633 - Wireless, Noninvasive Biosensors to Trend Physiologic Changes during Painful Stimuli in Patients with BPD

Publication Number: 4619.633

Deeb Omari, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States; Sofia Zitella, Northwestern University The Feinberg School of Medicine, Long Grove, IL, United States; Narayanan Krishnamurthi, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States; Aniruddh Srinivasan, Northwestern University The Feinberg School of Medicine, Chicago, IL, United States; sunju Kang, Northwestern University, Evanston, IL, United States; Tae Wan Park, Northwestern University, Evanston, IL, United States; Megan Lagoski, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States; john Rogers, Northwestern University The Feinberg School of Medicine, evanston, IL, United States; Susan M. Slattery, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States

Background: Infants in the NICU are subjected to painful stimuli during procedures, laboratory tests, and other stressors. Pain is difficult to accurately assess in infants, and those with bronchopulmonary dysplasia (BPD) present additional challenges. Accurately diagnosing pain is essential for treatment and to limit repeated stimuli effects on neurodevelopment, physiologic stability, and long-term stress response systems in this high-risk population.

Objective: In addition to vital signs, vocalizations, and movements used in neonatal pain assessment tools, we aim to determine whether infants with BPD exposed to painful stimuli exhibit changes in sensitive autonomic measures including heart rate variability (HRV), temperature, galvanic skin response (GSR), and cardiac sound intensity (CSI).

Design/Methods: Infants diagnosed with BPD (NICHD definition), and without other anomalies, with a scheduled laboratory draw, without acute pain and sedation management, and in the NICUs of Ann & Robert H. Lurie Children's Hospital and Northwestern Prentice Women's Hospital were monitored using wireless sensors recording heart rate (HR), high-frequency (HF) and low-frequency (LF) HRV, respiratory rate (RR), RR variability (RRV), continuous temperature, GSR, vocalizations, CSI, and limb movement during painful stimuli (heel lance/venipuncture) and baseline (quiet at rest). Descriptive measures (median, standard deviation, maximum, maximum-median difference) were compared using Wilcoxon signed-rank test.

Results: Ten infants were recorded (1 endotracheal intubation, 2 tracheostomy, 7 non-invasive positive pressure ventilation). Results of 9 patients were available (median corrected gestational age 43w1d ± 8w4d). Overall, maximum RRV and HR demonstrated a significant increase with pain (p=0.011; p=0.039) along with median number of limb movements (p=0.039). Maximum measures of CSI also approached significance (p=0.05). (Table 1)

Conclusion(s): Novel devices that capture measures of HR with RRV and CSI suggest increased sympathetic nervous system activation and reduced parasympathetic activity during painful stimuli in infants with BPD. These non-subjective Measures may enhance traditional evaluations of pain. Additionally, sensors that quantify movement may complement existing pain management tools in the NICU. Further evaluation of these findings in this population of infants is warranted.

Table 1: Maximum wireless biosensor measures to diagnosis pain during laboratory draws in infants with bronchopulmonary dysplasia in the NICU
microSiemens (µS), decibels (dB), units of gravity (g).
* p < 0.05


Figure 1: Chest and Limb Sensor Placement
Wireless, non-invasive bioelectric sensors placed on the chest and lower extremity continuously measure vital signs, high-frequency and low-frequency heart rate variability, respiratory rate variability, continuous skin temperature, galvanic skin response, vocalizations, cardiac sound intensity, and limb movement in an infant in the NICU to diagnosis pain.

Figure 2: Multimodal sensor waveform data from a single infant with bronchopulmonary dysplasia capturing physiologic measures during baseline and painful stimuli
Baseline (quiet at rest) = 1140-1380 seconds (shaded orange); pain (laboratory blood draw) = 840-1020 seconds (shaded purple). Pain was associated with increases in galvanic skin response, cardiac sound intensity, respiratory rate variability, and limb movement, with a relative reduction in high-frequency heart rate variability.