Outcome measures in perinatal medicine - pH or BE. The thresholds of these parameters in term infants

Hypoxia and severe foetal acidosis may lead to cerebral injuries and multi-organ failure. Base excess (BE) and actual pH determined in umbilical artery (UA) blood are valid parameters to measure (metabolic) acidosis. Until now there is no consensus worldwide as to which of the two parameters should preferably be used and which thresholds should be applied: the thresholds 7.000, 7.100 and 7.200 are discussed for pH,UA and - 16.0 mmol/l for BE,UA, respectively. The aim of this study was to redefine these thresholds for term infants taking into account the entire spectrum of complications in early neonatal life and to compare the diagnostic power of both variables under investigation.

512 foetuses all with a pH,UA < 7.100 were enrolled in this retrospective study. In this paper only term infants (n = 398) without major malformations were analysed. In order to quantify foetal morbidity and mortality an asphyxia complication (AC) score was designed using the Apgar score after 1 min and all possible complications encountered in the neonate until discharge (mainly from the NICU). Routine acid-base (AB) measurements (Radiometer, Copenhagen) were available (UA and UV blood) in nearly all cases. In this context, BE was corrected to 100 % oxygen-saturation (BE (oxy.)) using the (calculated) actual saturation (BE (act.oxy.)) of haemoglobin in each case because sO2 (%) becomes very low in severe acidosis. Oxygen saturation (%) was determined according to Ruiz et al. using haemoglobin F. Matched pairs (pH,UA > 7.10) were defined using the variables (i) gestational age, (ii) birth weight (+/- 100 g), (iii) sex, and (iv) parity.

Analysing term infants with a definitely good outcome (n = 389) led to the following AB variables in UA blood: 10 % had a pH < 7.000; the lowest pH was 6.717, the highest pCO2 was 118 mmHg and the lowest BE(oxy.) amounted to - 32.4 mmol / l. 90 % of these neonates had an oxygen saturation that was still > 3.0 %. However, early neonatal morbidity due to hypoxia and acidosis was remarkable; therefore these AB variables could not serve as thresholds. Relying on clinical criteria, no acidotic morbidity was accepted, except for respiration disorders in early neonatal live (C score) together with Apgar scores down to 4 after one minute. This led to an AC score of < or = 8 (n = 378). In this group the lowest pH,UA was 6.890 and BE(oxy.) was - 25.1 mmol / l, respectively. In order to have a "buffer zone" of 10 % in each variable distribution (10 (th) / 90 (th) percentile). the following thresholds could be evaluated in UA blood: pH: 7.000, pCO2: 84 mmHg, sO2: 3.0 % and BE(oxy.) - 20.0 mmol/l. Only 13 (matched) neonates suffered from an AC score > 8 and had all pH values > or = 7.100 (3.3 % overlap). These foetuses seemed to be not at risk concerning injuries and severe complications. In UA blood the actual pH always offers closer correlations with the AC score when compared with the BE (oxy.) value.

Thresholds in UA blood for pH, pCO2, sO2 and BE(oxy.) in term-infants are: 7.000, 84 mmHg, 3.0 % and - 20 mmol/l, respectively. Delivery of an otherwise healthy baby without getting in touch with these thresholds seems to be safe both for the baby and the obstetrician. In addition, severe neonatal depression (Apgar 1 min: 0 and 1) is usually avoided (0 / 398). BE(oxy.) does not offer a higher diagnostic power when compared with actual pH.