Discussion
This is the first study since the original study in the 60’s, that defines cut-offs from neonatal outcomes, without referring and comparing to previously published cut-offs (5). Based on our results we suggest the interventional cut-off for fetal acidaemia to be 5.2mmol/L, when analysing the scalp blood lactate with StatstripLactate® or StatstripXpress®. The argument for the chosen cut-off of 5.2mmol/L is the balance between obtaining the best sensitivity and specificity for common recognized prognostic factors of intrapartum asphyxia. By increasing the cut-off to more than 5.2mmol/L the sensitivity will decrease for AS <7 at 5 minutes and for the composite outcome. On the other hand, by choosing a lower cut-off than 5.2mmol/L the intervention rate will be unnecessarily high. We have previously shown that the mean value for LP + 2SD equals 5.2mmol/L in the second stage, although it is not known whether that cut-off is associated with an increased risk for adverse outcomes. It has been previously shown that the risk for disability at 4 years increases with increasing scalp blood lactate levels (20,21).
Lactate accumulates in tissues, blood and cerebral spinal fluid as a result of anaerobic metabolism due to hypoxia. Lactate levels, which are normally within a narrow range, can therefore be used to monitor tissue hypoxia since increased levels are associated with a high risk of compromised vital organ function (22,23). In the 90s, after valuable work with the POC measurement of lactate in scalp blood, Swedish scientists published cut-offs for normality and acidaemia for LP (6–8). These are the cut-offs recognized and used worldwide today (1,2,6). Drawbacks of that study are that the time intervals from FBS to delivery are not mentioned and that the cut-offs are partly based on the pre-existing cut-offs for pH (5,6).
Recently, it was shown that the scalp blood lactate level is closely associated to the level in central circulating blood (24). Not surprisingly, the correlation between FBS-lactate and cord blood lactate improves as the sampling-to-delivery time interval shortens (25). It is important to emphasize that the FBS-lactate level reflects the metabolic status of the fetus at the moment of FBS and depending on the fetal reserve and other factors the metabolic status can rapidly deteriorate. Consequently, comparing the FBS-lactate value to the fetal outcome will always be biased by the time-lag from FBS to delivery. To avoid preanalytical errors, it is important to ensure correct sampling technique and if there is any doubt about contamination with amniotic fluid or fetal products, sampling and analysing must be repeated directly (26).
For this study, we chose 25 minutes from FBS to delivery because it showed the highest correlation between SSLX and cord blood values and based on the reasoning above. In addition, 25 minutes is appropriate given that repeat sampling is recommended within 20-30 minutes if the CTG is still non-reassuring and it is also an expected interval from recognizing a high lactate value until expediated delivery, either by caesarean section or vacuum extraction (16,17).
There is no ideal outcome variable specifically indicating intrapartum asphyxia. Analysing cord blood provides the clinician with the most objective and accurate measurement of the metabolic status of the new-born, although the majority of babies with deteriorated cord blood gases will be vigorous and manifest no obvious short- or long-term neurological sequelae (27,28). Analysis of cord blood requires meticulous technique and knowledge of pitfalls such as timing of sampling/analysis and the different equations in the setting of the BGAs (2,29–31). There is no international consensus of the definition of fetal MA in cord blood, which compared to respiratory acidosis is a serious threat to the cell function (32–35). The ACOG definition of metabolic acidosis is pH < 7.0 + BDblood> 12 mmol/L, whereas the FIGO guideline uses a threshold of pH < 7.05 + BDecf >10mmol/L or lactate > 10mmol/L, because an association with adverse neonatal outcome is recognized at that level (18,22,36). Note that the value of BD is significantly dependent on the compartment used for calculation - blood or extracellular fluid - with highest values when calculated in blood (31). The lactate value in cord blood correlates with the lactate concentration in the fetal brain, which in turn is an established marker for the severity of cell damage and thereby for the degree of hypoxic ischemic encephalopathy. This supports the use of lactate in the definition of MA (22,23,37).
Low Apgar Scores, for most neonates, are likely to be due to hypoxia whereas an intermediate score often is due to other reasons, such as medications given to the mother, gestational age, or malformations (38). If the one-minute Apgar is low and the baby needs respiratory support or other intervention, then the five- and ten-minute Apgar will be affected by resuscitation procedures undertaken. This makes the AS an unreliable and nonspecific single marker for intrapartum hypoxia. In accordance with Kruger et al ., we found that the predictive ability of SSLX for low Apgar scores was relatively low (6).
In the pursuit of the optimal predictor of intrapartum hypoxia we created a parameter based on a pH < 2SD in a normal population plus one of the following outcomes: admission to neonatal intensive care unit, continuous positive airway pressure therapy, bilevel positive airway pressure, manual ventilation or AS < 7 at 5 minutes. In that context, it is notable that the negative LR were lowest for the three outcomes: MA, pH < 7.05 + BD ≥ 12mmol/L and pH < 7.05 plus lactate > 10mmol/L implying that an intervention cut-off of 5.2mmol/L can safely rule out intrapartum fetal acidaemia.
The high false positive rate of CTG remains a limitation in obstetrics and to achieve an appropriate intervention rate a secondary test is necessary. There is emerging evidence of the negative consequences of operative deliveries for both the mother and the baby (39,40). In our study, of all fetuses with a concerning CTG, only 16% of the FBS lactate were above the recommended cut-off for intervention, thus principally allowing 84% of all cases with a non-reassuring CTG to continue. A randomized controlled study to evaluate the effectiveness of FBS to reduce interventions without an increase in adverse neonatal outcomes is still lacking. The study by Haverkamp et al . was underpowered and the results from the Flamingo trial are not yet published (1,41).
Strengths of this study include the large sample size, which enabled cut-offs to be derived from lactate samples with a wide range and obtained close to birth, the routine of FBS and for umbilical cord blood sampling and use of continuous CTG in the second stage of labour.
The main limitation of the study is that, despite the large number of included patients, the numbers of adverse outcomes were low, resulting in relatively wide confidence intervals regarding the sensitivity. Also, not all cases had a successful blood gas analysis.
In conclusion, we suggest the scalp blood lactate cut-off for intervention to be ≥5.2mmol/L. If the test result is normal, labour can continue, but we recommend repeated FBS after 20 to 25 minutes if the CTG is still non-reassuring. To evaluate the efficacy of FBS, a randomized controlled trial is warranted with the use of the StatstripLactate®/StatstripXpress® Lactate system and the cut-off of 5.2mmol/L.