 20 ml/kg) after the first attempt(p=0.41972).

Zaloga [7] described success rates of 70 to 80%when the technique was performed by resident physi-cians. Our study differs from others that have comparedthe gastric insufflation to the standard insertion tech-nique [13, 14] since the procedures were performed byone or several experienced individuals in those trials.This could result in poor reproducibility of the results ifthe technique were performed by less experienced per-sonnel.

It is our impression that the simplicity and feasibilityof the studied technique (gastric insufflation) allows results superior to those of the standard technique even when performed by less experienced individuals.Salasadis et al. [13] showed that time required for theprocedure was 2–4 min when performed by experiencedpersonnel. As with Spalding et al. [15], we did not studythis variable. We believe that the additional time neces-sary for gastric insufflation is outweighed by the higherrate of success reached with such technique. In addition,the confirmation of a correct tube position after a shortperiod allows earlier enteral feeding.

One major flaw in this trial is that it was not blinded,and physicians knew whether they were using the stan-dard technique or the insufflation technique.

The gastric insufflation technique presents a highersuccess rate than the standard technique, even when per-formed by personnel in training. This is a valuable ad-junct that may result in earlier enteral feeding withoutthe need for complex procedures. The use of a lower airinsufflation volume (10 ml/kg) in the pediatric popula-tion does not alter the success rate significantly. Therewere no complications related to this procedure in ourstudy.

Acknowledgements This work was carried out at Hospital doServidor Público Municipal de São Paulo and the Escola Paulistade Medicina, Universidade Federal de São Paulo, Brazil.

945

References

1. Cerra FB, Benitez MR, Blackburn GL,Irwin RS, Jeejeebhoy K, Katz DP,Pingleton SK, Pomposelli J, RombeauJL, Shronts E, Wolfe RR, Zaloga GP(1997) Applied nutrition in ICU pa-tients A consensus statement of theAmerican College of Chest Physicians.Chest 111:769–778

2. Lanfranchi GA, Marzio L, Cortini C, etal (1978) Effects of dopamine on gas-tric motility in man. Evidence for spe-cific receptors. In: Duthie HL (ed) Gas-trointestinal motility in health and dis-ease. MTP, Lancaster, pp 161–171

3. Tarling MM, Toner CC, WithingtonPS, Baxter MK, Whelpton R, GoldhillDR (1997) A model of gastric empty-ing using paracetamol absorption in intensive care patients. Intensive CareMed 23:256–260

4. Levenson R, Turner WW, Dyson A,Zike L, Reisch J (1988) Do weightednasoenteric feeding tubes facilitate du-odenal intubations? JPEN J ParenterEnteral Nutr 812:135–137

5. Gabriel SA, Ackermann RJ, CastresanaMR (1997) A new technique for place-ment of nasoenteral feeding tubes using external magnetic guidance. Crit Care Med 25:641–645

6. Chellis MJ, Sanders SV, Dean JM,Jackson D (1996) Bedside transpylorictube placement in the pediatric inten-sive care unit. JPEN J Parenter EnteralNutr 20:88–90

7. Zaloga GP (1991) Bedside method forplacing small bowel feeding tubes incritically ill patients. A prospectivestudy. Chest 100:1643–1646

8. Kalliafas S, Choban PS, Ziegler D,Drago S, Flancbaum L (1996) Erythro-mycin facilitates postpyloric placementof nasoduodenal feeding tubes in inten-sive care unit patients: randomized,double-blinded, placebo-controlled trial. JPEN J Parenter Enteral Nutr20:385–388

9. Heiselman DE, Hofer T, Vidovich RR(1995) Enteral feeding tube placementsuccess with intravenous metoclopra-mide administration in ICU patients.Chest 107:1686–1688

10. Krafte-Jacobs B, Persinger M, CarverJ, Moore L, Brilli R (1996) Rapidplacement of transpyloric feedingtubes: a comparison of pH-assisted andstandard insertion techniques in chil-dren. Pediatrics 98:242–248

11. Hernandez-Socorro CR, Marin J, Ruiz-Santana S, Santana L, Manzano JL(1996) Bedside sonographic-guidedversus blind nasoenteric feeding tubeplacement in critically ill patients. Crit Care Med 24:1690–1694

12. Keidan I, Gallagher TJ (2000) Electro-cardiogram-guided placement of enter-al feeding tubes. Crit Care Med28:2631–2633

13. Salasidis R, Fleiszer T, Johnston R(1998) Air insufflation technique of en-teral tube insertion: a randomized, con-trolled trial. Crit Care Med 26:1036–1039

14. Schulz MA, Santanello SA, Monk J,Falcone RE (1993) An improved method for transpyloric placement ofnasoenteric feeding tubes. Int Surg78:79–82

15. Spalding HK, Sullivan KJ, Soremi O,Gonzalez F, Goodwin SR (2000) Bed-side placement of transpyloric feedingtubes in the pediatric intensive careunit using gastric insufflation. Crit Care Med 28:2041–2044

16. Pollack MM, Ruttiman UE, Getson PR (1988) Pediatric Risk of Mortality(PRISM) score Crit Care Med16:1110–1116

17. Zaloga GP (1998) Bedside placementof enteral feeding tubes in the intensivecare unit. Crit Care Med 26:987–988

18. Rees RG, Payne-James JJ, King C, SilkDB (1988) Spontaneous transpyloricpassage and performance of ‘fine bore’polyurethane feeding tubes: a con-trolled clinical trial. JPEN J ParenterEnteral Nutr 12:469–472

19. Ugo PJ, Mohler PA, Wilson GL (1992)Bedside postpyloric placement ofweighted feeding tubes. Nutr ClinPract 7:284–287

20. Kudsk KA, Croce MA, Fabian TC,Minard G, Tolley EA, Poret HA, KuhlMR, Brown RO (1992) Enteral versusparenteral feeding. Effects on septicmorbidity after blunt and penetratingabdominal trauma. Ann Surg215:503–513

946