Article Text

Download PDFPDF

Dose-dependent relationship between acidosis at birth and likelihood of death or cerebral palsy
  1. Rod Kelly1,2,
  2. SM Ramaiah3,
  3. Helen Sheridan4,
  4. Hilary Cruickshank5,
  5. Magda Rudnicka2,
  6. Chris Kissack2,
  7. Julie-Clare Becher2,
  8. Ben J Stenson2
  1. 1 Scottish Specialist Transport and Retrieval Service, Royal Infirmary of Edinburgh, Edinburgh, UK
  2. 2 NHS Lothian–Neonatology, Royal Infirmary of Edinburgh, Edinburgh, UK
  3. 3 The Newcastle upon Tyne Hospitals NHS Foundation Trust, Royal Victoria Infirmary, Neonatal Intensive Care Unit, Newcastle upon Tyne, UK
  4. 4 NHS Lothian–Paediatrics, Royal Hospital for Sick Children, Edinburgh, UK
  5. 5 NHS Lothian–Neonatal Physiotherapy, Royal Infirmary of Edinburgh, Edinburgh, UK
  1. Correspondence to Dr Rod Kelly, NHS Lothian–Neonatology, Royal Infirmary of Edinburgh, Scottish Specialist Transport and Retrieval Service, Neonatal Transport, Royal Infirmary of Edinburgh, Edinburgh, EH164SA, UK; rod.kelly{at}nhs.net

Abstract

Background The acid-base status of infants around birth can provide information about their past, current and future condition. Although umbilical cord blood pH <7.0 or base deficit ≥12 mmol/L is associated with increased risk of adverse outcome, there is uncertainty about the prognostic value of degree of acidosis as previous studies have used different variables, thresholds, outcomes and populations.

Methods Retrospective review of routinely collected clinical data in all live-born inborn infants of 35 weeks gestation or more delivered between January 2005 and December 2013 at the Simpson Centre for Reproductive Health, Edinburgh, UK. Infants were included if their lowest recorded pH was <7 and/or highest base deficit ≥12 mmol/L on either umbilical cord blood and/or neonatal blood gas within 1 hour of birth. Neurodevelopmental outcome of the infants with encephalopathy was collected from the targeted follow-up database.

Results 56 574 infants were eligible. 506 infants (0.9%) met inclusion criteria. Poor condition at birth and all adverse outcomes increased with worsening acidosis. Combined outcome of death or cerebral palsy was 3%, 10% and 40% at lowest pH of 6.9–6.99, 6.8–6.89 and <6.8, respectively, and 8%, 14% and 59% at a base deficit of 12–15.9, 16–19.9 and 20 mmol/L or more, respectively.

Conclusions There is a dose-dependent relationship between the degree of acidosis within an hour of delivery, and the likelihood of adverse neonatal and later neurodevelopmental outcome in infants born at 35 weeks gestation or more.

  • neonatology
  • neurodevelopment
  • acidosis
  • perinatal litigation
  • birth asphyxia

Statistics from Altmetric.com

What is already known on this topic?

  • Degree of acidosis at delivery is used as an outcome measure in obstetric clinical trials and as a factor used to support medicolegal claims.

  • Previous smaller studies have demonstrated a variable association between acidosis and adverse outcome of uncertain prognostic value.

What this study adds?

In more than 500 infants with severe acidosis at birth drawn from a population of 59 000 infants, pH and base deficit showed a clear dose-dependent relationship with adverse neonatal and 2-year outcome.

Introduction

Perinatal asphyxia is an important cause of neonatal morbidity and mortality. Umbilical cord blood pH <7.0 or base deficit ≥12 mmol/L are thresholds beyond which the risk of adverse outcome is considered to be increased.1 Most infants whose measurements reach these thresholds recover without sequelae, but some develop neonatal encephalopathy. The relationship between umbilical cord blood pH or base deficit at birth and the immediate condition of the infant and risk of subsequent encephalopathy is not well characterised. In perinatal litigation, considerable importance can be attached retrospectively to these measurements in infants who die or develop permanent disability, with arguments about the likely impact of earlier delivery on the outcome based on weak scientific literature.

The aim of this large population-based study was to describe relationship between pH and base deficit around the time of birth, and the immediate clinical condition of the infants after birth, the risk of neonatal encephalopathy and the risk of death or survival with permanent disability.

