Why Does a Baby's Heart Rate Drop During Labor
Summary
Groundwork
Continuous electronic fetal center-rate monitoring is widely used during labour, and computerised estimation could increase its usefulness. Nosotros aimed to establish whether the addition of determination-support software to assistance in the estimation of cardiotocographs affected the number of poor neonatal outcomes.
Methods
In this unmasked randomised controlled trial, we recruited women in labour aged 16 years or older having continuous electronic fetal monitoring, with a singleton or twin pregnancy, and at 35 weeks' gestation or more than at 24 maternity units in the Great britain and Republic of ireland. They were randomly assigned (one:1) to decision support with the Infant arrangement or no decision support via a estimator-generated stratified block randomisation schedule. The primary outcomes were poor neonatal outcome (intrapartum stillbirth or early on neonatal death excluding lethal congenital anomalies, or neonatal encephalopathy, admission to the neonatal unit inside 24 h for ≥48 h with testify of feeding difficulties, respiratory illness, or encephalopathy with evidence of compromise at birth), and developmental assessment at age ii years in a subset of surviving children. Analyses were done by intention to treat. This trial is completed and is registered with the ISRCTN Registry, number 98680152.
Findings
Between January 6, 2010, and Aug 31, 2013, 47 062 women were randomly assigned (23 515 in the conclusion-support group and 23 547 in the no-decision-back up group) and 46 042 were analysed (22 987 in the decision-support group and 23 055 in the no-decision-support grouping). We noted no difference in the incidence of poor neonatal result between the groups—172 (0·7%) babies in the determination-support group compared with 171 (0·7%) babies in the no-decision-support grouping (adjusted risk ratio ane·01, 95% CI 0·82–i·25). At 2 years, no significant differences were noted in terms of developmental assessment.
Interpretation
Use of computerised interpretation of cardiotocographs in women who accept continuous electronic fetal monitoring in labour does non better clinical outcomes for mothers or babies.
Funding
National Establish for Wellness Research.
Introduction
Continuous electronic fetal heart rate monitoring in labour is widely used but its potential for improving neonatal outcomes has not been realised.
i
- Alfirevic Z
- Devane D
- Gyte GML
Continuous cardiotocography (CTG) as a form of electronic fetal monitoring (EFM) for fetal assessment during labour.
- Google Scholar
The reasons for this are circuitous, only include difficulty interpreting the fetal heart charge per unit trace correctly during labour.
,
,
Computerised interpretation could be used to considerately detect abnormalities in fetal heart rate patterns during labour that are associated with asphyxia and bring them to the attention of birth attendants, who could then take action or expedite delivery and potentially prevent stillbirth or exposure to meaning asphyxia.
Guardian (K2 Medical Systems, Plymouth, Great britain) is an electronic information capture organization for managing information from labour monitoring.
v
K2s Medical Systems
K2MS Guardian: full electronic capture of patient information during childbirth.
- Google Scholar
It displays the cardiotocograph on a figurer screen alongside other clinical information either from external ultrasonographic transducers, or from fetal scalp electrodes (eg, partographs, maternal vital signs, details of maternal anaesthesia and analgesia) collected as part of routine clinical intendance. Guardian does non interpret any of the data gathered, but acts equally an interface to collect and brandish data at the bedside, centrally on the labour ward, in consultants' offices, or remotely.
Babe (K2 Medical Systems) is a decision-support software that was developed to run on the Guardian system. It analyses the quality of fetal heart signals and, if these signals are acceptable, displays baseline heart rate; heart-rate variability; accelerations and type and timing of decelerations; the quality of the signal; and the contraction pattern.
half-dozen
- Keith RDF
- Beckley South
- Garibaldi JM
- Westgate JA
- Ifeachor EC
- Greene KR
A multicentre comparative report of 17 experts and an intelligent figurer system for managing labour using the cardiotocogram.
- Crossref
- PubMed
- Scopus (119)
- Google Scholar
,
7
Skinner JF, Harris M, Greene KR. Computerised decision support for managing labour using the cardiotocogram: 500 cases with the range of abnormality. 28th British Congress of Obstetrics and Gynaecology; Harrogate, UK; June 30–July four, 1998.
- Google Scholar
,
8
- Keith RDF
- Greene KR
Evolution, evaluation and validation of an intelligent decision support tool for the management of labour.
- Google Scholar
INFANT and so makes an assessment of the overall pattern, which, if necessary, volition upshot in a colour-coded alert (blue is the least severe, yellow is moderate severity, and ruddy is the most severe; appendix). The decision-support software does not provide recommendations for any action that should be taken in response to abnormalities—such decisions are at the discretion of attending clinicians. In the United kingdom of great britain and northern ireland'due south National Health Service (NHS), all clinicians supervising labour are expected to have been trained in the appropriate response to an abnormal cardiotocographic reading—eg, past completing calculator-based training packages every 6–12 months, attending annual lectures, or attending regular cardiotocographic review meetings. All the same, training is non standardised, and individual competence is not assessed in virtually units.
Research in context
Evidence before the report
The National Institute for Wellness Research searched the Cochrane library, PubMed, NHS Bear witness, and the DARE database for articles published in the 5 years earlier it commissioned this trial in 2006. No studies of electronic decision support in interpretation of cardiotocography were identified. One previous trial of decision back up in labour has been washed, in which direct measurement of the fetal electrocardiograph—which requires the application of a fetal scalp electrode—was used. That trial recruited 7730 women and showed no evidence of a difference between the groups on the primary consequence of metabolic acidosis equally measured in cord blood.
