Scientific and technological advances in neonatal care in the last decades have been fundamental to decrease mortality and morbidity in the neonatal population. Continuous development of chest physical therapy in neonatal care units has encouraged the improvement in techniques and resources specific to this population.1
However, unlike in children and adults, where the use of chest physical therapy has proven to be beneficial to cardiorespiratory function and lung compliance, there are still scarce and conflicting studies in neonates. These studies used small sample sizes and different experimental designs.2 Although no clear consensus exists, chest physical therapy has been practiced in neonatal intensive care units around the world, especially in patients with chronic lung disease.3
In neonatal care units, bronchopulmonary dysplasia (BPD) is the most common cause for respiratory therapy assistance.4 The prevalence of infants undergoing treatment for gastroesophageal reflux (GER) is significant in this patient group.5
Although GER commonly occurs in infants with very low birth weight (VLBW), the characterization of gastroesophageal reflux disease (GERD) in these children remains unclear. Thus, the definition of GERD and its implications for this age group are unknown.6 No definite conclusion has been reached and 24.8% of the infants with VLBW are still treated for reflux.5
The effects of chest physical therapy on GERD in infants born preterm and at term have been controversial because diagnostic standardization is lacking and respiratory therapy techniques differ among studies.7 – 11 Some authors have suggested that respiratory therapy could exacerbate GER, potentially resulting in aspiration and worse lung function.7 , 8 There are no specific studies concerning this issue in the neonatal period.
The ideal postprandial period to perform physical therapy is also another important aspect, since the occurrence of reflux episodes can vary widely during the first hours after a meal.12 Determining the most appropriate time for respiratory therapy in neonatal units is therefore of practical interest.
In this study, the association between the expiratory flow increase technique (EFIT) at 2 and 3 hours after feeding, and increased exposure of the esophageal mucosa to acid in infants born preterm with BPD was investigated.
It was hypothesized that the number of reflux episodes during EFIT may be greater, especially when it is performed 2 hours after feeding.12
A crossover experimental study was designed. Neonates were included at a gestational age less than 37 weeks13 with BPD, according to a definition by Bancalari: use of mechanical ventilation or oxygen therapy in the first week of life for at least 3 days, with persistence of respiratory signs and symptoms, and need for supplemental oxygen after 28 days of life, accompanied by chest x-ray abnormalities.14 Infants at least 28 days of age with a clinical indication for chest physical therapy met the inclusion criteria. Data were collected from August 2007 to August 2008 at the neonatal care unit of a university hospital. Malformations and risk of barotrauma were exclusion criteria.
We studied 18 subjects, 8 of whom had a clinical indication for esophageal pH monitoring. Signs or symptoms suggestive of GERD after the introduction of postural and dietary measures were considered clinical indications for pH monitoring.15 The following signs and symptoms were included: 3 or more episodes/day of regurgitation or vomiting always associated with late-onset apnea unresponsive to caffeine or apnea that persisted beyond 36 weeks of postconceptual age. Isolated desaturation (<87%) or bradycardia (heart rate < 80 bpm), back-arching and laryngeal stridor comprised other GERD signs. In these cases, the first 8 hours from a total of 24 hours were used for study purposes. For the remaining 10 subjects without symptoms of GERD, the examination consisted of a single 8-hour period of esophageal pH monitoring.
During this 8-hour period, 2 physical therapy sessions were performed. Each session lasted 10 minutes, 1 in the morning and the other in the afternoon. The sessions took place in a different postprandial period, at 2 and 3 hours after feeding (Figure 1). Subjects were randomized into 2 groups (Figure 1) to receive chest physical therapy according to a postprandial schedule (group 1: EFIT 2 hours after the morning feed, and group 2: EFIT 3 hours after the morning feed).
The EFIT was applied to all subjects by the main researcher. The original technique was developed by Barthe16 and adapted by Vinçon17 for infants born preterm. A “bridge technique” was passively performed due to the infants' age. The EFIT was initiated if esophageal pH of 5.0 to 6.8 was present for at least 60 seconds.
