Acta Paediatr 81: 536-41. 1992
Hypoxaemia in infants with respiratory tract infections C F Poets', VA Stebbens', JR Alexander', WA Arrowsmith2, SAW Salfield3and D P Southall' Academic Department of Paediatrics. North Staffordshire Hospital Centre. University of Keele', Doncaster Royal Infirmar?. Doncaster and Rotherham District General Hospital'. Rotherham. U K
Poets CF, StebbensVA, Alexander JR, Arrowsmith WA, Salfield SAW, Southall DP. Hypoxaemia in infants with respiratory tract infections. Acta Piediatr 1992;81536-41. Stockholm. ISSN 0803-5253 Nineteen infants who were graduates from special care baby units underwent two overnight tape recordings of oxygen saturation (Sa02) and breathing movements; one during an upper (n= 12) or lower (n = 7) respiratory tract infection and the other when free of infection. Baseline Sa02was lower during infection (median 99.6 vs 1OQ%, p < 0.01), with four patients having values (84.3-95.5%) below the normal lower limit for full-term infants (97%). The median number of apnoeic pauses was also lower during respiratory tract infection (4.7 vs 15.7/h, p 0.05), with the exception of one patient who had extremely increased values during infection for both apnoeic pauses (63/h) and desaturations ( I 12/h). No infant, however, was considered clinically hypoxaemic. Clinically unsuspected hypoxaemia may thus occur during respiratory tract infection in a proportion of infants graduating from special care baby units. Such hypoxaemia may have potentially deleteriouseffects. 0 Arterial oxygen saturation, hypoxaemia. respiratory tract infection DP Southall. Academic Department of Paediatrics, North Stafforbhire Hospital Centre, Stoke on Trent, UK
Based on clinical history and histology there is evidence of a recent respiratory tract infection (RTI) in between 40 and 80% of victims of sudden infant death syndrome (1, 2). However, since almost every infant will contract at least one RTI (3), it remains unclear why an infection should trigger sudden and unexpected death in a minority. In 1980, Guntheroth et al. showed that a combination of RTI and hypoxaemia was accompanied by the development of intrathoracic petechiae, the latter being a consistent postmortem finding in sudden infant death syndrome (4). One possible cause for hypoxaemia is a prolonged apnoeic pause. In infants suffering RTI, however, Gould et al. (9,and Southall et al. (6) found that the number of apnoeic pauses was reduced during infection, and that prolonged pauses were also less frequent (9,or absent (6), during RTI. In contrast, other investigators reported more short (7), or more prolonged (8), apnoeic pauses during RTI. Clearly recordings of breathing movements and/or airflow do not provide direct information for the investigation of a potential relationship between RTI and hypoxaemia. In the present study, overnight tape recordings of breathing movements and arterial oxygen saturation (SaO2) were performed using a pulse oximeter in the beat-to-beat mode which has been validated against arterial line measurements (9). These were made on a group of infants who were graduates from special care baby units, most of whom had been born preterm. Preterm birth is associated with an increased risk of both hypoxaemia (10) and sudden death (1 1). Therefore,
these infants represent a group which may be particularly susceptible to the effects of RTI on oxygenation. Two specific questions are addressed in this report. First, what is the influence of a RTI on baseline oxygen saturation, and second, how does infection modify the incidence of apnoeic pauses and of short-lived episodes in which arterial oxygen saturation falls?
Patients and methods We have recently reported data on Sa02 and breathing patterns in preterm infants at discharge from special care (10). These data were from a population-based study on all infants admitted to three special care baby units immediately after birth during a one-year period. Infants were studied at discharge and about six weeks later. For the present study, parents of the above infants and hospital staff were asked to report any RTI occurring in the first six months after discharge. Infection was reported in 19 infants. Parental consent and hospital Ethics Committee approval were obtained to perform recordings during these illnesses. Table 1 summarizes the clinical data of these infants. Their median age at the time of the RTI recording was 76 days (range 23-181 days), their gestational age at birth was 32 weeks (26-41 weeks) and their birth weight was 1800 g (825-3780 g). Sixteen infants had been born preterm (gestational age < 37 weeks). Reasons for admission to special care were respiratory distress syndrome in 13, birth asphyxia in three and neonatal
Hypoxaemia in respiratory tract infections
ACTA PKDIATR 81 (1992)
Table I . Clinical histories and infection data in the 19 patients.
