Valerie A. Stebbens

Arterial oxygen saturation in preterm infants at discharge from the hospital and six weeks later

To obtain normal data on arterial oxygen saturation (SaO2) in preterm infants and to study early developmental changes in SaO2, we obtained overnight tape recordings of SaO2 and breathing movements in 160 preterm infants at their discharge from three special care baby units (mean gestational age at birth 33 weeks; at time of study, 37 weeks). One hundred ten infants (69%) underwent a second recording 6 weeks later. Median baseline SaO2 during regular breathing was 99.5% (range 88.7% to 100%) at discharge, and 100% (range 95.3% to 100%) at follow-up (p less than 0.001). The number of episodes of desaturation, defined as a fall in SaO2 to less than or equal to 80% for at least 4 seconds, corrected to the mean duration of recording (12.2 hours), decreased from a median of 3 (0 to 355) to 0 (0 to 17) (p less than 0.001). The median duration of each episode of desaturation remained unchanged (5.2 (4.0 to 22.7) vs 5.5 (4.2 to 24.0) seconds). At discharge, a small minority of infants had a clinically unrecognized low baseline SaO2 (lowest, 88.7%; 5th percentile, 95.7%) or a high number of desaturation episodes (the highest was six times the 95th percentile value). At follow-up, all outlying values had normalized. Follow-up recordings made between 42 and 47 weeks of gestational age (n = 53) were compared with similar recordings from 67 term infants at the same gestational age. The preterm infants had a significantly higher baseline SaO2 and no more desaturation than the infants born at term. Knowledge of normal ranges of oxygenation and their changes with age may be of value in identifying clinically undetected hypoxemia in preterm infants at discharge from the hospital. The potential influence of such hypoxemia on clinical outcome remains to be determined.

The Relationship between Bradycardia, Apnea, and Hypoxemia in Preterm Infants

ABSTRACT: The pathogenesis of bradycardias in preterm infants is poorly understood. Because their pathogenesis may involve both apnea and hypoxemia, we set out to analyze the proportion of bradycardias that were associated with an apneic pause and/or a fall in arterial oxygen saturation (Sao2), and the temporal sequence of the three phenomena, in overnight tape recordings of Sao2 (Nellcor N100 in beat-to-beat mode), breathing movements, nasal airflow, and ECG in 80 preterm infants at the time of discharge from hospital. A bradycardia was defined as a fall in heart rate of ≥33% from baseline for ≥4 s, an apneic pause as a cessation of breathing movements and/or airflow for ≥4 s, and a desaturation as a fall in Sao2 to ≤80%. A total of 193 bradycardias were found in 46 (58%) of the recordings (median, three per recording; range 1–18). There was a close relationship between bradycardias, apneic pauses, and desaturations: 83% of bradycardias were associated with apneic pauses and 86% with desaturations. Where all three phenomena occurred in combination, the time from the onset of apnea to the onset of the fall in Sao2 was shorter (median interval, 0.8 s; range −4.9-+11.5 s) than that from the onset of apnea to the onset of bradycardia (median, 4.8 s; range −4.0-+14.0 s). Hence, most bradycardias (86%) commenced after the onset of the fall in Sao2. We conclude that bradycardia, apnea, and hypoxemia are closely linked phenomena in preterm infants.

Apnoeic episodes induced by smothering: two cases identified by covert video surveillance.

Recurrent cyanotic episodes associated on some occasions with loss of consciousness due to cerebral hypoxia were investigated by long term tape recordings of breathing activity, oxygen saturation, air flow, electrocardiographic activity, and in some cases electroencephalographic activity. In 51 infants and children the mechanisms for the cyanotic episodes were identified (prolonged expiratory apnoea in 45, sleep related airway obstruction in three, seizure induced apnoea in one, behaviour induced apnoea in one). In one child apnoea was suspected as being caused by suffocation (smothering) by the mother. This was confirmed after enlisting the help of the police, who undertook covert video surveillance during cyanotic episodes. Each cyanotic episode was associated with a pattern of disturbance on the multichannel tape recordings which may be pathognomonic of this type of apnoea. A second infant with cyanotic episodes in whom smothering was suspected was referred for similar investigation after the availability of video recordings became established. Maternal smothering was again supported by specific patterns on multichannel tape recordings and confirmed by video surveillance. Diagnosis by video surveillance produces unequivocal evidence in these cases and avoids the need for medical and nursing staff to confront the mother with a possibly incorrect suspicion or in a court of law.

