Standardized Evaluation of Cord Gases in Neonates at Risk for Hypoxic Ischemic Encephalopathy

Our institution is a freestanding, academic, tertiary-care children’s hospital that offers obstetric and neonatal services, with ∼4500 deliveries annually. Newborn services include a newborn nursery, level II NICU, and level III and IV NICU. TH is standard care for neonates with moderate to severe HIE with criteria as outlined in the National Institute of Child Health and Human Development (NICHD) whole body hypothermia study.⁷  Only neonates meeting clinical examination criteria for moderate to severe HIE on the basis of the modified Sarnat examination are cooled. Biochemical criteria include an umbilical-cord blood gas or postnatal blood gas within 1 hour of birth with a pH ≤7.0 or a BD ≥16 mmol/L. If pH was between 7.01 and 7.15 or BD was between 10 and 15.9 mmol/L, additional clinical criteria are required. These included an acute perinatal event (eg, late or variable decelerations, cord prolapse, cord rupture, uterine rupture, maternal trauma, hemorrhage, or cardiorespiratory arrest) and either a 10-minute Apgar ≤5 or assisted ventilation initiated at birth and continued for at least 10 minutes. In those eligible, whole body cooling to a target temperature of 33.5°C is initiated within 6 hours of birth, with a planned duration of TH for 72 hours.

Cord blood gases are obtained at the discretion of the obstetrician, generally for concerns about the fetal status before or during delivery. Cord blood gases were run on an ABL 800 blood gas analyzer, and results were released in the maternal electronic medical record (EMR). A modification in the calculation of the BD occurred after our QI intervention, resulting in fewer abnormal cord gases. The clinical laboratory notified labor and delivery with a critical result of pH <7, but there was not a standardized process for communicating results to the pediatrics team or evaluating neonates, including the need for follow-up postnatal blood gases and clinical neurologic assessments.

Within a QI framework, a standardized clinical care pathway for inborn neonates ≥36 weeks' gestation with abnormal cord blood gases was developed and implemented. The multidisciplinary QI team was led by neonatal hospitalists and neonatologists, in conjunction with input from obstetricians, the clinical laboratory, and nursing. Abnormal cord blood gases were defined as a pH ≤7.0 or BD ≥10 and based in part on screening criteria established by the California Perinatal Quality Care Collaborative.⁸  An abnormal cord blood gas was automatically flagged in the clinical laboratory reporting system and resulted in a direct call from the clinical laboratory to the in-house neonatal hospitalist. We worked closely with the clinical laboratory to ensure that they had set the correct cutoffs for flagging pH and BD results value and the correct neonatal hospitalist phone number for direct laboratory to physician calls.

The hospitalist followed a clinical pathway (Fig 1) centered on postnatal blood gas testing and standardized documentation of a modified Sarnat examination. The Sarnat examination is a standardized grading system for HIE, categorizing neonates into mild, moderate, and severe categories on the basis of how the neonate scores on 6 items (level of consciousness, spontaneous activity, posture, tone, neonatal reflexes, and autonomic nervous system).⁹  Didactic education and training were provided to all neonatal hospitalists by 2 neonatologists who specialize in neonatal neurocritical care on how to perform a modified Sarnat neurologic examination and on the NICHD eligibility criteria for TH. The differential diagnosis of neonatal encephalopathy and the mechanism of HIE was also discussed. Education also was given on the use of a dotphrase within the EMR to facilitate standardized documentation of the Sarnat examination and ensure all the appropriate information was gathered and documented.

To gain institutional buy-in before “go live,” the clinical pathway was reviewed at neonatology faculty meetings and with nursing staff leadership in labor and delivery and newborn nursery. The pathway was also e-mailed out to the newborn nursery doctors and obstetricians for feedback. As part of the NICHD Neonatal Research Network, our institution has been involved in whole body hypothermia clinical trials for >15 years, and, therefore, most providers were already familiar with HIE, TH, and the importance of identifying those who might benefit. This familiarity likely helped facilitate “buy-in” of the pathway. Labor and delivery providers suggested consistent language be used with parents to discuss why a neonate would need a Sarnat examination and postnatal blood gas measurement, specifically so that it is clear the infant does not have a diagnosis of HIE on the basis of the blood gas result alone. We developed standardized language to guide nurses and hospitalists that was focused on the abnormal cord blood gas being a potential sign of stress to the infant around the time of delivery and the benefit of examining the infant more closely. The term HIE or brain injury was avoided.

