Central & peripheral body temperature in VLBW preterm infants during the neonatal period: Relationship to neonatal infection and necrotizing enterocolitis Objective: We will measure continuous abdominal temperatures (AT) and foot temperatures (FT) in 440 preterm infants from five neonatal units over 28 days to predict occurrence of infection. Abnormal values of the central-peripheral temperature difference (CpTD), greater than 2 and less than 0, are a sign of autonomic instability occurring during infection. We will model the temporal relationship between abnormal CpTD values and the diagnosis of infection to find the best predictive model to use continuous temperature monitoring as a non-invasive, inexpensive predictive monitoring tool for infection. For comparison, we will measure continuous heart rate characteristics (HRC) from a HeRO monitor which predicts infection in preterm infants using heart rate variability. Predictive models of CpTD and HRC will be compared and combined to find the optimal predictive monitoring tool for infection in our sample. We will examine covariates of infants' demographic factors, maternal obstetrical histories, and infants' clinical context for interaction effects between our predictors of CpTD and HRC for infection. Methods: Infants will be enrolled from five NICUs in North and South Carolina. Research nurses or site PIs will obtain parental consent for infants to be enrolled up to 6 hours of age, if they are less than 32 weeks gestational age and less than 1500 grams at birth. Study personnel will attach a thermistor to the sole of the foot and to the abdomen for each study infant. Thermistors will be attached to a data logger which will measure temperature every minute for the first 28 days of life. A HeRO monitor will be attached to each infant's standard cardiopulmonary monitor and a USB drive in the HeRO monitor will collect hourly HRC scores for 28 days. Maternal obstetrical history, infant demographics and infant clinical context variables will be entered into a RedCap data base to be analyzed with physiological data. Data management and analyses will be conducted at University of South Carolina, College of Nursing. Data Analysis: Our interest for this study is the CpTD (AT-FT=CpTD). We will investigate that measure as two derived variables HTD (number/percentage of minutes with CpTd>2) and NTD (number/percentage of minutes with CpTD<0) which are continuous variables. Predictive models will be built utilizing HTD and NTD information defined over various lengths of preceding time (24 to 72 hours) to ultimately yield the optimal predictive model for diagnosis of infection based on CpTD. Our predictive model will be compared to the HRC model utilizing a Monte Carlo cross-validation approach in which we use 70% of the data for training and 30% for validation. Finally, we will combine the HRC predictive covariates to the CpTD covariates to investigate whether the combined model offers further improvements to prediction. If our CpTD predictive model improves the HRC model, it will serve as the foundation of a subsequent proposal to validate the CpTD predictive with new data.
Over 63,000 premature babies are born each year with a birth-weight less than 3 pounds and these vulnerable babies experience a high rate of infections while in the neonatal unit. Our previous study uncovered a relationship of abnormal patterns of abdominal and foot temperature with infection, but our sample of babies was too small to detect statistical significance. This research study focuses on detecting infection through constant monitoring of body temperature in approximately 500 babies to confirm the relationship of abnormal patterns of body temperature with infection.