Arterial oxygen saturation in preterm neonates without respiratory failure

To obtain normal data on arterial oxygen saturation as measured by pulse oximetry (SpO2; Nellcor N200), we obtained 12-hour tape recordings of SpO2, photoplethysmographic waveforms, instantaneous pulse rate, and observations of breathing movements on 55 preterm neonates (25 girls) who had been admitted to one of four special care baby units but had no signs of respiratory distress and were breathing room air at 24 hours of age. Their median gestational age at birth was 35 weeks (range, 30 to 36), and their median age at the time of study 1 day (range, 1 to 7). Median baseline SpO2, measured only during regular breathing, was 99.4% (range, 90.7 to 100; 5th percentile, 95.5). Ten recordings (18%) contained a total of 83 episodes of desaturation (defined as a fall in SpO2 to < or = 80% for > or = 4 seconds). The 95th percentile for desaturation frequency was eight per recording. One infant had 55 episodes of desaturation and thus accounted for two thirds of all episodes observed. Only one of the episodes of desaturation in this infant, and none of those in the other nine infants, had been noted clinically, nor had the abnormally low baseline SpO2 (90.7%) in one infant. Baseline SpO2 in these nondistressed preterm neonates was higher than might be expected, given the SpO2 levels currently recommended for preterm infants with respiratory failure. A minority of infants, however, had a low baseline SpO2 or a high frequency of episodes of desaturation, the potential effects of which remain to be determined.

Cardiorespiratory Patterns in Infants Presenting with Apparent Life-Threatening Episodes

A total of 305 infants presenting with apparent life-threatening events (ALE) were referred by their paediatricians and underwent 24-hour tape recordings of electrocardiogram and abdominal breathing movements (from a pressure capsule transducer). Seventy-seven of these infants, all full-term (greater than or equal to 37 weeks of gestation), were randomly selected, followed up for clinical outcome, and their recordings subjected to a detailed analysis of heart and respiratory rates and breathing patterns. Recordings on 157 age-matched, full-term controls were similarly analysed for comparison purposes. One of the 77 patients suffering from ALE had a pre-existing neurodevelopmental problem, and 4 more cases showed this at follow-up, including 1 case whose ALE was subsequently diagnosed as originating from non-accidental injury. Compared with controls and as a group, the patients suffering from ALE showed higher numbers of apnoeic pauses (p less than 0.001), larger quantities of periodic breathing (p less than 0.01) and lower respiratory rates during regular breathing (p less than 0.01).

Pre- and perinatal clinical characteristics of infants who suffer sudden infant death syndrome.

In a population-based study, clinical and physiological data were collected in the neonatal period on 7,496 full-term (greater than or equal to 37 weeks) infants of birthweights greater than 2.50 kg born in 3 hospitals. The sample excluded a small number of infants who were born with severe illnesses from which there was no period of recovery before death in the neonatal period. Twenty-seven of the infants had died by the time all of the population had passed their third birthday. Twenty-one of these cases died suddenly and unexpectedly, and in 13 there was no adequate explanation for death (sudden infant death syndrome-SIDS). Pre- and early postnatal clinical data on the infants who had died were compared to a randomly selected sample of 478 survivors. This analysis showed that the infants who suffered SIDS showed similar clinical characteristics to those previously reported. Unlike the SIDS group, however, the prenatal histories of the non-SIDS deaths were similar to controls, showing no evidence of a sub-optimal intra-uterine environment.