Methods

The study took place in the Simpson Centre for Reproductive Health at the Royal Infirmary of Edinburgh, UK. This is a regional referral centre with 7000–8000 births per annum. Data were sought for all live-born infants of 35 weeks gestation or more who were delivered between January 2005 and December 2013. It was routine practice for umbilical arterial and venous blood sampling to be attempted after all deliveries. Umbilical cord blood gases occasionally do not reflect the true acid-base status of a newborn infant (eg, in cases where there has been umbilical cord obstruction). Therefore, both umbilical cord blood gas and early neonatal blood gas results were collected to capture the blood gas values when infants were at their most acidotic. Hospital policy is to aim to provide routine postnatal ward care alongside the mother to all infants who are born at 35 weeks gestation or more. Where an infant was admitted to the Neonatal Unit, blood gas data were collected from the admission blood gas sample, if it was obtained within 1 hour of birth. Infants were included in the study if the lowest recorded pH value was <7.0 or the base deficit value was ≥12 mmol/L either on an umbilical cord blood specimen or on a blood gas specimen obtained from the infant within an hour of birth.

Umbilical cord blood samples and neonatal unit admission blood gas samples were analysed at the point of care using Radiometer ABL805M Flex blood gas analysers in the delivery suite and neonatal unit. Blood gas results were entered into the clinical records. Individual infants could have umbilical arterial and venous blood gases and a neonatal unit admission blood gas (capillary, venous or arterial) obtained within the first hour after birth. Most infants therefore had more than one blood gas sample and for each infant the lowest pH and the most acidotic base deficit from any of these specimens were included. Results were excluded where a transcription error had resulted in an impossible pH or base deficit result being recorded such as with values with more digits than are reported by the analyser or a result outside the clinically observable range. We did not differentiate between umbilical arterial and venous specimens, simply taking the most acidotic specimen for each infant.

The clinical details of the infants were extracted from the electronic patient record systems (Clevermed Badger and TRAK). Data recorded for each infant included pH, base deficit, highest blood lactate level within the first hour, lowest blood glucose level in the first hour, Apgar scores at 1 and 5 min, need for Intermittent Positive Pressure Ventilation by mask or endotracheal tube during initial stabilisation, need for cardiac massage during initial stabilisation, neonatal encephalopathy as defined by Sarnat and Sarnat2 and mortality (both in the labour suite and within 2 years of birth). Neurodevelopmental follow-up until 2 years of age was offered to all surviving infants diagnosed with grade 2 or 3 hypoxic-ischaemic encephalopathy (HIE), as these infants are at highest risk of neurodevelopmental problems. Follow-up was not organised for infants with perinatal acidosis who did not develop encephalopathy. For infants diagnosed with cerebral palsy, a Gross Motor Function Classification System (GMFCS)3 score of 1 was classed as ‘mild’, 2 as ‘moderate’ and 3–5 was classed as ‘severe’. Development was classified as normal in infants in whom there was no concern about developmental delay and delayed in any infant with ongoing concerns.

Advice from the South East Scotland Research Ethics Service indicated that formal NHS ethical review under the terms of the Governance Arrangements for Research Ethics Committees was not needed to review routinely collected data in the neonatal records.

Results

There were 59 092 live inborn infants during the study period, of which 56 574 were born at 35 weeks gestation or more; 4776 eligible infants (8.4%) were admitted to the neonatal unit during the study period. Overall, there were 506 infants (0.9%) who met the inclusion criteria.

pH

This included 504 infants with a lowest pH <7.0 of whom 468 had their lowest pH on umbilical cord blood and 37 had their lowest pH on blood gas specimen within an hour of birth. Two further infants had a pH above 7 but met inclusion criteria on base deficit.

Base deficit

There were 223 of the 506 infants who had a base deficit ≥12 mmol/L. Of these, 151 infants had their highest base deficit on an umbilical cord blood specimen and 72 infants had their highest base deficit on a blood gas specimen within an hour of birth. Of the 504 infants with pH <7.0, there were 199 infants whose base deficit was not 12 or greater and there were 85 infants who did not have a valid base deficit value recorded.

A valid pH was recorded in 66.9% of infants. Base deficit was poorly documented in the electronic patient data record. A valid base deficit was recorded in 5.1% of eligible infants. A flow chart of data acquisition is shown in figure 1.

Figure 1

Flow chart of data acquisition from eligible infants during the study period.