Added value of this report
This trial, INFANT, was the first to assess the utilise of conclusion back up in the interpretation of the cardiotocographs in women undergoing continuous electronic fetal heart-rate monitoring. The aim was to accost the effect on substantive outcomes of neonatal mortality and morbidity. The size of the Babe trial provides sufficient ability to detect pocket-sized differences in the principal outcome of perinatal mortality and serious morbidity, and outcomes such equally metabolic acidosis. No differences were noted. Additionally, it had very loftier ability to assess whether use of this decision-support system had any consequence on the run a risk of operative delivery, and over again nosotros noted no evidence of whatever deviation.
Implications of all the available prove
The quality of intendance in labour is a major business organisation for women, their families, and the health professionals providing intendance for them. Technologies to improve intrapartum monitoring that identify early signs of fetal hypoxia during labour are needed and then that clinicians can intervene and endeavor to prevent poor neonatal outcomes. Our results prove that the method of determination back up nosotros tested does non achieve this aim.
We did a randomised controlled trial to examination the hypotheses that a substantial proportion of substandard care results from failure to correctly identify abnormal fetal heart rate patterns, that improved recognition of abnormality would reduce substandard care and poor outcomes, and that improved recognition of normality would decrease unnecessary intervention.
Methods
Study pattern and participants
We did a pragmatic, unblinded, randomised controlled trial in motherhood units in the UK and Ireland. All 24 sites in the Great britain and Ireland that used Guardian at the time of the trial took office. Doctors and midwives were able to recruit participants. Eligible women were judged to crave continuous electronic fetal centre charge per unit monitoring by the local clinical team on the basis of their existing do, had a singleton or twin pregnancy, were at 35 weeks' gestation or farther forth, had no known gross fetal abnormality, including whatsoever known fetal eye arrhythmia such every bit heart cake, aged 16 years or older, and able to give consent to participate equally judged by the attending clinicians. Continuous electronic fetal heart rate monitoring during labour is not routine in the Uk. Clinical guidance for the NHS recommends that women assessed every bit having a low take a chance of complications should be offered intermittent auscultation during labour.
10
National Collaborating Centre for Women's and Children'south Health (UK)
- Google Scholar
Significant women attention recruiting hospitals were provided with written information near the trial during pregnancy and in labour. For women who met the eligibility criteria, written informed consent was sought past ways of a dated signature from the woman and from the person who obtained informed consent. Enquiry ideals committee blessing for the written report was granted by the National Research Ethics Service—Northern and Yorkshire Research Ethics Committee (09/H0903/31). The study protocol has been published.
xi
- Brocklehurst P
A study of an intelligent system to support decision making in the management of labour using the cardiotocograph—the Infant report protocol.
- Crossref
- PubMed
- Scopus (nineteen)
- Google Scholar
Randomisation and masking
The Guardian organization was used to confirm that all necessary eligibility criteria were met and then to randomly allocate women (ane:1) to cardiotocography with or without decision support. The allocations were computer generated in Stata (version 10.1) by the trial statistician, who used stratified block randomisation, in which variable block sizes were used to remainder betwixt the two trial arms by whether the pregnancy was a singleton (cake sizes 12, fourteen, 16, 18, 20, 22, and 24, allocated in proportion to the elements of Pascal's triangle—1:half-dozen:15:20:fifteen:half-dozen:1) or twin (block sizes 2, 4, and 6, allocated in proportion to the elements of Pascal'southward triangle 1:2:one), and within each participating centre. The trial was not blinded, which allowed indirect measurement of any changes in clinician behaviour, such equally how much fourth dimension the attending midwife spent with the woman on the basis of the knowledge that the decision-support system was active or not.
Procedures
Clinicians in participating centres were trained in the use of the determination-support software by staff from the trial office. A grooming team at each site was responsible for cascading grooming amidst the local clinicians. Women in the trial were managed according to standard procedures. No additional grooming was provided to clinical staff in how to respond to fetal heart rate abnormalities.
Labour data and outcomes were stored contemporaneously on the Guardian system, which were so sent electronically to the trial office. Data were extracted from notes of babies admitted to the neonatal unit and for all neonatal deaths. All children surviving were flagged at the NHS Information Centre for those born in England and NHS Greater Glasgow and Clyde Safe Oasis for those born in Scotland, which meant that all deaths occurring later discharge in these countries could be identified. Ireland does non have a similar system for monitoring deaths, then Irish gaelic data had to be excluded from the denominator for these calculations. A sample of surviving children were followed upward to age two years via a parent-completed questionnaire to assess the child's health, development, and wellbeing (appendix).
All babies with an adverse outcome potentially associated with intrapartum asphyxia (ie, the trial primary outcome plus cord artery pH <7·05 with base deficit 12 mmol/L or more) and all neonatal deaths and intrapartum stillbirths had their care in labour assessed past review of de-identified case notes by a panel comprising a senior obstetrician, neonatologist, and midwife, to run across if intendance was suboptimal—ie, if it was possible or likely that different management would take prevented the adverse outcome.
,
Outcomes
The trial had two main outcomes. The first was a binary (present or not present) blended of poor neonatal outcome, including deaths (intrapartum stillbirths plus neonatal deaths up to 28 days after birth, except for congenital anomalies deaths) and significant morbidity (moderate or severe neonatal encephalopathy, defined as the use of whole-body cooling or admission to the neonatal unit of measurement within 48 h of nascency for 48 h or more than with evidence of feeding difficulties or respiratory illness, with evidence of compromise at nativity suggesting mild asphyxia or mild encephalopathy, or both). The second was a continuous outcome of developmental progress measured past the Parent Report of Children's Abilities—Revised (PARCA-R) composite score
,
at age ii years for a subset of children.
