Abdominal Binders to Minimize Enteral Nutrition Disruptions for Preterm Infants On Non-Invasive Respiratory Support

Background Preterm neonates who require respiratory support benefit from the use of non-invasive respiratory support over invasive mechanical ventilation to decrease rates of chronic lung disease [1]. Non-invasive support typically involves positive end-expiratory pressure which allows for distending pressures to keep the alveoli of the lungs open [1]. In our institution, this typically includes continuous positive airway pressure (CPAP) or non-invasive mechanical ventilation (NIMV). Mechanistically, due to the positioning of the trachea adjacent to the esophagus, the provision of pressure support for the lungs often results in additional air and pressure being supplied to the enteric tract. This can result in what is called "CPAP belly" [2]. Although CPAP belly was initially described as a benign gaseous distension, emerging evidence shows it has negative impacts on the functionality of the enteric tract, including delaying gastric emptying time [2,3]. A recent study characterizing CPAP belly has shown that it can lead to longer time to full enteral feeds, disruptions to feeding practices, and additional abdominal radiography [4].

Several strategies have been tried to decrease feeding intolerance in infants with abdominal distension, including early enteral feeds, which is our current practice, abdominal massages, and probiotic therapy [5,6]. Recently, a novel idea was proposed to combat the progression of CPAP belly: abdominal binding to support the abdominal wall musculature [7]. It was initially reported as a case study of two infants with significant abdominal distension while on respiratory support who had improvements in their distension back to expected range within one week of therapy initiation and without any reported adverse effects [7]. The author of this article has created a product, the NeoBellyBand for sale by the DandleLION Medical company, for use in her own institution. It is now in use in several institutions, given its minimal risk to patients and the potential benefits, as the new standard of care. In speaking with the developer of this product, we understand that there are also ongoing pre-and post-intervention studies looking at its impact on decreasing rates of chronic lung disease or length of stay in the NICU.

There is little published evidence for or against the use of abdominal binding in this population and there are no current studies that the research team is aware of looking at its impact on feeding practices in the NICU. Prior to the use of abdominal binding, we would like to study its safety and efficacy for the neonatal population. We hypothesize that the use of abdominal binders will lead to a decrease in feeding interruption events as well as a decrease in adverse feeding and respiratory outcomes for preterm neonates on non-invasive respiratory support.

The specific aims of this study are to:

1. Evaluate the impact of abdominal binders on decreasing interruptions to our unit feeding protocol due to feeding intolerance symptoms, such as increased abdominal distension or high-volume gastric residuals, compared with our current standard of care.
2. Analyze the effect of abdominal binding on time to major neonatal milestones, including discontinuation of respiratory support, discontinuation of parenteral nutrition, full oral feeds, discontinuation of therapies to promote regular stooling (such as glycerin suppositories or prune juice), and discharge from hospital.
3. Determine how the use of abdominal binders changes respiratory management (such as increasing respiratory pressures for worsening distension) and respiratory outcomes (such as chronic lung disease).
4. Evaluate the safety of abdominal binders by tracking possible adverse effects, including temperature regulation concerns, skin integrity concerns, and incidence and complications associated with umbilical or inguinal hernias.
5. STUDY DESIGN We propose a randomized controlled trial of hospitalized preterm neonates born less than 32 weeks and 0 days gestational age (GA) on non-invasive respiratory support to evaluate respiratory and enteral nutrition outcomes of infants using abdominal binders compared with infants who are not using them.

Primary outcome Days with interruptions to the feeding protocol, defined as interruptions to feeds or advances per our unit protocol for the infant's GA and birth weight (BW), due to feeding intolerance, defined as abdominal distension, increase in abdominal girth, abnormal stooling pattern, abdominal tenderness, increase in gastric residuals, emesis, and feeding-related cardiorespiratory events.

