Continued Challenges of the Car Seat Tolerance Screen

Since 1991, the American Academy of Pediatrics has recommended all infants born <37 weeks gestation have a period of monitored observation in their car seat before discharge – the “car seat tolerance screen (CSTS)”¹ . This is based on evidence that infants in the inclined, semiupright position can experience cardiorespiratory instability. Although widely implemented, the CSTS has been a controversial topic among physicians who provide newborn care.²  Historically this was due, in part, to a variation in screening parameters, including what constitutes a “pass” or a “fail,” questions regarding the clinical significance of the cardiorespiratory instability, and lack of clear guidance about how to respond to CSTS failures.³  Indeed, citing lack of sufficient evidence to support the practice, the Canadian Pediatric Society discontinued their recommendation for routine CSTS for preterm infants in 2016.⁴  The literature seeking to address concerns around the practice of screening is unfortunately meager, and we lack the data necessary regarding outcomes for infants who either pass or fail the CSTS.

As providers of newborn care in 2 busy academic medical centers, the writers of this commentary hold differing views on the merits of CSTS, although both endorse practices and research aimed at identifying infants at risk for adverse outcomes. As such, we read with interest the article in this month’s Hospital Pediatrics by Harrison et al⁵  examining birth hospitalization length of stay and 30-day hospital revisits for a cohort of late preterm infants in 4 hospitals who follow the American Academy of Pediatrics recommendation for CSTS. The objective of this study was not to describe the diagnoses of those infants who failed the CSTS, but rather to investigate subsequent healthcare utilization. This retrospective study examined 5222 newborns over a 5-year period who qualified for CSTS based on gestational age or birth weight. Adherence to screening recommendations was high (91%), and CSTS failure rate is similar to that reported in the literature (8%). Using the small cohort of newborns who met criteria for screening but whom were not screened as the reference group, they report a similar length of stay for those newborns who passed the screening. Newborns who failed, however, had significantly longer lengths of stay. The increased time varied by location, with a mean of +12.6 hours in the newborn nursery and a mean of +71.2 hours in the NICU or pediatric unit. Consistent with outcomes reported in other studies, most newborns passed CSTS on subsequent attempts and were discharged (90%). Hospital revisits were low and did not differ significantly between the reference group (7.3%), those who passed CSTS (5.2%), and those who initially failed (4.4%). For those who did have hospital emergency department visits or readmission, the authors sought to characterize them as “definitely CSTS-related (eg, apnea while in the car seat), “potentially CSTS-related (eg, laryngomalacia) or “CSTS-unrelated (eg, fever). Interestingly, no infants who experienced a hospital revisit were categorized as “definitely CSTS-related,” and of those with the “potentially CSTS-related” diagnosis of brief resolved unexplained event on hospital revisit, all passed their CSTS before discharge.

On the surface, this retrospective study adds to the controversy by noting a longer length of stay for infants who fail their initial screen without subsequent healthcare utilization, at least in the short term. However, is a longer length of stay for infants who are known to be at increased risk of cardiorespiratory events who fail the CSTS beneficial if there is potential for uncovering a previously undiagnosed condition? Unfortunately, studies to date suggest such diagnoses are rarely made. Sharma et al reported outcomes for infants who failed CSTS with a 2 to 3 year follow up. They found an increase in diagnoses of respiratory conditions such as asthma, obstructive sleep apnea, and any albuterol use.⁶  These diagnoses were likely self-evident when clinically notable irrespective of a history of a failed CSTS. Magnarelli et al describes infants who failed CSTS in a 2020 retrospective review of 918 screened newborns. The authors note that for most infants who failed the initial CSTS, the majority passed on subsequent attempts. Their cohort included only 2 newborns who failed more than twice and underwent further evaluation that resulted in supplemental oxygen before discharge.⁷ 

The authors note that for any screening program to be justified, it must “uncover clinically significant pathology during the birth hospitalization that would otherwise go unnoticed” and must prompt “interventions…to improve patient outcomes.” This question of “clinically significant” proves difficult to predict. Shah reports 39% of preterm newborns placed on monitors after failing a CSTS experienced periods of apnea.⁸  It is unknown, of course, if these newborns were doing the same during their presumably asymptomatic hospital course before the screening, raising the question of clinical significance of apneic episodes that are self-limited and brief. Similarly, both term and preterm newborns monitored supine in a NICU observational study experienced multiple self-limited oxygen desaturation events – 4.7 episodes per hour for preterm infants, that would otherwise go unnoticed had they not been monitored for the purposes of this descriptive study.⁹ 

This is where much of the future discussion about CSTS will lie. Whereas it is fairly apparent that CSTS outcome will neither predict nor prevent future hospitalization or serious morbidity and mortality, it is also clear that there are fundamental changes in biomechanics and cardiorespiratory regulation for young infants when positioned at an angle. Studies commissioned by the US Consumer Product Safety Commission, and subsequently published in the peer-revied journals have demonstrated that “an incline of 20 degrees or greater is not safe” for infant sleep.¹⁰  This is due, in part, to the findings that “the trunk experienced increased flexion and limited movement, which is concerning for both breathing and the potential for chin-to-chest incidents compared to a flat crib mattress.”¹¹  These studies were conducted at angles >10 and <30 degrees from horizontal. Car safety seats position infants in a much more upright position, between 45 and 60 degrees from vertical, increasing the risk for cardiorespiratory instability.

Perhaps the outcome of interest is not catastrophic morbidity or mortality, but the cumulative effects of recurrent hypoxemia? In a survey of members of the Better Outcomes Through Research for Newborns, almost 90% of respondents stated they would have concerns about discharging an otherwise well 36-week newborn who desaturated while supine for 20 seconds or more.¹²  Should we not have the same concern for an infant placed in a fundamentally dangerous position in a product that they must use to be safely transported from hospital to home, and afterward? What is the cumulative impact of recurrent cardiorespiratory instability on infant brain development?

This paper adds to the urgency with which we must answer these questions. It, like many of its predecessors, reviews data from multiple institutions, each with different criteria for failure. This problem screams out for consistent protocols and coordinated, multicenter collaboration. We must shift the focus from the absence of existing evidence as justification for abandoning CSTS and acknowledge the biomechanical evidence. It is incumbent upon us as child health professionals and advocates for children to demonstrate evidence of absence of harm.