Infant secondary outcomes, all of which are binary unless specified, were intrapartum stillbirth (excluding deaths from congenital anomalies); neonatal deaths up to 28 days after birth (excluding deaths from congenital anomalies); moderate or severe encephalopathy; admissions to the neonatal unit within 48 h of nascence for 48 h or more with evidence of feeding difficulties or respiratory disease (when there was testify of compromise at nascence suggesting mild asphyxia or mild encephalopathy, or both); access to a higher level of intendance; an Apgar score of less than four at five min; distribution of cord claret gas information for cord artery pH; metabolic acidosis (cord artery pH <vii·05 and base deficit of 12 mmol/50 or more); resuscitation interventions (categorical); seizures; destination immediately after nativity (categorical); length of hospital stay (continuous); health and development outcomes at 24 months (continuous); score on the not-verbal cognition scale (continuous), vocabulary sub-calibration (continuous), and sentence complexity sub-calibration (continuous) of PARCA-R; deaths to 24 months; major disability and not-major disability at 2 years; and cognitive palsy. Maternal secondary outcomes were mode of commitment (chiselled); operative intervention (caesarean department and instrumental delivery) for fetal indications, failure to progress, a combination of fetal indications and failure to progress, or any other reason; grade or urgency of caesarean section
15
- Lucas DN
- Yentis SM
- Kinsella SM
- Holdcroft A
- May AE
- Wee M
- et al.
Urgency of caesarean department: a new classification.
- PubMed
- Google Scholar
(categorical); episiotomy; whatever episode of fetal blood sampling; length of outset stage, length of second stage, and total length of labour from trial entry (continuous); destination immediately after birth (categorical); and admission to a college level of intendance.
Because trial resource allotment was non blinded, it was important to measure any change in clinical care that could outcome from clinicians being aware of whether the decision-support organisation was in operation. We measured the total number of cardiotocographic abnormalities and the proportion of women with cardiotocographic abnormalities in each arm; the time taken between concluding red alert and delivery (for these first iii outcomes, we retrospectively used the decision-back up software later on the trial was over to analyse the cardiotocographic trace, and used these data to make up one's mind when the alarm would take occurred); the number of routine measurements recorded during labour, including the number of vaginal examinations, use of epidural analgesia, employ of labour augmentation, and presence of meconium; and the number of thumb entries (similar to a signature in paper notes) per hr from fourth dimension of trial entry to offset yellowish level of concern or until the cervix was fully dilated (every bit a proxy mensurate to assess presence of a health professional in the delivery room during the labour).
Statistical assay
A sample size of 46 000 births was needed.
eleven
- Brocklehurst P
A study of an intelligent system to back up decision making in the direction of labour using the cardiotocograph—the Baby report protocol.
- Crossref
- PubMed
- Scopus (19)
- Google Scholar
Nosotros postulated an incidence of the main issue of three per 1000 births by summing the previous reported rates of intrapartum stillbirth, neonatal death, moderate and severe encephalopathy, and mild encephalopathy (reliable data for significant asphyxial morbidity were not available and and then could only be estimated).
xvi
- Kurinczuk JJ
- Barralet JH
- Redshaw M
- Brocklehurst P
Study to the Patient Prophylactic Research Programme (Policy Research Programme of the Department of Health). Monitoring the incidence of neonatal encephalopathy—what side by side?.
- Google Scholar
,
,
The effect size that could exist detected with 46 000 women (23 000 in each grouping), assuming a 5% level of significance and 90% ability, was a 50% reduction in poor neonatal outcomes from iii to 1·5 per g. In a written report of preterm infants,
the mean PARCA-R blended score at two years was fourscore (SD 33) and the mean Mental Development Alphabetize (Bayley Scales of Babe Development II) was effectually half an SD beneath the mean of 100. Assuming that a healthy group of term infants would accept a PARCA-R blended score half an SD college than this sample of preterm infants, we estimated a mean 2 year score of 96 (SD 33). A follow-up sample of size 7000 (3500 per arm) had more 90% power to find a divergence of 3 points in the PARCA-R component score with a two-sided 5% significance level. The incidence of severe metabolic acidosis (cord avenue pH <7·05) is ten per 1000.
,
,
,
22
- Yelland LN
- Salter AB
- Ryan P
Relative risk estimation in randomized controlled trials: a comparison of methods for contained observations.
- Google Scholar
A sample of 46 000 women enabled the states to discover a 28% relative risk reduction in this incidence with more than 80% power, bold a 5% level of significance, in babies in whom string artery pH was measured.
During the early on part of the trial, and with advice from the information-monitoring committee, the primary outcome definition was refined to ensure that it captured babies who were likely to take experienced hypoxia during labour. The original component of the primary outcome—access to neonatal unit of measurement within 48 h of birth for 48 h or more than with evidence of feeding difficulties, respiratory illness, or encephalopathy—resulted in the inclusion of many babies with a range of disorders, many of which were unlikely to exist related to hypoxia. Each instance fulfilling this component of the principal outcome was reviewed by an independent console of neonatologists (who were blinded to allocation) to accredit it as fulfilling the revised definition or not (appendix).
The trial steering commission approved the statistical analysis plan before the analysis (appendix). For the main comparative assay, participants were analysed in the groups into which they were randomly allocated, irrespective of allocation received. All women and babies with available data were included, except for women for whom a valid signed consent form could not exist located or women who withdrew consent. The numbers and percentages of babies in whom the main outcomes were noted are for each group, and the run a risk ratios plus 95% CIs were calculated. Run a risk ratios were estimated with generalised linear models with a binomial distribution and a log link (or a Poisson distribution with a log link if convergence could not exist achieved). Chance ratios were estimated with Cox regression and rate ratios with Poisson regression. We adapted for the stratification factors used in the randomisation (centre and singleton or twin pregnancy), and used robust variance estimators in all models to business relationship for the correlation in outcomes between twins and siblings delivered in a subsequent pregnancy during the trial menses.
11
- Brocklehurst P
A study of an intelligent system to support decision making in the direction of labour using the cardiotocograph—the INFANT study protocol.