THE RELATIONSHIP BETWEEN EPISODES OF BRADYCARDIA, APNOEA AND HYPOXAEMIA IN PRETERM INFANTS

Echocardiography assessment of L-R ductal shunting has relied on indirect estimates of flow ie. left atrial, and left ventricular to aortic ratios (LA:Ao, LVEDD:Ao). Since left ventricular output (LVO) increases with L-R ductal shunting it could provide an alternative means of assessment. Measurement by echocardiography is innaccurate due to error in measurement of aortic root cross-sectional area (AoCSA). Doppler measurement of the aortic flow integral or ‘stroke distance’ (AoSD) is highly reproducible. LVO -AoCSA × AoSD × heart rate, therefore, AoSD & minute distance (MinDis=AoSD × heart rate) were evaluated as alternative methods, excluding AoCSA. Studies before and after ductal constriction or closure were compared in babies <32 weeks gestation. 48 courses of indomethacin (indo) were needed in 39 babies. 8 babies, in whom indo failed, were assessed prior to ligation. Results were also compared with 107 scans from 53 babies, aged <28 days. & <32 weeks with a closed duct*.Conclusion: Aortic stroke & minute distance can be used to assess ductal shunting & can be usefully combined with other methods. Significant shunting is likely when AoSD >12cm or MinDis > 1900cm.Little is known about the pathogenesis of bradycardias in preterm infants. We therefore studied episodes of bradycardia and their relationship with apnoeic pauses and desaturations in overnight tape recordings of arterial oxygen saturation (SaO2, Nellcor N100 in beat to beat mode), breathing movements, nasal airflow and ECG from 80 preterm infants at the time of discharge from hospital (mean gestational age at birth 32.5 wk (SD 2.6 wk), and at the time of study 36.3 wk (SD 2.3 wk). A bradycardia was defined as a fall in heart rate to ≤67% of baseline level for ≥4 s, an apnoeic pause as a cessation of respiration for ≥4 s, and a desaturation as a fall in SaO2 to ≤80%. 196 bradycardias were found in 46 (58%) of the recordings.There was a close relationship between bradycardias, apnoeic pauses and desaturations: excluding artefact, 83% of the bradycardias were associated with apnoeic pauses, and 87% with desaturations. Where all 3 phenomena occurred in combination (n=117), they commenced almost simultaneously, apnoeic pauses preceded the onset of the bradycardias by a median of +4.6 s (range -4.0 to +11.0 s); desaturations commenced +5.4 s (-2.2 to +12.7) before the bradycardias and +0.5 s (-5.1 to +5.8) before the apnoeic pauses (after correction of the SaO2 values for the response time of the pulse oximeter in each infant (median 6.0 s)).These observations suggest that a reflex mechanism may be involved in the pathogenesis of some bradycardias in these preterm infants. This reflex can involve an inhibition of respiratory movements, a change in ventilation to perfusion matching (resulting in a fall in SaO2), and a slowing of the heart rate.

CARDIORESPIRATORY PATTERNS OCCURRING IN INFANTS DURING AND AFTER RECOVERY FROM RESPIRATORY TRACT INFECTION

Twelve cardiorespiratory measurements were taken from 24 hour recordings of ECG and respiratory activity in 29 infants during and after respiratory tract infections that involved admission to hospital. These measurements were compared to similar data obtained from 117 age matched control infants without infection. Respiratory and heart rate levels during the state of regular breathing were elevated during infection, compared with recordings made after recovery. The numbers of short apnoeic pauses 3.6 - 6.0 seconds and >6.0 to 12.0 seconds in duration, together with the duration of the overall longest apnoeic pause per recording, were reduced during infection compared with recovery. The total durations of periodic breathing and of periodic apnoea per recording also appeared to be reduced during infection compared with recovery. These effects were consistent in 27 of the 29 cases but,in two, periodic breathing levels during infection exceeded the 90 centile in age matched controls and were reduced on recovery. Measurements made after recovery tended to conform more closely to values in the control infants. None of the 29 infants studied subsequently died or suffered chronic respiratory problems. This study suggests that prolonged apnoeic pauses or increased numbers of short pauses are not usually a consequence of respiratory tract infections in normal infants.