Secondary outcomes

• Age at which infant first achieved 120 mL/kg/day of enteral feeds, which is when IV fluids are typically discontinued in our unit
• Length of hospital admission and post-menstrual age (PMA) at discharge
• Number of abdominal radiographs obtained for abdominal concerns, as noted on the indication for the image or otherwise noted in the chart
• Rates of chronic lung disease (CLD), defined as oxygen requirement at 36 weeks PMA
• PMA when successfully off all respiratory support
• Rates of abnormal abdominal circumference (AC): head circumference (HC), namely AC:HC, defined as AC:HC >95th %-ile for GA and PMA
• PMA at which infant achieves full oral feeds

Inclusion criteria for infants Born at <32w0d GA Chronologic age >7 days, treatment or control protocol started by 21 days of life On non-invasive positive pressure respiratory support (NIMV or CPAP), either as primary support or after extubation from invasive mechanical ventilation Has received 80 mL/kg/day of feeds at least once

Exclusion criteria for infants Skin integrity concerns Known abdominal or respiratory anomalies Umbilical lines in place Transferred to AMC NICU after 7 days of life Known genetic or major anatomic anomalies

Study Methodology Preterm infants who meet inclusion criteria will be identified upon admission to the AMC B4 NICU by the study team. They will be enrolled within the first 2-3 weeks of life by one of the physician study team members after obtaining written consent from the legal guardians.

There is no current method of tracking interruptions to feeds or disruptions to feed advancement. The study team did a preliminary analysis of infants born at <32w0d who were discharged between September 2024 - January 2025. We found the mean number of feeding interruption days to be 1.9 and the standard deviation to be 2.7 for the duration of their hospital admission, using documentation in daily progress notes and nursing notes. Of note, one outlier infant with 69 feeding interruption days was excluded from analysis. However, we do not believe that all feeding interruption days are being captured given our understanding of the current arrangement for documentation. Therefore, we estimate that infants may have a mean of ~3 feeding interruptions during their hospital course. A power analysis was performed to detect a decrease in feeding interruption days with a mean of 3 days per infant (and a presumed standard deviation of 2.7) to a mean of 1 day per infant with 80% power and an alpha of 0.05. An adequate sample size would be 74 infants, 37 in each group (case and control), who complete the study. Given that we are allowing families the ability to withdraw from the study at any point in time, we will aim to enroll ~100-120 infants to ensure that at least 74 infants complete the study.

This study is expected to be completed within the next three years, or once 120 patients have been enrolled in and have completed the study, whichever is sooner. We currently have ~120 annual admissions of infants at this gestational age to the NICU. Assuming 25% declines enrollment and a smaller percentage do not meet the inclusion criteria, we would still have adequate numbers to complete this study within this time frame.

The infants will be participating in the study during their admission in the NICU, the duration of which ranges anywhere from 2 weeks to 4-6 months.

Study Design Randomization will occur using a block method within two separate strata, infants ≤28 weeks GA and infants between >28 weeks to 32 weeks and 0 days. We will use block assignments of 8 infants, indicating that for each group of 8 infants enrolled, 4 will be assigned to the control group and 4 will be assigned to the treatment group. Unfortunately, we are unable to use blinding due to the treatment being a physically apparent intervention and the control being no intervention (which is our standard of care currently). The study team would ensure that infants allocated to the treatment group would receive the treatment starting between 8-21 days of age and once the infant has received at least one feed of 80 mL/kg/day, depending on the infant's skin maturity, as determined by the study team and in conjunction with the treating medical team. Infants who are intubated in the first week of life would qualify if they can be safely extubated within 2-3 weeks of life. Band placement would be started when they are extubated and on non-invasive support.

Device The device used in the treatment arm of this study will be the DandleLION NeoBellyBand. The NeoBellyBand is a Class 1 exempt abdominal binder which is included in DandleLION's Quality System and FDA registration. It is listed as a Medical Therapeutic Binder under product group MDR on Tarry Medical Products FDA establishment listing and does not require 510K clearance. (See: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfrl/rl.cfm?lid=872457\&lpcd=MDR) Per the manufacturer, it is made of a medical-grade blend of fabric and foam that has been used safely on premature skin and has a soft foam inlay that is lightweight, breathable and moisture-wicking.