- Crossref
- PubMed
- Scopus (19)
- Google Scholar
The hateful (SD) PARCA-R blended score was calculated for each group, and the mean deviation between groups plus 95% CI were calculated and compared with a Gaussian model with identity link. For secondary outcomes including the components of the master outcome, a one% level of significance was employed.
We did prespecified subgroup analyses with the statistical examination for interaction for singletons versus twins, suspected fetal growth brake at labour onset versus no growth restriction, body-mass index group, and middle. These analyses were done for the trial primary outcomes, all neonatal outcomes, instrumental vaginal deliveries, and caesarean section. Additionally, we did a subgroup assay of all process outcomes past eye. Major inability at 2 years was classified in terms of neuromotor part, seizures, auditory function, communication, visual office, cerebral function, and other physical disability.
,
Nosotros used Stata/SE for Windows (version 13.1) for all analyses. The trial was overseen by an independent trial steering committee and an independent data-monitoring committee. The data-monitoring committee used the Haybittle-Peto approach
for interim analyses, with 3 SEs every bit the cutoff for consideration of early cessation, preserving the blazon-1 mistake across the trial.
The trial is registered with ISRCTN, number 98680152.
Role of the funding source
The funder of the written report had no role in the trial design; data collection, assay, or estimation; or writing of the report. The corresponding writer had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results
Between January 6, 2010, and Aug 31, 2013, 47 062 women were recruited to the Infant trial (appendix). 1020 women (two·2%) were excluded from the analysis of the main outcome (figure; appendix), mostly because of missing or incomplete consent forms. Information at the time of birth were available for 100% of women and babies eligible to be analysed. Follow-up data at 2 years were available for 7066 of those contacted; data were sufficiently consummate for 6707 children.
Figure Trial contour
Show full caption
The 46 042 women included in the analysis includes 448 women with two singleton birth episodes and half dozen women with one singleton and one twin nascency episode in the study period. The allocation received for the subsequent commitment was contained of the beginning allocation received. xxx of 21 509† infants in the decision-support grouping and 37 of 21 599† infants in the no-decision-support group died before 24 months. Parent Report of Children'due south Abilities—Revised (PARCA-R). *One adult female who withdrew (with consent to utilize of data) was besides randomly assigned in error. †Data from the Ireland not included in the numerator (north=ane decision support) or denominator (n=1754 in decision back up and northward=1752 in no decision support) because data for deaths after discharge were non available; deaths in the period nautical chart include stillbirths (n=1 in decision support and northward=2 in no determination back up).
- View Large Image
- Figure Viewer
- Download Hi-res epitome
- Download (PPT)
Baseline characteristics were similar betwixt the 2 groups (tabular array 1). Median maternal age was 29 years (IQR 25–33). Effectually threescore% of women were having their first baby, and well-nigh women had a gestational age between 38 and 41 completed weeks (table one). Very few women had a previous stillbirth (1%) and effectually 6% had previously had a previous caesarean section (tabular array 1). Almost 60% of women had their labour induced.
Table 1 Maternal characteristics at trial entry
| Determination support (n=22 987) * Women with more than 1 nativity episode in the study flow are included more than in one case (n=454). | No decision support (north=23 055) * Women with more than one nascency episode in the study period are included more than once (n=454). | ||
|---|---|---|---|
| Median age, years | 29 (25–33) | 29 (25–33) | |
| Ethnic grouping † As coded past the U.k. National Health Service. | |||
| White | 17 234 (83·3%) | 17 213 (83·0%) | |
| Indian | 743 (3·six%) | 724 (3·5%) | |
| Pakistani | 736 (iii·half-dozen%) | 802 (iii·ix%) | |
| Bangladeshi | 98 (0·5%) | 113 (0·5%) | |
| Black Caribbean | 116 (0·6%) | 135 (0·vii%) | |
| Black African | 461 (2·2%) | 505 (two·4%) | |
| Any other ethnic group | 1296 (6·three%) | 1249 (6·0%) | |
| Unknown | 2303 | 2314 | |
| Twin pregnancy | 276 (1·two%) | 296 (1·3%) | |
| Gestational age, completed weeks | |||
| Median | 40 (39–41) | 40 (39–41) | |
| <35 weeks | 4 (<1%) | 6 (<ane%) | |
| 35 weeks to 37 weeks, 6 days | 2529 (xi·0%) | 2522 (10·9%) | |
| 38 weeks to 39 weeks, 6 days | 7322 (31·9%) | 7266 (31·five%) | |
| 40 weeks to 41 weeks, 6 days | xi 688 (l·9%) | 11 795 (51·2%) | |
| ≥42 weeks | 1437 (half-dozen·3%) | 1457 (6·3%) | |
| Body-mass alphabetize (at booking visit) | |||
| Median | 25 (22–30) | 25 (22–thirty) | |
| <18·5 | 379 (2·5%) | 384 (2·six%) | |
| 18·5– 24·ix | 6302 (42·1%) | 6225 (41·6%) | |
| 25·0–29·nine | 4531 (30·2%) | 4560 (thirty·5%) | |
| xxx·0–34·9 | 2178 (14·5%) | 2237 (14·nine%) | |
| 35·0–39·9 | 1024 (vi·8%) | 1025 (6·8%) | |
| ≥twoscore·0 | 565 (3·viii%) | 544 (iii·6%) | |
| Unknown | 8008 | 8080 | |
| Smoking (at booking visit) | |||
| Yep | 2448 (14·3%) | 2536 (14·7%) | |
| No | 14 724 (85·seven%) | fourteen 722 (85·3%) | |
| Unknown | 5815 | 5797 | |
| Parity | |||
| Nulliparous | 13 736 (59·8%) | xiii 650 (59·two%) | |
| Parous | 9247 (xl·2%) | 9390 (forty·8%) | |
| Obstetric history | |||
| Stillbirth | 273 (one·two%) | 223 (1·0%) | |
| Elective caesarean section | 208 (0·9%) | 253 (ane·1%) | |
| Emergency caesarean section | 1240 (v·4%) | 1224 (5·3%) | |
| Neonatal decease | 80 (0·4%) | 95 (0·4%) | |
| Cervical dilatation at trial entry (cm) | |||
| Median | iv (2–vi) | 4 (two–v) | |
| Unknown | 16 184 | 16 339 | |
| Fetal growth restriction suspected at labour onset | 859 (3·7%) | 914 (4·0%) | |
| Labour consecration | |||
| Induced | thirteen 516 (59·2%) | 13 568 (59·2%) | |
| Spontaneous | 8955 (39·two%) | 8967 (39·two%) | |
| No labour | 376 (one·7%) | 367 (1·vi%) | |
| Epidural analgesia ‡ These data were nerveless simply from 2013 at most centres. | |||
| Aye | 2682 (26·0%) | 2766 (26·8%) | |
| No | 7628 (74·0%) | 7549 (73·ii%) | |
| Unknown | 12 677 | 12 740 | |
| Presence of meconium ‡ These data were collected only from 2013 at most centres. | |||
| Yep | 449 (4·five%) | 454 (four·v%) | |
| No | 9454 (95·5%) | 9535 (95·five) | |
| Unknown | 13 084 | 13 066 | |
Data are median (IQR), n, or n (%). Missing data are <1%, unless otherwise presented; there were no apparent differences in missing data between trial arms.