Neurodevelopmental outcome

Of the 506 eligible infants, 3 (0.6%) died in the delivery suite, 271 (53.6%) were admitted to the neonatal unit and 63 (12.4%) developed HIE grade 2 or 3. There were 24 deaths of infants overall (three deaths in the delivery suite and 21 in infants with a diagnosis of HIE grade 2 or 3). There were 42 surviving infants who developed grade 2 or 3 HIE (29 grade 2 and 13 grade 3). Information about developmental delay on follow-up was available for 38 out of 41 cases (93%). This was to at least 2 years for 33 out of 41, and to at least 21 months for 5 out of 41. One infant (who had grade 2 HIE) had extensive cystic encephalomalacia, microcephaly, cerebral palsy and developmental delay when follow-up was taken over in another service at 5 months. A further two infants were lost to follow-up (1 had grade 2 HIE and 1 had grade 3 HIE). There were 15 cases of cerebral palsy in the 39 infants with grade 2 or 3 HIE with follow-up data. All of the infants with cerebral palsy were also considered to have delayed development. Infants are categorised according to degree of acidosis as shown in table 1, which details Apgar scores, resuscitation required and early mortality. For those admitted to the neonatal unit, table 2 shows highest lactate and lowest glucose within 1 hour, highest grade of HIE during admission, cerebral palsy and mortality. Long-term outcomes for those diagnosed with HIE grade 2 or 3 is shown in table 3. Outcomes compared with degree of acidosis are shown in figures 2 and 3.

Figure 2

Condition at birth according to acid-base status at birth. IPPV, intermittent positive pressure ventilation; NNU, neonatal unit.

Figure 3

Outcome at 2 years according to acid-base status at birth. HIE, hypoxic-ischaemic encephalopathy.

Table 1

Condition at birth categorised by pH and base deficit

Table 2

Neonatal outcomes for those infants admitted to the neonatal unit and neurodevelopmental outcomes for infants undergoing targeted follow-up

Table 3

Long-term outcomes for those infants who developed HIE grade 2 or 3

Worsening acidosis as measured by either pH or base deficit was associated with poor condition at birth and the development of HIE, and at 2 years with death or cerebral palsy. The risk of death or cerebral palsy was 3%, 10% and 40% at a lowest pH of 6.9–6.99, 6.8–6.89 and <6.8, respectively, and 8%, 14% and 59% at a base deficit of 12–15.9, 16–19.9 and 20 mmol/L or more, respectively. When data are restricted to 2009 onwards after the introduction of therapeutic hypothermia for HIE grade 2 or 3, the findings were very similar with risk of death or cerebral palsy 3%, 9% and 35% at lowest pH of 6.9–6.99, 6.8–6.89 and <6.8, respectively, and 9%, 15% and 64% at a base deficit of 12–15.9, 16–19.9 and 20 mmol/L or more, respectively. Long-term outcomes of infants before and after the introduction of therapeutic hypothermia are shown in table 4.

Table 4

Long-term outcomes for those infants who developed HIE grade 2 or 3, pre and post the introduction of therapeutic hypothermia

Discussion

This population study of infants born at 35 weeks gestational age or more demonstrates a dose-dependent association between the degree of acidosis around the time of birth (measured by the most acidotic result from either umbilical cord blood gas or subsequent blood gas within 1 hour of birth) and adverse neonatal and later neurodevelopmental outcome. It is the largest single cohort, which examines both routine umbilical samples and subsequent gases within 1 hour of birth, which incorporates venous pH, arterial pH and base deficit, and which includes neurodevelopmental follow-up of 93% of encephalopathy survivors.

Umbilical cord blood gas analysis is recommended in all high-risk deliveries4 and where a baby is born in poor condition,5 and in some centres this practice is routine following all deliveries. Early identification of infants at risk of complications such as encephalopathy is important as prompt intervention can be instituted. However, there is not clear consensus on how to define fetal acidosis and there is a lack of robust umbilical cord acid-base reference values.6

Previous observational studies have drawn inconsistent conclusions with regard to the association between cord pH and adverse outcome. These studies have used a variety of different thresholds to define significant acidosis (pH 7.0–7.24),7 have used different variables such as arterial cord pH, venous cord pH and base deficit,8–10 have examined a variety of different neonatal and long-term outcomes11–22 and have used both selected high-risk populations23–30 and unselected populations where umbilical sampling is routine.31–39 This has led to uncertainty about the prognostic value of the degree of acidosis at birth, despite cord pH being used as an outcome me(9-29%) asure in obstetric clinical trials and as a factor that may be used to support medicolegal claims.1 4

Malin et al 7 conducted a systematic review and meta-analysis into the association between umbilical cord pH and both perinatal and long-term outcomes. They demonstrated an association between low arterial cord pH and neonatal mortality, HIE, intraventricular haemorrhage, periventricular leucomalacia and cerebral palsy. However, both the variable and the threshold used to define significant acidosis differed significantly between papers, and the population varied including both term and preterm infants.