Treatment Protocol The band is intended to be in place at all times outside of cares times, skin-to-skin kangaroo care with the family, baths, and any procedures involving the abdomen. Once the bands have been placed on, they will stay in place while the infant is on non-invasive or invasive pressure support (CPAP, NIMV, or MV) and will be weaned off once the patient has been stable on high flow nasal cannula (HFNC) for two days. This means that for patients who have been exposed to non-invasive pressure ventilation and require re-intubation, the bands will also be worn during the time they remain re-intubated. The weaning will include three days of a schedule of 6 hour stretches with the band on and 6 hour stretches without the bands. If the patient requires pressure support again after the bands have been weaned off, the bands will be placed back on until they're weaned off pressure support again and weaned in the same way as above.

Control Protocol Infants will receive the standard of care for the NICU, except for additional data collection, which will include data on feeding interruptions and abdominal circumference measurements.

Data Collection from Chart The information that will be collected from the medical chart is information that is currently collected for patient care. The exceptions to this are the paper chart that tracks feeding disruptions and the nursing flow-sheet that tracks the times that the band is on or off, comments for when the band is off, and the abdominal circumference weekly. Data collected will include demographic information such as name, MRN, date of birth, sex, race and ethnicity, gestational age at birth and birth weight. It will include process measures, such as length of time the bands remained on the infant, documentation of abdominal circumference, abdominal circumference to head circumference measurement ratios, number of radiographs obtained for feeding intolerance, and glycerin use. It will include outcome measures, such as number of feeding interruption days, time to goal enteral feeds, time to all oral feeds, rates of chronic lung disease, necrotizing enterocolitis, spontaneous intestinal perforations, hernias and rates of incarcerated hernias, length of stay, and post-menstrual age at discharge.

Data Protection The study team members will collect retrospective data from their medical records securely, either while at the hospital, or through the secure Citrix system. The data will then be stored in a secure database, the RedCap database, for further analysis. The data contained within the RedCap database will be password-protected. Although the data collected will include protected health information and identifiable health information, all data will be de-identified prior to publication. Once the research study has been published, all identifiable data will be destroyed. De-identified data will be preserved as many journals now require a data availability statement.

Data Analysis For the primary outcome, the mean, standard deviation, standard error of the mean, and 95% confidence intervals will be calculated. The means will be compared between the cases and controls using a two-tailed t-test. If the primary outcome is noted to have a non-normal distribution, medians and interquartile ranges will be calculated and non-parametric Mann Whitney analyses will be conducted. For infants who do not withdraw from the study, an intent-to-treat analysis will be conducted. For secondary outcomes which have a continuous variable, the same statistical analyses will be used as the primary outcome. For secondary outcomes with a categorical variable, we will use proportions as well as chi-square tests. The statistical analyses will be conducted on the SPSS software and reviewed by a statistician.

Risk Monitoring Although we anticipate minimal risk to infants based on prior published data available about the NeoBellyBand, we will be tracking adverse events weekly. Rates of adverse events between the treatment and control arm will be analyzed after enrollment of the first 25 infants and will include documented pressure injuries and over-heating requiring removal of the band for >6 hours. Although we do not anticipate any internal abdominal pathology to be a result of the placement of the bands, we will also be tracking rates of necrotizing enterocolitis, intestinal perforation, surgical abdomen, and incarcerated or strangulated hernias. Once the initial 25 patients have been enrolled (ideally about 11-14 from control group and treatment group), we will not start research interventions for new patients. We would continue to recruit more patients but will not carry out the research intervention until safety monitoring has been completed. The risk monitoring data will be reviewed by the research team, but also independently reviewed by a non-conflicted neonatologist at an outside institution.