* Women with more one nativity episode in the study period are included more than once (n=454).
† Equally coded by the UK National Wellness Service.
‡ These data were collected just from 2013 at most centres.
- Open table in a new tab
The incidence of the primary outcome—poor neonatal outcome—did not differ significantly betwixt the groups (tabular array ii). 172 (0·vii%) of 23 263 babies had a poor outcome in the determination-back up group compared with 171 (0·seven%) of 23 351 babies in the no-conclusion-support group (adjusted risk ratio [RR] ane·01, 95% CI 0·82–1·25; table ii). Similarly, we noted no show of a departure in any component of the composite primary outcome betwixt the groups (table 2). A prespecified sensitivity analysis, in which nosotros used a different cutoff for defining compromise at birth (a score of 7 or greater indicating very severe compromise rather than a score of iii or greater on a scale from 0 to 14), made no divergence to the interpretation of the measure out of effect for the primary upshot (adapted RR 0·97, 95% CI 0·58–1·63; appendix). We noted no evidence of any differences in any of the trial's secondary outcomes for the baby (table 2), including Apgar scores, admission to the neonatal unit, metabolic acidosis of string blood samples, the need for neonatal resuscitation, or duration of infirmary stay.
Table ii Primary and secondary outcomes
| Determination support (north=23 263) | No decision support (north=23 351) | Adjusted risk ratio (CI) | ||
|---|---|---|---|---|
| Composite neonatal main outcome | ||||
| Composite primary outcome * The components of the main outcome are mutually exclusive and outcomes listed higher accept precedence over those listed below them—eg, if a baby with neonatal encephalopathy died within 28 days, the upshot would be recorded as neonatal death. | 172 (0·7%) | 171 (0·7%) | 1·01 (95% CI 0·82–one·25) | |
| Intrapartum stillbirths † Excludes stillbirths due to built anomalies. | i (0) | two (0) | 0·50 (95% CI 0·05–5·53) | |
| Neonatal deaths up to 28 days after nascency ‡ Excludes deaths due to built anomalies; deaths afterward infirmary discharge non reported for Ireland. | six (0) | four (0) | ane·51 (95% CI 0·42–5·33) | |
| Moderate or astringent neonatal encephalopathy (requiring cooling) | eighteen (0·one%) | 21 (0·1%) | 0·86 (95% CI 0·46–1·61) | |
| Access to neonatal unit within 48 h of nascency for ≥48 h considering of feeding difficulties, respiratory illness or symptoms, or encephalopathy and evidence of compromise at birth | 147 (0·6%) | 144 (0·six%) | 1·02 (95% CI 0·81–1·29) | |
| Other neonatal outcomes | ||||
| Access to a higher level of care | 1389 (6·0%) | 1429 (6·1%) | 0·98 (99% CI 0·89–1·08) | |
| Apgar score <4 at 5 min | 43 (0·2%) | 65 (0·iii%) | 0·67 (99% CI 0·xl–1·11) | |
| String artery pH | ||||
| <7·15 | 1625 (11·three%) | 1695 (11·viii%) | 0·96 (99% CI 0·88–1·04) | |
| <7·05 | 268 (1·9%) | 278 (1·9%) | 0·95 (99% CI 0·77–ane·nineteen) | |
| Mean (SD) | 7·24 (0·08) | seven·24 (0·08) | ·· | |
| Unknown | 8829 | 8981 | ·· | |
| Metabolic acidosis § Cord artery pH <7·05, base of operations deficit ≥12mmol/50. | ||||
| Yes | 148 (1·1%) | 131 (1·0%) | i·12 (99% CI 0·82–1·52) | |
| No | 13 538 (98·9%) | xiii 533 (99·0%) | ·· | |
| Unknown | 9577 | 9687 | ·· | |
| Resuscitation | ||||
| None | 18 457 (87·three%) | 18 605 (87·6%) | ·· | |
| One intervention | 2139 (10·i%) | 2116 (10·0%) | 1·03 ¶ Risk ratio of 1 or more than interventions vs none. | |
| Two or more interventions | 554 (2·6%) | 524 (2·5%) | ·· | |
| Unknown | 2113 | 2106 | ·· | |
| Seizures in infirmary | 39 (0·two%) | 41 (0·2%) | 0·95 (99% 0·54–one·70) | |
| Destination of baby immediately after nascence | ||||
| Postnatal ward | 21 571 (93·half dozen%) | 21 664 (93·6%) | ·· | |
| Habitation | 467 (2·0%) | 485 (2·1%) | 1·00 Risk ratio of ward or domicile vs all other destinations (if known). | |
| Transitional care unit | 277 (ane·2%) | 235 (1·0%) | ·· | |
| Neonatal unit | 653 (2·8%) | 690 (3·0%) | ·· | |
| Transferred hospital | 4 (0) | seven (0) | ·· | |
| Stillbirth | 1 (0) | 2 (0) | ·· | |
| Other | 69 (0·3%) | 53 (0·2%) | ·· | |
| Median length of hospital stay, days | 2 (one–three) | 2 (1–3) | 0·99 ** This effigy is a risk ratio rather than a risk ratio. | |
| Delivery outcomes | ||||
| Mode of delivery | ||||
| Spontaneous cephalic vaginal | eleven 823 (l·viii%) | xi 959 (51·ii%) | 0·99 (99% CI 0·97–ane·01) | |
| Caesarean department | 5669 (24·four%) | 5555 (23·8%) | ·· | |
| Instrumental | 5698 (24·5%) | 5765 (24·7%) | ·· | |
| Vaginal breech | 73 (0·iii%) | 72 (0·iii%) | ·· | |
| Indications for whatever operative intervention (caesarean section and instrumental delivery) | ||||
| Fetal distress | 4278 (18·four%) | 4262 (18·3%) | 1·04 Risk ratio for fetal distress and gamble ratio for failure to progress include (in the numerator) deliveries in the third category for which both fetal distress and failure to progress were recorded. | |