The final pathway was posted in the neonatal hospitalist’s workroom. Neonatal hospitalists attend all high-risk deliveries at our institution and are available 24 hours per day. In addition, the neonatal hospitalist group is small in size (<12 providers). If any questions or concerns arose, they could be quickly addressed via their group’s e-mail listserv or text string or at daily patient sign-out. Therefore, they represented an ideal group to lead implementation of the pathway and facilitate adoption.

After QI implementation (August 2016), neonates with abnormal blood gases were identified and tracked via a query of the clinical laboratory database, and an in-depth chart review was performed over August 2016 through June 2019. Additionally, neonates ≥36 weeks' gestation with abnormal blood gases over a 6-month period before QI implementation (January 2016 to June 2016) were identified, and data were collected for comparison with the post-QI time period.

Main process measures included the percentage of neonates with an abnormal cord gas who had a documented Sarnat examination (either specific examination time documented or when the documentation was signed) and a postnatal neonate gas, and percentage of abnormal cord gases for which a direct clinical laboratory to neonatal hospitalist call was documented by the clinical laboratory staff. Confirmation of direct contact was collected via chart review and on the basis of laboratory documentation in the EMR. Additional information collected included demographics, birth history, cord gas results, and the relevant clinical course, including need for TH.

Control charts for process measures were constructed with control limits of 3 SDs. To test for process change after implementation of the QI initiative, a special cause shift in the center line shift was considered at 8 consecutive points, either all above or all below the mean of the pre-QI period. If a process change was detected, a new center line and control limits were calculated for the post-QI period. Comparisons of clinical characteristics of neonates between the pre- and post-QI period were made by using a χ² test, Mann–Whitney U test, or Student's t test for categorical, ordinal, and continuous data, respectively. Statistical significance was set at P < .05. Data were analyzed by using Stata 13 (Stata Corp, College Station, TX).

This QI initiative was reviewed by the local institutional review board and determined to be a local QI project that did not meet the definition of human subjects research.

Over the 35-month period after the start of the QI initiative, 11 981 neonates ≥36 weeks’ gestation were born, of whom 203 (1.7%) had an abnormal cord gas. In comparison, over the 6-month pre-QI period, 2024 neonates ≥36 weeks’ gestation were born, of whom 104 (5.1%) had an abnormal cord gas. After the start of the QI initiative, an error was identified in the clinical laboratory calculation of BD. This error was corrected February 2017 and, subsequently, resulted in fewer abnormal cord gases (15 per quarter versus 30 per month). It is unclear how long this calculation error was occurring before the QI period.

Clinical characteristics of neonates with abnormal cord gases during the pre‒ and post-QI time period are shown in Table 1. In general, neonates were similar between periods, except those in the post-QI period had lower Apgar scores at 1 minute (P = .04) and 5 minutes (P < .01), lower umbilical-cord pH (P = .005), and higher umbilical-cord BD (P < .001) and were more likely to require continuous positive airway pressure (CPAP) (P = .008) or positive pressure ventilation (PPV) (P = .002) in the delivery room.

The percentage of neonates with an abnormal cord gas who were evaluated and had a documented Sarnat examination in the EMR (Fig 2A) increased from 16.1% (14 of 87) in the pre-QI period to 94.1% (191 of 203) in the post-QI period (P < .001). Postnatal blood gas measurement (Fig 2B) increased from 11.5% (10 of 87) in the pre-QI period to 93.6% (190 of 203) in the post-QI period (P < .001). Of the 203 neonates in the post-QI period, the documented Sarnat examination was performed within 2 hours after birth in 76% and within 4 hours after birth for 91%. For postnatal blood gas measurements, 70% were measured within 2 hours after birth and 90% were measured within 4 hours after birth. Of the 203 neonates who had an abnormal cord gas in the post-QI period, the laboratory made direct contact with the neonatal hospitalist to confirm the abnormal value in 187 cases (92.1%). Of the 16 neonates who did not have a documented laboratory call to the hospitalist, 10 abnormal cord gases (4.9%) were called back to the labor and delivery nurse or nurse caring for the neonate, and 6 (3%) were not called back to either a nurse or physician.