To perform this maneuver, the abdominal hand bridges the ribs and the thumb and forefinger (or middle finger) are used as pillars. The thoracic hand fits the size of the thorax and moves toward the abdominal hand, creating a downward compressing flow (see Video, Supplemental Digital Content 1, for a demonstration of the maneuver, http://links.lww.com/PPT/A27). The goal is to protect the abdomen, creating a mechanical limit to the thoracic hand. The lack of abdominal countersupport allows the increased pressure generated in the chest to be dissipated by a more compliant abdomen.16
Esophageal pH Monitoring
Semi-disposable pediatric pH catheters, 2.0 mm in diameter with an antimony electrode (Alacer) and LA-2 computer program, version 1.24b3 (Alacer, São Paulo, Brazil) were used. The pH catheter was inserted transnasally 1½ hour before starting pH recording, until a pH turning point below 4 (gastroesophageal junction) was identified. The final catheter tip position was determined by the mean distance between the turning point and distance from the gastroesophageal junction, based on a reference curve.18 Catheter fixation was performed 3 cm above the average value established. Its position was confirmed by x-ray and considered adequate when the catheter tip was situated at the level of the T6–T7 vertebral bodies. Researchers performed all pH-monitoring procedures.
The subjects were maintained in the supine position during the 8-hour study. Aimed at enhancing the sensitivity of the examination to acid episodes, a 5% glucose solution (pH∼4.5) was given before applying the EFIT. The volume was similar to that used in previous milk feedings (Figure 1). All subjects received expressed breast milk or formula by indewelling orogastric tube, 4 hours before glucose feeding. Drugs administered to the infants, for example, caffeine, budesonide, dexamethasone, and phenobarbital, were continued during examination. All infants were given oxygen by nasal cannula and intermittent feeding by orogastric tube.
The dependent variable was the reflux index (RI), defined as the percentage of time the esophagus is exposed to pH less than 4 during the recording period. The RI was studied in the 20 minutes preceding the EFIT and also during 10 minutes of the procedure. On examination, the descriptive variables gestational age, birth weight, gender, chronological age, and weight were also assessed.
Mean RI values were obtained before (14%) and during (1.4%) physical therapy performed 2 hours after feeding in a pilot study of 9 subjects. On the basis of these values, and selecting an alpha of 5% and a power of 80%, a sample size of 15 subjects was determined.19
To compare the numerical and categorical descriptive variables between time points (at 2 and 3 hours), we used Fisher exact test and the Mann-Whitney test.
In a comparison of the numerical variables between both time points (2 and 3 hours) and between both assessments (before and during therapy), analysis of variance for repeated measures was used for a crossover experiment. Tukey multiple comparison test was used to compare groups in each evaluation. The contrast profile test was used for intergroup comparison. Since a normal distribution was lacking, variables were transformed into ranks. SAS for Windows (Statistical Analysis System), version 9.1.3 SAS Institute Inc, 2002–2003 (Cary, North Carolina) software program was used for analysis. The alpha level for significance was set at 5%.
The study was approved by the research ethics committee and family members signed an informed consent form.
The 18 infants that were studied showed good tolerance to EFIT and pH monitoring. No complications were recorded during these procedures. The mean ± SD value of the RI obtained for the 18 subjects was 17.51 ± 15.08%, median = 17.15%, maximum = 62.10%, minimum = 2.20%. Nine subjects had a RI of 10% or more, during the 8-hour study. Four of these subjects were from group 1, and 5 from group 2. The descriptive variables corresponding to the study groups are presented in Table 1.
All subjects required ventilatory support during the first week of age and received surfactant, caffeine, and inhaled and systemic steroids. Oxygen was given by nasal catheter to all infants during the pH study. All subjects received parenteral nutrition until the total volume of enteral nutrition was reached. An orogastric tube was used for bolus feedings (every 2 or 3 hours) in all infants and the tube remained inserted during feeding intervals. Nine subjects were fed by continuous infusion or partially continuous infusion (intermittent feeding infused over 120 minutes). Gastroesophageal reflux disease was suspected in 8 infants who received ranitidine in addition to postural measures for a few weeks, without symptom remission. These infants then underwent pH monitoring.
