OBJECTIVE. The objective of this study was to compare the respiratory physiologic features of healthy term infants placed in either a car bed or a car safety seat.
METHODS. Within the first 1 week of life, 67 healthy term infants were recruited and assigned randomly to be monitored in either a car bed (33 infants) or a car safety seat (34 infants). Physiologic data, including oxygen saturation and frequency and type of apnea, were obtained and analyzed in a blinded manner.
RESULTS. The groups spent similar amounts of time in the devices (car bed: 71.6 minutes; car seat: 74.2 minutes). The mean oxygen saturation values were not different between the groups (car bed: 97.1%; car seat: 97.3%). The percentages of time with oxygen saturation of <95% were also similar for the 2 groups (car bed: 18.3%; car seat: 11.8%). In both groups, a number of infants spent high percentages of study time with oxygen saturation of <95%. The 6 infants with the most time at this level were all in the car safety seat group (54%–63% of study time). Values for the 6 infants in the car bed group with the most time at this level were lower (20%–42%). This difference in the duration of oxygen saturation of <95% was not statistically significant. The mean end-tidal carbon dioxide levels and the numbers of episodes of apnea were similar for the 2 groups.
CONCLUSIONS. The respiratory physiologic features of infants in the 2 car safety devices were observed to be similar. Of note, substantial periods of time with oxygen saturation of <95% were surprisingly common in both groups.
Infant car seats play a critical role in the safe transportation of young infants and have reduced the rates of deaths and injuries during motor vehicle accidents. As with all safety devices, however, there are some limitations. Respiratory instability is a potential concern because of the upright position in the car seat. This is particularly true for premature newborns,1,2 which has resulted in the recommendation for car safety seat testing before discharge from the hospital for such infants.3,4 Concerns have also been raised about term infants. Merchant et al5 showed that mean oxygen saturation declined for both term and premature infants, reaching a nadir of 95% after ∼70 minutes of placement in a car safety seat; 7% of infants were noted to have oxygen saturation values of <90% for >20 minutes. In an earlier study by Bass and Mehta,6 selected, high-risk, term infants were found to have increased risks of developing hypoxia.
It has been accepted generally that the potential for adverse effects resulting from these hypoxic events is substantially offset by the crash protection afforded by these devices. Unfortunately, the portability of car seats and busy contemporary lifestyles are resulting in infants spending extended periods of time in the car seat for reasons other than transport. Callahan and Sisler7 found that, of 187 infants, 94% spent >30 minutes in seating devices (including car seats) every day. The mean time spent in seating devices was 5.7 ± 3.5 hours (range: 0–16 hours). Prolonged use of car seats by infants too young to sit unsupported also may result in prolonged periods of oxygen desaturation.
The hypoxia while in the car safety seat is most likely attributable to the relative vulnerability of the airway in premature and term infants.8–11 The cause of the airway narrowing is slouching of the head forward while the infant is asleep in the car seat, which results in closure of the mouth, pressing of the tongue against the posterior pharynx, and flexion of the airway. The shape of the newborn head, with a prominent occiput, and the reduced muscle tone encourages head slouching. A pilot study suggested that placement of an insert that accommodated the occiput, thus reducing the tendency toward flexion, improved airway size and reduced the frequency of episodes of oxygen desaturation.11 The flexion position, which is unique to car seats, may have important implications for the airway of newborns.
The American Academy of Pediatrics (AAP) recommends currently that infants with documented oxygen desaturation in a car seat should be transported in a car bed.4 Car beds should overcome the position-dependent airway instability. However, gastroesophageal reflux might occur in the flat position, which could result in apnea. The purpose of this study was to define whether term infants are at increased risk of apnea and oxygen desaturation when positioned in a semi-reclined car safety seat, compared with a car bed.