Yeh et al 40 conducted a large observational cohort study, including the entire pH range. They described an increased risk of adverse neurological outcome at a pH <7, and demonstrated that above a pH of 7 neonatal acidaemia is weakly associated with adverse outcome. Base deficit, or subsequent infant blood gases, were not analysed and the rate of acidotic infants was significantly higher (2.2%) than in our study, presumably secondary to targeted umbilical cord gas sampling.

The literature which relates outcome to a specific level of base deficit in term infants is relatively sparse. One study of cord base deficit from the 1990s in which term and preterm infants were matched with a more historic cohort, showed an increased risk of neonatal complications at a base deficit of >16 mmol/L.41 A further study describes the neonatal outcomes of term infants born in the same time period in relation to various base deficit values taken in the first hour of life. Although this was a highly selected population and no mortality is described, it shows an increasing likelihood of encephalopathy with increasing acidosis, with a threshold base deficit of 14 mmol/L giving 81% sensitivity for developing encephalopathy.42 Both studies showed a higher incidence of encephalopathy than our more recent cohort perhaps reflecting less rigorous postnatal support such as effective resuscitation protocols and the absence of therapeutic hypothermia, the latter well-recognised in altering the typical trajectory of HIE.43

In relation to long-term sequelae, only one study reports the outcome at differing buffer base values.44 Like base deficit, buffer base measures the metabolic acid state of the fetus but from an opposite perspective (presence of base vs absence of base). This study showed a dose-dependent relationship between metabolic acidosis and both major and minor deficits at 1 year of age but the study was limited in being small, including both term and preterm infants, and in not describing infants who died.

The strength of this paper is that it includes a large cohort selected only by gestational age and therefore captures both high-risk and low-risk deliveries. Unlike previous studies, it also includes blood gas samples obtained within an hour of birth and therefore includes infants whose umbilical cord values may not truly reflect their degree of acidosis at the time of delivery (eg, cases of cord obstruction). This study also examined both pH and base deficit, and the centre involved has a system of targeted follow-up which provides neurodevelopmental follow-up until 2 years of age in infants who have survived neonatal encephalopathy.

A limitation of this study is that while pH is well-recorded, base deficit data are less complete in the maternal electronic patient record. There is therefore the possibility that infants who had a high cord base deficit alone were missed from inclusion. All cases of HIE were captured from the neonatal unit admission data, but because valid cord blood gas data were only available for 66.9%, it is likely that there were other acidotic infants who did not have cord gases and remained well. This means the extent to which acidosis may predict adverse outcome may be overestimated by this dataset. In this paper, we have only estimated the risk associated with the degree of acidosis. Other factors which might influence outcome such as hypoglycaemia,45 intrauterine growth restriction46 and chorioamnionitis47 were not studied.

The risk of adverse long-term outcomes associated with acidosis was broadly similar in this cohort of infants before and after the introduction of therapeutic hypothermia. The study design does not measure the efficacy of therapeutic hypothermia.

This study defines a clear dose-dependent relationship between the degree of acidosis and outcome in a cooling era. It may help to facilitate the setting of entry criteria in clinical trials and better inform medicolegal opinion as to the likely outcome had delivery been expedited and the severity of acidosis been reduced.

Appendix

, . . ;:–.

References

Footnotes

  • Contributors SMR and HS conducted an initial internal audit using data from 2005 to 2009 from unit admission records and patient notes. RK expanded this to a longer study period, and included data from electronic patient records (provided by CK), blood gas analysers and follow-up data (provided by HC and MR). RK wrote the manuscript under the supervision of J-CB and BJS, who both conceived and refined the study design.

  • Competing interests None declared.

  • Ethics approval South East Scotland Research Ethics Service.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.