The analysis of variance study for crossover analysis was focused on comparisons as follows: (a) between 2 groups (1 and 2); (b) between 2 evaluations (2 and 3 hours after feeding); and (c) interactions between groups and evaluations. In the comparison between groups, there were no significant differences between groups 1 and 2 (P = .245). In the time based comparison, a significant difference in the results (P < .001) was detected, between 2 and 3 hours after feeding, independent of the group (Table 2).
The Tukey test showed that 2 hours before EFIT, group 2 had a significantly higher RI than group 1 (Figure 2). Regarding group 1, no significant differences in RI were found before and during EFIT, in the 2-hour and 3-hour postprandial periods. Group 2, however, showed significant differences in the RI in the following paired comparisons: (1) before EFIT versus during EFIT at 2 hours; (2) at 2 hours versus 3 hours for evaluation before EFIT; (3) before EFIT (at 2 hours) versus during EFIT (at 3 hours) (P < .001). In addition, for this group a comparison between the RI before and during EFIT applied 3 hours after feeding was not significant (Figure 2).
Calculation of the power of this sample was 92.6% for a significance level of 5%, considering comparisons before and during therapy and between postprandial time points.
The results showed that the EFIT in the supine position was not associated with acid reflux episodes in premature infants with BPD, with or without symptoms suggestive of GERD.
In this study, the diagnostic method used to identify acid reflux episodes was pH monitoring, although multiple intraesophageal electrical impedance combined with pH recording (which captures acid and nonacid reflux) is replacing the single use of pH monitoring. However, the aim of this study was to evaluate acid reflux only. Since acid reflux is the most important aspect in the pathogenesis of GERD in infants born preterm,20 the choice of pH monitoring over impedance techniques was justified. In addition, impedance techniques have shown important technical limitations in detecting acid reflux in infants with VLBW.21 , 22
Since pH monitoring identifies only acid reflux episodes, the events occurring during the first 2 hours after feeding may not be detected due to neutralization of gastric acid contents by milk. Thus, to avoid a buffering effect, we chose to use a 5% glucose solution.23 , 24
Previous studies of traditional chest physical therapy techniques (postural drainage, percussion, vibration, cough stimulation, aspiration)7 , 8 , 25 have reported an increased RI, probably due to the use of these techniques in combination with refluxogenic positions (placing the subject in the Trendelenburg position).7 , 8
To date, only 2 studies investigating the effect of the EFIT on infants with GERD are known and none involves infants born preterm.10 , 11 In infants who wheeze with and without GERD diagnosed by scintigraphy,11 EFIT did not exacerbate reflux episodes. In another study, EFIT performed in the supine position significantly increased RI when compared with the elevated position.10
This study indicated that the modern EFIT was not associated with acid reflux episodes, even considering that a gastric tube was used to feed all subjects, which favors the occurrence of reflux episodes.26 In contrast, in the second postprandial hour both groups showed a decrease in RI during EFIT, with group 2 demonstrating a significant reduction. One possible explanation for this observation may be that the EFIT produces an increase in intrathoracic pressure, raising the pressure gradient in the lower esophageal sphincter preventing the occurrence of reflux. Further studies are required to confirm this hypothesis.
In this study, the subjects were evaluated during a period of clinical stability (average age: 50 days) using less aggressive therapy27 (EFIT in the “bridge mode”) to promote slow and prolonged expiration. Thus, secretions of the small bronchi are gently mobilized and moved into the proximal airways, probably reducing the risk of deleterious effects of physical therapy on blood flow to the central nervous system.
It is known that children fed with milk every 3 or 4 hours have more acid reflux episodes in the 3rd and 4th postprandial hours than in the 2nd postprandial hour.3 Our study showed a different result. In both groups, RI was higher in the 2nd postprandial hour than in the 3rd postprandial hour, which can be explained by prior acidic ingestion (5% glucose solution).