The institutional review board of the Massachusetts General Hospital approved the study. Informed consent was obtained from the parents of all participating infants. Sixty-seven healthy term infants were recruited from the newborn nursery. Delivery could be vaginal or cesarean section. Maternal health and medication usage can alter newborn physiologic features; therefore, children were included only if parents were healthy and used no medications other than vitamins. Inclusion criteria included weight of >3 kg, gestational age of >37 weeks, normal physical examination results, and normal vital signs. Exclusion criteria included any abnormality on examination and any maternal or infant medications other than iron and/or vitamins. In principle, it would have been preferable to study each child in both the car bed and car seat. This was not possible because of the length of time the child would have been away from the parents and because it would have interfered significantly with discharge planning. Therefore, each infant was assigned randomly to one device. Initially, we had planned to study each child for 90 minutes, to identify any changes with prolonged car seat usage. At the request of the institutional review board, the study was terminated if the child was excessively irritable.
The study was performed after feeding (∼15 minutes) and a diaper change. Each infant was assigned randomly to either a car bed (Dream Rides Ultra; Cosco, Columbus, IN) or car seat (Graco Snug Ride, model 8471UVB; Graco, Exton, PA). These models were chosen because they were the most popular car bed and car seat used in our area at the time of the study. Each infant underwent cardiorespiratory monitoring with the Sandman Elite sleep system (Sandman, Quebec, Canada).
Position in Device
The car seat position was identical to that used by Merchant et al.5 The infant's head position was maintained with placement of rolled towels on both sides of the head, with the towel extending above the ears. Towel rolls were rolled tightly and had diameters of 2 to 2.5 inches. The straps were adjusted to allow no more than 1 finger to fit between the strap and the clavicle, in accordance with the National Highway Traffic Safety Administration guidelines. The harness clip was placed at the armpit level. The seat was positioned so that the angle indicator was at the approved level (45° angle, relative to horizontal). After placement in the car seat position, the appropriate position was confirmed with a check-sheet. A similar protocol was adopted for the car bed. We adjusted the strap tightness and harness clip as described above. No towel rolls were used, and the bed was placed flat on the floor.
Polygraphic recordings (Sandman Elite sleep system) were made to record heart rate, respiratory rate, chest wall excursion, and oronasal airflow with a thermistor, and pulse oximetry was performed from the large toe of the right foot. Oxygen saturation was measured with a Mallinckrodt/Sandman digital recorder, which incorporates the Nellcor Oxismart XL advanced signal-processing oximetry technology to overcome motion and low-perfusion conditions beyond the abilities of other oximetry.12 Polygraphic data were collected for the entire study time. The same technician performed all of the studies and was not involved in the analysis. For this study, the following parameters were recorded for each infant in each position: (1) time and percentage of time with oxygen saturation of >95%, 90% to 94.9%, 85% to 89.9%, and <85%; (2) bradycardia, with a heart rate of <90 beats per minute; (3) time and percentage of time with end-tidal carbon dioxide levels of >40 mm Hg and <40 mm Hg; and (4) apnea (>10 seconds) frequency, duration, and type. A blinded reviewer performed the analysis. We excluded oxygen saturation data on the basis of lack of correlation of the pulse on the oxygen saturation recording and the electrocardiogram. To calculate the percentage of time with oxygen saturation of >95%, 90% to 94.9%, 85% to 89.9%, and <85%, a simple program was established to identify how many seconds were spent at each oxygen saturation level. Once we knew the number of seconds and the total length of the study, we calculated the percentage.
The planned sample size was 100 subjects per group, which, with a 2-sample, χ2 analysis with a 2-sided significance level of .050, would provide 80% power to detect a difference from 10% to 25% in the percentage of infants in each group categorized as having low oxygen saturation. However, because of unanticipated difficulties in recruitment, the study was stopped after a total sample of 67 subjects was obtained.
All data are presented as mean ± SD. Statistical comparisons of mean values between groups were based on Student's t tests.
A total of 67 infants were studied, with 34 in the car seat and 33 in the car bed. No infants had abnormalities identified or were receiving any medications. The characteristics of the infants and mothers in the 2 groups were similar (Table 1). Of particular note, the means and ranges of weights for infants in the 2 groups were similar. The ages at the time of study were similar in the 2 groups (car seat: 45.87 hours; car bed: 44.43 hours).
