The Nonpharmacological Methods in Reducing the Pain Caused by Orogastric Tube Insertion in Preterm Infants

Orogastric tube feeding is a feeding method used to support the feeding of newborns who cannot be fed orally and OGT insertion causes pain. Ottawa Neonatal Pain Interest Group (2015) and the studies have shown that OGT insertion causes acute pain in newborns. It was stated that OGT insertion causes pain in newborns and non-pharmacological methods were effective in reducing this pain. However, there have been no studies in which non-pharmacological methods (expressed breast milk, swaddling and facilitated tucking methods) are alone and in combination with OGT insertion procedural pain for preterm infants. Therefore, the purpose of this study is to evaluate the efficacy of using expressed breast milk, swaddling and facilitated tucking methods together and alone in reducing the pain caused by OGT insertion process in preterm infants. Based on the above literature review and mechanisms, the investigators hypothesized that: (1) Swaddling, expressed breast milk, and facilitated tucking alone could be more effective than routine care on preterm infant pain before, during, and after OGT insertion procedure. (2) Pain occur less frequently before, during, and after OGT insertion procedure in preterm infants treated with combinations of swaddling, expressed breast milk, and facilitated tucking than in those receiving routine care. (3) Combined intervention of swaddling+expressed breast milk, and facilitated tucking+expressed breast milk could be more effective than any single intervention on preterm infant pain before, during, and after OGT insertion procedure.

Preterm infants were recruited by convenience sampling from level III neonatal intensive care units of three hospitals in Turkey/Antalya from November 2017 to 2018. Preterm infants meeting study criteria (n=219) from 2017 to 2018. Participation was refused by 24 parents who did not want their infants to be videotaped data collection (12), refused anything extra done to their infants (n=10), were not interested (n=2) and 8 infants did not need OGT insertion; thus, 187 infants participated in this study. Preterm infants who did participate did not differ significantly in terms of sex, GA, postnatal age, and body weight. To calculate study power, the investigators first determined that the effect size was 0.51, based on the mean Premature Infant Pain Profile (PIPP) scores, respectively, and the correlation (r=0.385) of PIPP scores between the six groups. Based on this effect size and a significance level of 0.05, the study power (two tailed) with 187 infants was 0.99. Based on the between- and within-group variances of PIPP scores during OGT insertion procedures the effect size was 0.33. Thus, a sample size of 187 preterm infants was sufficient. Preterm infants were randomly allocated before the OGT insertion by a neonatal nurse using a random closed envelope manner to one of the six groups: routine care group, swaddling, expressed breast milk, facilitated tucking, swaddling+expressed breast milk, and facilitated tucking+expressed breast milk group. After obtaining parental written consent, neonatal nurse who apply OGT insertion randomized the infant and learned the allocation group. The nurses could not be blinded to the allocation because of the nature of the intervention. However, the outcome assessment of the videos was blinded. OGT insertion procedure was conducted by the clinical nurse in NICU within the scope of treatment for preterm infants in the case when clinical physician was deemed as necessary. The assigned treatment condition was administered by one researcher. In the routine care group, while a neonatal nurse performed the OGT insertion procedure, physiological measurements of the highest value of heart rate and the lowest value of oxygen saturation were recorded by one researcher 1 min before the procedure, during the process and after the process in 1st and 2nd minutes acquired for each infant in the unit with an individual monitor. In the swaddling group, swaddling process was applied 10 minutes before the painful procedure. The swaddled newborns remained in the swaddling during the procedure and for 5 minutes after the procedure. In the study, breast milk was given slowly to the upper part of the tongue of each preterm infant from his/her own mother as a single dose before OGT insertion. In the expressed breast milk group, preterm was ensured to take all of 2 ml breast milk by reducing the aspiration risk of breastmilk. The breast milk was given without touching the tip of the injector to the newborn's mouth. The preterm infants were not allowed to suck the tip of the injector. In the facilitated tucking group, facilitated tucking was initiated 3 minutes before OGT insertion in order for the newborn to feel the fetal position and to cope with the painful procedure. The preterm was kept in fetal position during the procedure and for 5 minutes after the procedure. In the combined swaddling and expressed breast milk group, swaddling procedure was conducted by the researcher to the preterm infant 10 minutes before the OGT insertion procedure. 2 ml breast milk was administered by the researcher to preterm using a sterile injector as a single dose for 2 minutes before OGT insertion procedure. In the combined facilitated tucking and expressed breast milk group, facilitated tucking method was applied right after giving 2 ml breast milk by the researcher to the preterm 3 minutes before OGT insertion procedure. The Premature Infant Pain Profile (PIPP) scale was used for the pain assessment. Physiological indicators were continuously monitored and behavioral indicators (facial images) were videotaped by a real-time colour video recorder. The digital camera was fixed at a certain angle via the tripod 30 minutes before the procedure. Pain was scored from videotapes of infants' faces 10 min before, 2 min during, and 5 min after OGT insertion procedures. PIPP score was measured by two specialist pain doctors who were blinded to the study purpose, and to the study group allocation. The inter-rater reliability of the Premature Infant Pain Profile ranged from 0.97-0.99 (0.97 for baseline [phase 1], 0.97 phase 2, 0.98 phase 3, 0.99 phase 4, respectively). All preterm infants, heart rate and oxygen saturation were measured using an electrocardiographic bedside monitor and continuously recorded by custom computer software.

Each OGT insertion included four phases: (1) Baseline: 1 min of baseline was collected at the end of the 30 min without stimuli. (2) OGT insertion: conducting the OGT measurement, inserting and fixing OGT (3) Recovery: one min after OGT insertion. (4) Recovery: two min after OGT insertion. Heart rate, oxygen saturation, and PIPP scores were evaluated by four phases.

For the statistical analysis of the data obtained in the study, SPSS (Statistical Package for Social Science) for Windows 22.0 and SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA) packaged software were used. In the analysis of the measurements such as heart rate and oxygen saturation taken from the same newborns at different times, both graphical methods and repeated measures analysis of variance were used. Paired-sample t-test was applied to the features found to be significant as a result of repeated measures analysis of variance. Advanced analysis of the Bonferroni post-hoc test was also performed. Concordance between the two first evaluators for the PIPP measurements were evaluated using the intraclass correlation coefficient (ICC).

In order to compare the preterm infants with and without pain in six different application groups, those having PIPP values of ≥6 were defined as 1 "pain" and values less than 6 were defined as 0 "no pain". In the analysis of the repeated measure data obtained from the preterm infants in six application groups and at different measurement times, the Generalized Estimating Equation (GEE) method's multiple logistic regression models were used. When the observations are related with each other as in the data with repeated measures, GEE models give more effective and unbiased estimates than ANOVA-based models. Statistical significance was defined as p<0.05.