| Failure to progress | 5059 (21·eight%) | 5175 (22·2%) | 1·01 Gamble ratio for fetal distress and take a chance ratio for failure to progress include (in the numerator) deliveries in the 3rd category for which both fetal distress and failure to progress were recorded. | |
| Fetal distress and failure to progress | 1774 (7·half dozen%) | 1599 (6·9%) | ·· | |
| Other reason | 229 (1·0%) | 247 (one·1%) | ·· | |
| Indication for instrumental vaginal deliveries | ||||
| Fetal distress | 2608 (eleven·ii%) | 2559 (11·0%) | 1·03 Risk ratio for fetal distress and risk ratio for failure to progress include (in the numerator) deliveries in the tertiary category for which both fetal distress and failure to progress were recorded. | |
| Failure to progress | 2262 (9·7%) | 2396 (ten·iii%) | 0·97 Risk ratio for fetal distress and risk ratio for failure to progress include (in the numerator) deliveries in the 3rd category for which both fetal distress and failure to progress were recorded. | |
| Fetal distress and failure to progress | 700 (3·0%) | 660 (two·8%) | ·· | |
| Other reason | 117 (0·five%) | 134 (0·6%) | ·· | |
| Caesarean section | ||||
| Grade 1 (immediate threat to life) | 1138 (4·9%) | 1121 (4·8%) | 1·02 Gamble ratios based on cumulative totals for grade—ie, grade 1 vs all other deliveries, grade 1–2 vs all other deliveries, and grade 1–3 vs all other deliveries. | |
| Class 2 (some threat of compromise) | 3754 (sixteen·ii%) | 3605 (15·5%) | 1·04 Chance ratios based on cumulative totals for grade—ie, class ane vs all other deliveries, course i–2 vs all other deliveries, and grade 1–3 vs all other deliveries. | |
| Grade 3 (no threat of compromise) | 645 (2·eight%) | 689 (3·0%) | 1·02 Risk ratios based on cumulative totals for grade—ie, grade 1 vs all other deliveries, grade 1–2 vs all other deliveries, and grade i–3 vs all other deliveries. | |
| Grade 4 (elective) | 12 (0·ane%) | 12 (0·one%) | ·· | |
| Episiotomy due north=22 987 in the decision-support group and 23 055 in the no-decision-back up grouping (ie, the number of women, rather than the number of infants) for all outcomes after this row. | ||||
| Yes | 6396 (28·nine%) | 6498 (29·three%) | 0·99 (99% CI 0·95–ane·03) | |
| Unknown | 826 | 840 | ·· | |
| Any episode of fetal claret sampling northward=22 987 in the decision-support group and 23 055 in the no-decision-support group (ie, the number of women, rather than the number of infants) for all outcomes subsequently this row. | 2366 (x·3%) | 2187 (9·5%) | 1·08 (99% CI one·01–ane·16) | |
| Destination of female parent immediately after birth n=22 987 in the decision-support group and 23 055 in the no-conclusion-back up group (ie, the number of women, rather than the number of infants) for all outcomes after this row. | ||||
| Ward | 21 554 (94·half-dozen%) | 21 614 (94·v%) | ·· | |
| Home | 429 (one·ix%) | 462 (2·0%) | 1·00 Chance ratio of ward or home vs all other destinations (if known). | |
| Intensive-care unit | xv (0·i%) | 19 (0·1%) | ·· | |
| High-dependency unit | 793 (3·five%) | 768 (three·iv%) | ·· | |
| Theatre | 0 (0) | 0 (0) | ·· | |
| Other hospital | 0 (0) | 8 (0) | ·· | |
| Access to a higher level of care n=22 987 in the decision-back up grouping and 23 055 in the no-decision-support group (ie, the number of women, rather than the number of infants) for all outcomes after this row. | 1245 (five·4%) | 1193 (5·2%) | 1·05 (99% CI 0·95–1·16) | |
| Duration of labour | ||||
| No labour | 378 | 371 | ·· | |
| Seemingly randomised after delivery | 92 | 120 | ·· | |
| Length of labour from trial entry (min) || Denominators exclude women with no labour, seemingly randomised after delivery, and unknown length of labour. | ||||
| Geometric mean and geometric hateful ratio | 379 | 381 | 0·99 (99% CI 0·98–1·01) | |
| Median | 404 (234–638) | 408 (236–640) | ·· | |
| Unknown | 871 | 924 | ·· | |
| Length of outset stage from trial entry (min) || Denominators exclude women with no labour, seemingly randomised after delivery, and unknown length of labour. | ||||
| Geometric mean and geometric mean ratio | 169 | 168 | 1·01 (99% CI 0·98–1·04) | |
| Median | 200 (100–351) | 201 (96–354) | ·· | |
| Unknown | 6422 | 6292 | ·· | |
| Length of 2nd stage from trial entry (min) || Denominators exclude women with no labour, seemingly randomised after delivery, and unknown length of labour. | ||||
| Geometric mean and geometric mean ratio | 39 | 39 | 0·99 (99% CI 0·96–ane·03) | |
| Median | 49 (15–113) | 50 (16–114) | ·· | |
| Unknown | 6036 | 5934 | ·· | |
Information are n (%), n, or median (IQR), unless otherwise specified. Missing data are <1% unless otherwise presented. Risk ratios were adapted for stratification factors used in the randomisation (eye and twin birth) and clustering because of twins and multiple nativity episodes. Minimisation factors were not adapted for in the analysis of intrapartum stillbirths, neonatal deaths, and neonatal encephalopathy considering of the modest number of events. Crude event measures were not presented every bit identical to i decimal place (two decimal places for most outcomes). CI=confidence interval.