The 2 groups spent similar amounts of time in the devices (car bed: 71.6 minutes; car seat: 74.2 minutes) (Table 2). The mean oxygen saturation values were similar for the groups (car bed: 97.1%; car seat: 97.3%). Two infants (1 in each group) had their studies terminated prematurely (according to the protocol) because of persistent dips of oxygen saturation to <85%. The mean total time spent with saturation of <95% was greater in the car seat group (872 seconds) than in the car bed group (448 seconds). When this time was adjusted for study duration (percentage of study time with saturation of <95%), however, this difference did not reach statistical significance. Each group was assessed for the time spent at a number of oxygen saturation levels (Table 2 and Fig 1). The mean percentage of studies with saturation of <85% was ∼2% in both groups; however, the percentages of studies with saturation values between 85% and 89% and between 90% and 94% were higher in the car seat group than in the car bed group. In both groups, there were a number of infants who spent very high percentages of their study time with oxygen saturation of <95% (Fig 2). It is noteworthy that 6 infants with the greatest amounts of time with oxygen saturation of <95% were all in the car seat group. Six infants in the car seat group spent between 54% and 63% of the study with saturation of <95%, whereas the 6 infants in the car bed group with the most time with saturation of <95% spent between 20% and 42% of the study at that level. Each study was scored for apnea. Very little central apnea was seen in either group. The groups had similar numbers of obstructive and mixed apneic events. End-tidal carbon dioxide levels were low in each group. This was not unexpected, because the normal rapid respiratory rate did not allow plateauing of the end-tidal carbon dioxide trace.
Car safety seats, which have been recommended by the AAP since 1974, are essential safety devices that have resulted in substantial reductions in motor vehicle-related deaths.13 In 1986, it was first demonstrated that premature infants, because of their reduced muscle tone, might develop airway obstruction when placed in these devices.1 On the basis of that initial study, as well as a subsequent study with a larger group of infants, the AAP recommended a period of monitoring in the car safety seat before discharge for preterm infants (<37 weeks of gestation).3 If the infant demonstrates evidence of respiratory impairment in the car safety seat test, then currently it is recommended that the child be transported in a car bed.4 However, no study has compared directly the respiratory physiologic features in car safety seats and car beds. This study focuses on this area for the first time.
Despite the limited study size, significant physiologic changes were observed. Overall, it was surprising how much desaturation occurred in each group (Fig 2). The car seat group seemed to spend more time at lower saturation levels. We had expected that the car bed would perform better, because the airway did not seem to be as vulnerable in that device. This finding raises an important question, namely, why an infant should experience desaturation in a car bed. Partial obstructive apnea was seen in both groups, but the amount was not excessive. Previously, Tomkin et al11 corrected the upper airway vulnerability in car seats with the use of an insert that pushed the shoulders forward while allowing the head to extend. In this study, despite correction of the head position, some desaturation persisted.
It is possible that the desaturation is attributable to a cause other than airway closure. The harness is common to both groups. The tension of this harness was similar for the 2 devices. We speculate that this feature may contribute to the vulnerability to desaturation. The chest wall of infants is more pliable than the chest wall of adults. The rigidity of the chest wall keeps the lungs from collapsing in both adults and children. In infants, however, the more-pliable chest wall is also kept in check by a dynamic balance of the inspiratory and expiratory muscles. With the harness fitting snugly against the chest wall, this finely balanced system is likely to be tipped in favor of collapse. This change is likely to limit respiration by reducing tidal volume and residual volume; therefore, the respiratory reserve is reduced and lower oxygen saturation is expected, particularly when the system is stressed, such as during partial obstruction. However, from the perspective of crash protection, tight adherence of the harness to the chest wall is a critical feature for crash tolerance. Additional studies are needed to address the optimal tightness of the harness. Also common to both groups is the buckle attached to the harness. Because of its hardness and position on the center of the chest, it might increase the vulnerability of the chest wall of the infant. In addition, because infants are abdominal breathers, it is possible that compression on the abdomen is a factor contributing to respiratory compromise. Although this is possible, in our opinion it is less likely, because the harness is not focused in this area.