A significant difference in the mean RI was observed between both study groups before EFIT in the 2nd postprandial hour. The lower RI in group 1 can be justified by excessive salivation due to probe insertion, which favors an increase in esophageal pH.28 The same was not observed in group 2 when the study was performed several hours after probe insertion.
This study compared respiratory therapy at 2 postprandial time points (2 and 3 hours), confirming that there was no difference in the RI between either time point. One important practical implication of this result is that this technique can be achieved without restriction in these 2 time periods. Acid reflux episodes occur frequently in both periods.2
Nine subjects had a RI of 10% or more, during the 8-hour study, which could be considered diagnostic for GERD.29 However, the accuracy of the 8-hour pH records in relation to 24-hour records for the diagnosis of GERD has been questioned.30 In addition, a high RI in asymptomatic subjects is difficult to interpret. Thus a separate analysis of subjects with and without GERD cannot be performed.
The EFIT can be safely performed in infants born preterm with BPD, since there is no increase in the frequency of acid reflux episodes. Furthermore, a comparison between 2 postprandial periods (2 and 3 hours) identified no differences in esophageal acid exposure.
The authors thank the statistical analyst, Helymar da Costa Machado, from the Research Commission, Department of Biostatistics, University of Campinas (UNICAMP) for the statistical analysis of this study. The authors also thank the medical staff of the Division of Neonatology, Department of Pediatrics, Faculty of Medical Sciences, UNICAMP, Brazil.
1. Demont C, Vinçon C, Cambas CH, et al. Effets de la technique d' augmentation du flux expiratoire sur la resistance and oxigen saturation du premature a l'enfant a term. Ann kinésithér. 1999;(5):227–231.
2. Hough JL, Flenady V, Johnston L, Woodgate PG. Chest physiotherapy for reducing respiratory morbidity in infants requiring ventilatory support (Review). In: The Cochrane Collaboration. New York: John Wiley & Sons, Ltd; 2008;(3):1–24.
3. Lewis JA, Lacey JL, Herderson-Smart DJ. A review of chest physiotherapy in neonatal intensive care units in Australia. J Pediatr Child Health. 1992;(28):297–300.
4. Postiaux G. La fisioterapia respiratória en neonatologia. In: Fisioterapia respiratoria en el niño. 1st ed. Madrid: McGraw-Hill Interamericana de Espana, S.A.U; 2000.
5. Malcolm WF, Gantz M, Martin RJ, et al. Use of medications for gastroesophageal reflux at discharge among extremely low birth weight infants. Pediatrics. 2008;121(1):22–28.
6. Molloy EJ, Di Fiore JM, Martin RJ. Does gastroesophageal reflux cause apnea in preterm infants? Biol Neonate. 2005;87:254–261.
7. Vandenplas Y, Dierix A, Blecker U, Lanciers S, Deneyer M. Esophageal pH monitoring data during chest physiotherapy. J Pediatr Gastroenterol Nutr. 1991;13:23–26.
8. Button BM, Heine RG, Catto-Smith AG, Phelan PD, Olinsky A. Chest physiotherapy, gastro-esophageal reflux, and arousal in infants with cystic fibroses. Arch Dis Child. 2004;34:330–334.
9. Phillips GE, Pike SE, Rosenthal M, Bush A. Holding the baby: head downwards positioning for physiotherapy does not cause gastro-oesophageal reflux. Eur Respir J. 1998;12:954–957.
10. Demont B, Escurrou P, Vinçon C, Cambas CH, Grisan A, Odièvre M. Effects of respiratory physical therapy and nasopharyngeal suction on gastroesophageal reflux in infants less than a year of age, with and without abnormal reflux. Arch Fr Pediatr. 1991;48:612–615.
11. Ribeiro MA, Cunha ML, Etchebehere EC, Camargo EE, Ribeiro JD, Condino-Neto A. Effects of cisapride and chest physical therapy on the gastroesophageal reflux of wheezing babies based on scintigraphy. J Pediatr (Rio J). 2001;77:393–400.