* The components of the primary consequence are mutually sectional and outcomes listed college take precedence over those listed beneath them—eg, if a infant with neonatal encephalopathy died within 28 days, the consequence would be recorded as neonatal death.
† Excludes stillbirths due to congenital anomalies.
‡ Excludes deaths due to congenital anomalies; deaths subsequently hospital discharge non reported for Ireland.
§ Cord avenue pH <vii·05, base of operations deficit ≥12mmol/L.
¶ Risk ratio of one or more than interventions vs none.
|| Take chances ratio of ward or abode vs all other destinations (if known).
** This figure is a hazard ratio rather than a risk ratio.
†† Chance ratio for fetal distress and risk ratio for failure to progress include (in the numerator) deliveries in the tertiary category for which both fetal distress and failure to progress were recorded.
‡‡ Adventure ratios based on cumulative totals for grade—ie, form 1 vs all other deliveries, form 1–two vs all other deliveries, and grade 1–3 vs all other deliveries.
§§ north=22 987 in the determination-support grouping and 23 055 in the no-decision-support group (ie, the number of women, rather than the number of infants) for all outcomes after this row.
¶¶ Risk ratio of ward or dwelling vs all other destinations (if known).
|||| Denominators exclude women with no labour, seemingly randomised after delivery, and unknown length of labour.
- Open table in a new tab
Simply over half of all births were spontaneous vaginal births and the frequency did not differ significantly between the ii groups (adjusted RR 0·99, 99% CI 0·97–ane·01; table ii). More women underwent fetal claret sampling in the decision-support group than in the no-conclusion-back up grouping (2366 [10·3%] vs 2187 [ix·5%]; adjusted RR 1·08, 99% CI 1·01–1·16). No other significant differences were noted between the two groups from trial entry to nascency in terms of clinical outcomes (table two).
For babies with an adverse event and cord metabolic acidosis who underwent skilful review, the overall proportion of babies judged to take received suboptimal care likely to accept affected the outcome was 38%—14 of 35 babies in the decision-support grouping and xiii of 36 babies in the control grouping—which is similar to that reported previously.
Nosotros could not investigate whether in all cases not reviewed appropriate action was taken in response to recognised abnormality.
In women with any level of business concern equally measured past INFANT (table iii), bluish levels of business organization were well-nigh frequent (median seven alerts during labour—roughly 1·1 per h), followed by yellowish alerts (median 2 alerts per labour), and and so scarlet alerts (median one per labour). A lower charge per unit of yellow levels of concern was noted in the decision-support group compared with the no-decision-support group (adjusted rate ratio 0·87, 99% CI 0·84–0·89; table iii). Frequency of blue and ruby alerts did non differ significantly between groups (tabular array 3).
