We performed a single-center pre-post intervention prospective cohort study and followed STROBE reporting guidelines [9]. Beginning in April 2020, a unit-wide clinical EPM program was launched in CHOP’s level-IV NICU. Implementation of the program was multifaceted and included development of a holding and mobility job aid inclusive of nurses, physical and occupational therapists, and caregivers; a mandatory electronic learning module for nurses, frontline clinicians, and physical and occupational therapists; supplemental bedside education of nursing staff by physical and occupational therapists; a hospital-wide standard therapy order; multiple educational reinforcement techniques via signage and visual reminders; and a rounding tool in which a question about the infants’ out of bed experiences was answered by the medical providers each Friday during clinical work rounds. Therapy is routinely ordered in the electronic medical record for all patients who are developmentally and clinically appropriate for movement as evidenced by ability to maintain state control and hemodynamic stability, and who do not fall under specific exclusion criteria (for example dependence on a high frequency ventilator or immediately post operative).

For this study, we enrolled 32 intubated infants with severe BPD who were managed by the NeoCLD Program between April 2019 and January 2022. The NeoCLD program is a multidisciplinary regional referral service specializing in the management of infants with severe BPD. Study participants were enrolled during two epochs: pre- and post-EPM implementation. Infants in epoch 1 were enrolled and completed study activities between April 2019 and April 2020 and infants in epoch 2 were enrolled between June 2020 and January 2022. We initially planned for 20 participants in each epoch, however therapy staffing was unexpectedly limited by hospital mandate in the setting of the COVID-19 pandemic. Thus, 16 subjects were enrolled in the pre-EPM implementation epoch; this number was then matched in the post-EPM implementation epoch. Eligible infants had a diagnosis of severe BPD, had a post menstrual age (PMA) of 40 weeks or greater, and were intubated with an endotracheal tube (ETT) at the time of enrollment. We excluded infants with diagnosis of critical airway, infants who were pharmacologically muscle relaxed, and infants for whom therapy was not ordered because they were deemed medically unstable by their clinical providers.

Clinical and demographic data were extracted from the electronic medical record. Baseline demographic data were infant sex, gestational age (GA) in completed weeks at birth, and birth length and weight. Additional clinical and outcome variables included: length of NICU and hospital stay, serial growth data, ultimate tracheostomy status, and survival.

A modified version of the Test of Infant Motor Performance (TIMP) was used to characterize short-term motor skill development in the study participants. The TIMP is a validated, standardized test of functional motor skills in infants that can discriminate among young infants with varying risk for motor delays, starting at 40 weeks PMA [10,11,12]. The TIMP was not developed for use in children who are intubated and our study team determined that most prone items were inappropriate for use in this population. Therefore, a modified version of the TIMP, which includes 13 observed items and 4 elicited items related to head control and is used routinely to assess developmental progress in patients managed by the NeoCLD program, was used as the primary outcome for the current study. Assessments are documented in a physical or occupational therapy progress note in the electronic medical record. For the purposes of this study, the TIMP was administered to each participating infant during routine therapy sessions on two occasions approximately 4–6 weeks apart, depending on therapist availability and infant clinical status, beginning no sooner than 40 weeks PMA.

To evaluate feasibility and demonstrate successful implementation of EPM, granular data on infant movement experiences were collected from the electronic medical record. These data included numbers of physical or occupational therapy sessions; positioning by therapists or nurses in supine, prone, sitting up in bed or in a seating device; skin-to-skin holding with a parent or caregiver; times out of bed (OOB) on the floor/mat; and positioning in tilt and side-lying during the time interval between the 2 TIMPs. Safety data included monitoring for any cases of unplanned extubations in study subjects, exposure to additional sedating medications (any opioid, benzodiazepine, or other medication administered for sedation and/or anxiolysis), and respiratory severity score (product of mean airway pressure and fraction of inspired oxygen) on days of TIMP assessments as indicators of infant tolerance of progressive mobility.

The planned sample size of 20 participants in each group was expected to provide >80% power to detect a difference in 3 points in the change in TIMP over time between the pre-EPM group (epoch 1) and post-EPM group (epoch 2). As described above, the final sample size was smaller due to pandemic-related impacts on therapist presence in our NICU.

The primary analysis is a difference in differences approach to compare the change in performance of the 17 items of the TIMP over 4–6 weeks in epoch 1 as compared to epoch 2. Bivariate comparisons between groups and between time points in each group were performed with Fisher’s exact, Kruskal-Wallis, and t-tests. We used linear regression models to evaluate change in elicited, observed, and total TIMP scores adjusted for baseline or starting values in the two study groups. All analyses were performed with STATA/SE v16.1 (College Station, TX) and p < 0.05 was considered significant. The study was deemed exempt by the CHOP Institutional Review Board; therefore, consent was not required.