An interesting paradox might exist here. In the car seat, airway opening may be limited.1,5,11,14 In contrast, the effects of compression of the chest may be more severe in the car bed. In the car seat, the child is sitting, and diaphragmatic displacement downward is facilitated. This downward displacement might overcome the potential compression of the chest. Therefore, we postulate that the reasons for desaturation might be different for the 2 groups.
This clinical study has a number of limitations. Insufficient power might have made it less likely to identify differences between the groups. We did not anticipate the low oxygen saturation in the car bed group, and this lower value made differences harder to distinguish without larger study groups. In view of the findings for the car bed group, another control group in which the infant was placed flat on a bed without a harness would have been useful. Of interest, when the car bed group is compared with the Collaborative Home Infant Monitoring study group, in which oxygen saturation was monitored among normal infants,15,16 the car bed group had lower oxygen saturation than those historical control infants. To help exclude the possibility that these desaturations are normal occurrences among infants not placed in devices, we reviewed all of our data again and looked at the times when desaturation occurred. Our data completely parallel the data reported by Merchant et al,5 in which there was an increase in desaturation with progression of the study in both groups. This finding suggested that desaturation was device related rather than being related to the baseline characteristics of the child. Other possible informative control positions for future studies might include the placement of subjects in the car safety seat or the car bed with the harness very loose or detached. Another limitation is that we did not stage sleep. Sleep stage could be altered by the device, which might provide an explanation for these results.
Our studies were relatively short, at 70 minutes; a previous study showed that the extent of desaturation while the infant was in the car seat became more obvious with time.5 However, the nadir seemed to be at ∼70 minutes, and our study was within that period. In addition, it was not possible to compare each infant in both devices. Such studies would be ideal but, in view of the limited time infants spend in the hospital after birth, are difficult to perform. Also, all of the infants in our study groups were placed by a trained nurse with many years of experience in this area. In real-life conditions, we would expect much more variability in placement, because of both parental practice and differing characteristics of various car safety seat brands. These real-life conditions would need to be tested for generalization of these results to larger populations of infants.
Car safety seats and car beds remain very important transport devices for children. Overall, our study suggests that additional refinements of the car seat and bed are necessary to minimize or to prevent respiratory compromise. Such modifications are possible and need to be addressed seriously by the manufacturers of these devices. In the interim, we suggest caution when considering the use of these devices outside the setting of transportation in the car, for which they were designed.
This study was funded by Aprica, Osaka, Japan. This unrestricted grant provided a travel stipend to a conference for Drs Kinane, Bass, and Corwin.
- Accepted March 27, 2006.
- Address correspondence to T. Bernard Kinane, MD, Pediatric Pulmonary Unit, MassGeneral Hospital for Children, 55 Fruit St, Boston, MA 02114. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵American Academy of Pediatrics, Committee on Injury and Poison Prevention and Committee on Fetus and Newborn. Safe transportation of premature and low birth weight infants. Pediatrics.1996;97 :758– 760
- ↵American Academy of Pediatrics, Committee on Injury and Poison Prevention and Committee on Fetus and Newborn. Safe transportation of premature infants. Pediatrics.1991;87 :120– 122
- ↵Merchant JR, Worwa C, Porter S, Coleman JM, deRegnier RA. Respiratory instability of term and near-term healthy newborn infants in car safety seats. Pediatrics.2001;108 :647– 652
- ↵Bass JL, Mehta KA. Oxygen desaturation of selected term infants in car seats. Pediatrics.1995;96 :288– 290
- Roberts JL, Reed WR, Mathew OP, Menon AA, Thach BT. Assessment of pharyngeal airway stability in normal and micrognathic infants. J Appl Physiol.1985;58 :290– 299
- ↵Tonkin SL, McIntosh CG, Hadden W, Dakin C, Rowley S, Gunn AJ. Simple car seat insert to prevent upper airway narrowing in preterm infants: a pilot study. Pediatrics.2003;112 :907– 913
- ↵American Academy of Pediatrics, Committee on Injury and Poison Prevention and Committee on Fetus and Newborn. Auto safety for the infant and young child. Pediatrics.1974;5 :758– 760
- Copyright © 2006 by the American Academy of Pediatrics