12. Omari TI, Barnett CP, Benining MA, et al. Mechanism of gastro-esophageal reflux in preterm and term infants with reflux desease. Gut. 2002;51:475–479.
13. Ballard JL, Khoury JC, Wedig L, et al. New Ballard score, expanded to include extremely premature infants. J Pediatr. 1991;119:417–423.
14. Bancalari E, Gerhadt T. Bronchopulmonary dysplasia. Pediatr Clin North Am. 1986;1:3–27.
15. Newell SJ, Booth IW, Morgan ME, Durbin GM, McNeish AS. Gastro-oesophageal reflux in the preterm infants. Arch Dis Child. 1989;64:780–786.
16. Barthe J, Binoche C, Brossard V, et al. Techniques de kinésithérapie respiratoire. In: J. Pneumokinesitherapie. 2nd ed. Paris: Doin Éditeurs; 1990.
17. Vinçon C, Fausser C. Les techniques de kinésithérapie. In: Kinésithérapie Respiratoire en Pédiatrie - du Prématuré au Petit Enfant. 1st ed. Paris: Masson; 1989.
18. Omari TI, Benninga MA, Haslam RR, Barnett CP, Davidson GP, Dent J. Lower esophageal sphincter position in premature infants cannot be correctly estimated with current formulas. J Pediatr. 1999;135:522–525.
19. Medronho RA, Carvalho DM, Bloch KV, et al. Epidemiologia. São Paulo, Brazil: Atheneu; 2002.
20. Wenzl T. Data analysis and interpretation in impedance outcome studies. J Pediatr Gastroenterol Nutr. 2010;50(3):238–239.
21. Corvaglia L, Mariani E, Aceti A, Capretti MG, Ancora G, Faldella G. Combined oesophageal impedance-pH monitoring in preterm newborn: comparison of two options for layout analysis. Neurogastroenterol Motil. 2009;21:1027–1081.
22. Di Fiore JM, Arko M, Churbock K, Hibbs AM, Martin RJ. Technical limitations in detection of gastroesophageal reflux in neonates. J Pediatr Gastroenterol Nutr. 2009;49:177–182.
23. Awadalla JK, Hopper AO, Deming DD, et al. Is twenty four hour esophageal pH monitoring necessary for detecting gastroesophageal reflux in premature mechanically ventilated infants? Ped Research. 1994;35(4) (part 2):124A.
24. Sutphen JL, Dillard VL. Effects of maturation and gastric acidity on gastroesophageal reflux in infants. Am J Dis Child. 1986;140:1062–1064.
25. Orenstein SR, Orenstein DM. Gastroesophageal reflux and respiratory disease in children. J Pediatr. 1988;112(6):847–858.
26. Peter CS, Wiechers C, Bohnhorst B, et al. Influence of nasogastric tubes on gastroesophageal reflux in preterm infants: a multiple intraluminal impedance study. J Pediatr. 2002;141:277–279.
27. Delaunay JP. Conférence de consensus en kinésithérapie respiratoire. Place respective des différentes techniques non instrumentales de désencobriment bronquique. Paris: Cah kinésithér. 1998;192(4):14–22.
28. Vandenplas Y. Oesophageal pH monitoring hardware and software. In: Oesophageal pH monitoring for Gastro-oesophageal Reflux in Infants and Children. 1st ed. Chichester, England: John Wiley & Sons Ltd; 1992.
29. A standardized protocol for the methodology of esophageal pH monitoring and interpretation of the data for the diagnostics of gastroesophageal reflux. Working Group of the European Society of Pediatric Gastroenterology and Nutrition. J Pediatr Gastroenterol Nutr. 1992;14:467–471.
30. Rudolph CD, Mazur LJ, Lipitak GS, et al. Pediatric GE reflux clinical practice guidelines. J Pediatr Gastroenterol Nutr. 2001;32(2 suppl 2):S1–S31.