Throughout the post-QI study period, 15 of 203 (7.4%) neonates with an abnormal cord gas received TH. Of these, 13 (87%) were already admitted to the NICU at the time of notification from the laboratory, but 2 (13%) were identified in the newborn nursery in couplet care. The first was a 41 + 1 weeks’ gestation boy born via vaginal delivery with maternal chorioamnionitis. The neonate required 3 minutes of PPV and 1 minute of CPAP. At 12 minutes of age, the neonate had diminished tone and was taken to the newborn nursery for observation. At 40 minutes of age, the hospitalist was called for an abnormal cord gas (arterial pH: 6.94; BD: 14 mmol/L), and a Sarnat examination was performed at 1.1 hours of age with only the category of tone scoring as moderate. A postnatal gas was performed at 2.4 hours of age, with a pH of 7.24 and BD of 15 mmol/L. Per the pathway the neonate had a repeat examination at 2.5 hours of age. After repeat examination, 5 of 6 categories of the Sarnat examination were scored as moderate and thus the neonate met criteria for TH. TH was started by 2.75 hours of age. The second neonate was a 40 + 4-week boy with prolonged rupture of membranes and maternal chorioamnionitis born via cesarean delivery. The neonate required PPV for 5 minutes and began to breath spontaneously thereafter without distress. By 9 minutes of age, the neonate was well appearing but noted to have poor tone. The neonate was placed skin to skin at 12 minutes of age. The hospitalist was called for an abnormal cord gas at 40 minutes of age (arterial pH: 6.85; BD: 14.7 mmol/L). A Sarnat examination was performed by 1.1 hours of age, with 3 of 6 categories scored as moderate, and the neonate met criteria for TH. TH was started by 1.6 hours of age. Both neonates had brain MRIs that were normal.

We successfully implemented a standardized clinical care pathway to screen neonates with abnormal cord blood gases who are at risk for HIE and may be eligible for TH. Adoption of this pathway resulted in timely evaluations, including a documented Sarnat examination in 91% within 4 hours of age and postnatal blood gas evaluation in 90% of neonates within 4 hours of age. Practice changes were sustained throughout the post-QI study period.

During the study period, 2 neonates outside of the NICU with moderate HIE were identified as part of the standardized clinical care pathway who were eligible for and went on to receive TH. Both were identified promptly after a Sarnat examination and postnatal gas were performed within 2.5 hours of age, and TH was started at <2.75 hours of age in both. Although we can only speculate, without the standardized pathway, these neonates may have had a more delayed presentation, with a potential for late recognition outside the recommended 6 hour time frame to initiate cooling for optimal outcome. Overall, our standardized pathway led to the timely and consistent evaluation of at-risk neonates with abnormal cord blood gases and identified 2 neonates with moderate HIE outside the NICU.

The development and implementation of our QI pathway highlighted the need for a multidisciplinary effort that included obstetricians, nurses, and clinical laboratory services, in addition to pediatricians and neonatologists to make it successful. We formally worked with the clinical laboratory to set the laboratory limits for defining an abnormal cord gas and established a new protocol for a direct verbal notification process to the neonatal hospitalist. After our intervention, the laboratory notified the physician 92% of the time and only the nurse 5% of the time. Of the 6 abnormal cord gases that were not called to a provider, all had a BD of exactly 10. Further clarification to call the physician for a BD ≥10 resolved the problem.