Ageat Ageat Gestational Birth infection control recording recording Subject Pathogen In hospital age at birth weight no. Nature of infection" isolated (Yes/No) (weeks) (g) Sex (days) (days) I 2 3 4 5 6 7 8 9 10
I1 12 13 14 I5 16 17 18 19 a
URTI URTI URTI URTI URTI URTI URTI URTI URTI URTI URTI URTI Bronchiolitis Bronchiolitis Bronchiolitis Bronchiolitis Bronchiolitis Bronchiolitis Bronchiolitis
RSV RSV RSV RSV RSV RSV RSV
Yes Yes No No No No Yes Yes No No No Yes Yes Yes Yes Yes Yes Yes Yes
31 34 39 32 36 28 32 31 33 36 26 33 30 34 37 41 31 31 29
1545 2350 3250 1825 1800 1000
1650 1750 2100 2380 825 2260 1500 2100 2700 3780 1425 I250 850
F F M M M M F M F M M F F M M F F F F
54 23 47 100
49 86 36 44 58 60 130 81 76 36 137 98 180 181 163
98 61 91 17 20 48 21 37 20 18 102 72 97 49 17 62 I13 113 141
Treatment given for infection (excluding antipyretic;) Antibiotics Antibiotics Nasal decongestant Nil Nil Antibiotics Nil Nil Nil Nil Nil Nil Antibiotics Antibiotics Ipratropium bromide, ketotifen Antibiotics Antibiotics Antibiotics Antibiotics
URTI =Upper respiratory tract infection; RSV = respiratory syncytial virus.
anaemia, meconium aspiration and hypoglycaemia in the remaining three infants. No infant showed congenital anomalies and only one had required additional inspired oxygen for more than the first 28 days of life (case 19, for 72 days). At the time of the recording, 12 patients had an upper RTI and seven had bronchiolitis. Recordings were performed within four days of the onset of the first symptoms in all but one of the infants with upper RTIs, and shortly after the acute phase of the disease (between day 4 and day 9) in the patients with bronchiolitis. Thirteen patients had been admitted to hospital because of their RTI. In the remaining six patients recordings were performed at home, although one patient (case 10) was admitted to hospital the day after his recording. The diagnosis of RTI was made on clinical findings and, in patients admitted to hospital, from chest X-rays. All patients had a pharyngeal and/or nasal specimen taken for virus isolation and immunofluorescence. No patient was ventilated during the RTI and only one patient (case 14) received additional inspired oxygen (during the first two days of hospital admission). All infants survived. As part of the normative study (lo), all patients had at least one recording performed at a time when they were free of infection. The recording performed most closely to the infection recording was taken for comparison (control recording). Fourteen patients had this recording prior to, and five following, the infection recording, with a difference between median age at infection and at control recording of 15 days. In 15 infants (79%) the time interval between infection and control recording was less than 45 days.
All infants underwent 12-h overnight tape recordings (Racal FM4) of S a 0 2(Nellcor NlOO with new software equivalent to N200 and specially modified to provide beat-to-beat data), each photoplethysmographic waveform from the oximeter (for the validation of the saturation signal) and breathing movements from a volume expansion capsule (Graseby) or from respiratory inductance plethysmography (Studley Data Systems) (12). Nasal airflow was also recorded but not analysed in this study. Recordings were printed and analysed by two experienced workers who were blind as to whether they were made during the RTI or when well. The duration of artefact-free SaOz signal, as observed from examination of the photoplethysmographic waveforms, was measured. Periods of regular and non-regular breathing movement patterns were defined as described previously (1 3). Apnoeic pauses were identified wherever there was a pause in abdominal wall breathing movements of 2 4 s in duration; these were counted. Pauses were classified into three groups by duration: 4.0-7.9, 8.0-1 1.9 and 2 12 s (12). Episodes with at least three successive apnoeic pauses, each separated by less than 20 breaths, were classified as periodic apnoea (1 3). Baseline S a 0 2was calculated by measuring the Sa02 values a t five consecutive breaths at the centre of each episode of regular breathing; these breaths being at least 10 s away from sighs and apnoeic pauses (12). Episodes in which Sa02 fell to 5 80% were counted, and the time SaO2 remained 580% measured for each episode. These desaturations were classified according to their relationship with an apnoeic pause. A positive association was noted if the desaturation commenced within
CF Poets et al.
ACTA PEDIATR 81 (1992)
Table 2. Medians (ranges) for variables measured in RTI and control recordings. Infection Age at study (days) Baseline Sa02 Apnoeic pauses/h Periodic apnoeic pauses/h %, apnoeic pauses 4.0-7.9 s '%,apnoeic pauses 8.0- I I .9 s "AI apnoeic pauses 2 I2 s Desaturations (SaOz sSO%,)/h Mean duration of desaturations(s) Longest duration of desaturations(s) "A, of pauses with desaturation YOof desaturations with apnoeic pause
76 (23- I8 1) 61 (17-141) 99.6 (84.3-100) 100 (98.4-loo)** 4.7 (0.1-62.6) 15.7 (2.9-50.9)* 0.7 (0.0-47.7) 5.0 (0.0-26.3) 95.7 (72.5-100) 94.4 (80.6-100) 5.6 (0.0-19.4) 4.4 (0.0-26.6) 0.0 (0.0-9.2) 0.0 (0.0-2.2) 1.3(0.0-111.8) I .9 (0.0-24.4) 1.2 (0.5-2.5) 1.5 (0.3-3.5) 3.4 (0.3-3 1.4)
14.7 (0.0-83.0) 47.4 (0.0-100)
9.8 (0.0-43.1) 76.4 (0.0-loo)**