An error in the BD calculation was discovered several months into our QI study period, resulting in overestimation of the BD, which had led to unusually high rate of abnormal cord blood gases. After correction of this calculation error in February 2017, fewer neonates subsequently had a BD >10, and, therefore, fewer needed to be screened postnatally. This might account for the higher rates of PPV and CPAP and lower Apgar scores seen in the post QI group because of the inappropriate classification of some neonates in the pre-QI period as having acidemia, despite a more clinically stable appearance. The laboratory calculation error did not appear to impact our process change because a high rate of documented Sarnat examination and postnatal gas measurement was seen throughout the course of the post-QI period (Fig 2). In addition, if we only examined neonates in the post-QI period who qualified on the basis of pH ≤7.0 criteria (which was not impacted by the laboratory calculation error), all 13 neonates before the error and 21 of 22 neonates after correction of the error had a documented Sarnat examination and postnatal gas measurement.

Neonates with less severe blood gas findings (BD: 10–15.9 mmol/L) represented the majority of patients evaluated (77%), with 46% of neonates having a BD of 10 to 11.9 mmol/L. None with a BD of 10 to 11.9 had abnormal neurologic signs on Sarnat examination, and none received hypothermia. Most were well-appearing newborns in our labor and delivery unit who proceeded with routine skin-to-skin care after birth. Because our intervention necessitates a brief period of separation of mother and neonate during the first hour of life to conduct our postnatal assessment, we have since updated our pathway so that neonates with a BD of 10 to 11.9 mmol/L no longer require an assessment unless they also meet secondary criteria for hypothermia (ie, 10-minute Apgar ≤5 or the need for assisted ventilation for at least 10 minutes after birth). Our updated pathway (Supplemental Fig 3) is in-line with criteria from the NICHD whole body hypothermia study, in which neonates with a BD of 10 to 15.9 mmol/L required additional secondary criteria to qualify for hypothermia.⁷  Although the optimal blood gas threshold to use for screening is not yet known, the cutoff used for cord pH or BD must consider the balancing measure of unnecessary interventions in the neonate and disruption of mother–neonate bonding.

Central to identifying neonates with HIE is the Sarnat examination, and educating neonatal providers was an important component in the implementation of our pathway. Neonatal providers received a didactic presentation on the Sarnat examination, incorporating video examples of normal, mild, moderate, and severe findings for the 6 categories.¹⁰  Severe findings are typically clear and easier to identify, but the differences between mild to moderate examination findings can be subtle and, examiner experience becomes an important variable. Further hands-on education may be helpful for providers to understand nuances associated with the Sarnat examination. This could be important because the use of TH in neonates with mild HIE is now being studied (www.clinicaltrials.gov identifier NCT04176471).

This QI study has several limitations. The timely evaluation of neonates was possible at our institution because of the presence of a 24/7 in-house neonatal hospitalist. Our approach, therefore, may not be generalizable to centers without in-house physicians. Such centers will need to individualize their approach to identify and screen neonates at risk for HIE. Our institution does not define criteria for when to obtain cord blood gases at delivery. Although the American College of Obstetricians and Gynecologists has published recommendations for when cord gases should be obtained, the decision to do so at our institution remains at the discretion of the delivering obstetrician.¹¹  Thus, variability in practice may have influenced our results, and there remains the possibility that neonates eligible for TH were not recognized, despite our QI process. An additional limitation is the limited amount of pre-QI data that we had. Although the pre-QI period comprised data from 6 months before the intervention, additional months of data would have been ideal. Finally, our sample size was not sufficient to detect meaningful differences in clinical outcomes on account of the low rate of abnormal cord gases and the small absolute number of HIE cases during our study period. However, the relatively small numbers of neonates identified through our process suggests that postnatal screening is feasible, with potential to identify eligible neonates for TH and improve clinical outcomes. Further experience will be needed in multiple care settings and in larger populations of at-risk neonates to substantiate the clinical utility of our pathway.

A standardized clinical care pathway for the screening of neonates with abnormal cord blood gases can help facilitate the timely identification and evaluation of neonates at risk for HIE. Adherence to the pathway was sustainable over the 35-month study period. Ongoing process improvement will be helpful to optimize the number of neonates to screen so that those with HIE are readily identified, while minimizing unnecessary evaluations and disruption of neonate–mother bonding.

We thank all of the neonatal providers who implemented this pathway to improve care for our neonates and families. We also thank our obstetric colleagues, our nursing staff, and the clinical laboratory for their collaboration and support for this QI initiative.