Table 3 Process outcomes after trial entry
| Determination back up (due north=22 517) | No decision support * For the command group with cardiotocographic monitoring only, decision-support software was used retrospectively to determine when an alert would have sounded. | Adjusted event measure out (99% CI) | ||
|---|---|---|---|---|
| Epidural analgesia | ||||
| Yes | 2770 (27·3%) | 2689 (26·5%) | Hazard ratio 1·03 (99% CI 0·97–1·09) | |
| No | 7383 (72·7%) | 7453 (73·5%) | ·· | |
| Unknown † These data were only recorded and uploaded for analysis from 2013 for most centres. | 12 364 | 12 422 | ·· | |
| Labour augmentation | ||||
| Yep | 2705 (30·9%) | 2750 (31·3%) | Risk ratio 0·99 (99% CI 0·93–1·04) | |
| No | 6047 (69·1%) | 6042 (68·vii%) | ·· | |
| Unknown † These data were only recorded and uploaded for analysis from 2013 for virtually centres. | 13 765 | 13 772 | ·· | |
| Presence of meconium | ·· | ·· | ||
| Yes | 440 (4·v%) | 469 (iv·8%) | Take chances ratio 0·94 (99% CI 0·80–1·11) | |
| No | 9316 (95·5%) | 9346 (95·2%) | ·· | |
| Unknown † These data were only recorded and uploaded for analysis from 2013 for most centres. | 12 761 | 12 749 | ·· | |
| At least 1 blueish, yellow, or blood-red level of business | 21 950 (97·five%) | 22 021 (97·6%) | Risk ratio 1·00 (99% CI one·00–1·00) | |
| At least one bluish level of business organisation (mild aberration) | 21 863 (97·1%) | 21 913 (97·1%) | Risk ratio 1·00 (99% CI ane·00–1·00) | |
| At least one yellow level of business concern (moderate abnormality) | 16 765 (74·5%) | 16 722 (74·1%) | Run a risk ratio ane·00 (99% CI 0·99–1·02) | |
| At least one ruby level of business organization (severe abnormality) | 2335 (10·viii%) | 2413 (11·1%) | Risk ratio 0·97 (99% CI 0·ninety–1·04) | |
| Unknown ‡ Data for timing of carmine level of concerns not available for two centres—Warwick (n=823) and Derby (n=832)· | 822 | 833 | ·· | |
| Blue, yellow, or red levels of business organisation in women with at least one level of business organization | ||||
| Median | nine (5–15) | ix (five–15) | ·· | |
| Rate (per h) | 1·37 | 1·xl | Charge per unit ratio 0·98 (99% CI 0·96–1·00) | |
| Unknown § Women with missing length of labour were not included in calculation of rates and rate ratios. | 765 | 824 | ||
| Blue levels of concern in women with a blue level | ||||
| Median | 7 (four–xi) | 7 (iv–eleven) | ·· | |
| Rate (per h) | 1·06 | ane·05 | Rate ratio 1·01 (99% CI 0·99–1·03) | |
| Unknown § Women with missing length of labour were not included in calculation of rates and charge per unit ratios. | 740 | 800 | ·· | |
| Yellow levels of concern in women with a xanthous level | ||||
| Median | 2 (one–4) | 2 (1–5) | ·· | |
| Rate (per h) | 0·35 | 0·xl | Rate ratio 0·87 (99% CI 0·84–0·89) | |
| Unknown § Women with missing length of labour were non included in calculation of rates and rate ratios. | 354 | 421 | ·· | |
| Red levels of concern in women with a blood-red level | ||||
| Median | 1 (ane–one) | 1 (1–1) | ·· | |
| Rate (per h) | 0·xiv | 0·xiv | Rate ratio 0·98 (99% CI 0·92–one·04) | |
| Unknown ‡ Information for timing of red level of concerns not bachelor for two centres—Warwick (northward=823) and Derby (northward=832)· § Women with missing length of labour were not included in calculation of rates and charge per unit ratios. | 41 | 55 | ·· | |
| Interaction with Guardian arrangement ¶ Measured via number of thumbprint entries from fourth dimension of trial entry to first yellow level of business, or until cervix fully dilated if no abnormality detected. | ||||
| Median | 5 (0–16) | 4 (0–15) | ·· | |
| Charge per unit (per h) | 4·22 | 4·21 | Rate ratio 0·99 (99% CI 0·95–i·03) | |
| Unknown | 1723 | 1603 | ·· | |
| Vaginal examinations | ||||
| Median | 2 (1–3) | ii (1–three) | ·· | |
| Rate (per h) | 0·28 | 0·27 | Rate ratio 1·03 (99% CI i·00–1·05) | |
| Unknown | 877 | 929 | ·· | |
| Time from last blood-red level of concern to delivery (mins) | ||||
| Median | 58 (13–279) | 58 (thirteen–264) | HR 0·99 (99% CI 0·92–1·06) | |
| Unknown ‡ Information for timing of ruby-red level of concerns not available for two centres—Warwick (n=823) and Derby (n=832)· | 822 | 823 | ·· | |
Data are n (%), n, or median (IQR), unless otherwise specified. Women with no labour or seemingly randomised after commitment were not included in calculations in this table, which is why the denominators differ from those in the footnotes of table 2. Effect measures were adjusted for stratification factors used in the randomisation (centre and twin birth) and clustering as a outcome of twins and multiple nascency episodes. Crude result measures are not presented as identical to one decimal place (two decimal places for most outcomes). CI=confidence interval. HR=hazard ratio.
* For the control group with cardiotocographic monitoring but, decision-back up software was used retrospectively to determine when an alert would accept sounded.
† These data were only recorded and uploaded for assay from 2013 for most centres.
‡ Data for timing of ruddy level of concerns not available for two centres—Warwick (n=823) and Derby (n=832)·
§ Women with missing length of labour were not included in calculation of rates and rate ratios.
¶ Measured via number of thumbprint entries from time of trial entry to kickoff yellow level of concern, or until neck fully dilated if no aberration detected.
- Open up tabular array in a new tab
Although at that place was a worry that women in the conclusion-back up grouping would exist left solitary more frequently during labour than those in the no-decision-support group, the frequency of thumbprint entries on the Guardian did not differ significantly between groups (four·22 per h vs 4·21; adapted rate ratio 0·99, 99% CI 0·95–1·03; table 3).
Time from the last red level of concern to nascence was like in both groups (median 58 min; tabular array 3). In a subgroup of 473 readable traces from a sample of 500 taken as a similar number of sequent cases from each contributing eye, the last ruby-red level of concern was judged (past expert investigator, PS) to exist a valid fetal concern for 276 (58%) traces. Maternal centre rate triggered the red level of business in 128 (27%) cases, misclassified accelerations in 36 (8%) cases, and other reasons in 33 (vii%) cases.
Families were contacted when their surviving child or children born reached age 2 years. Nearly 7000 families returned a questionnaire. The characteristics of the mothers who responded differed significantly from those of the entire trial accomplice and from those of mothers who did non answer (appendix). Compared with the entire trial cohort, responders were more than likely to be white, to have given nascence at a afterward gestational age, and to have been having their commencement baby (appendix).
Of the 7066 infants for whom a questionnaire was returned, data could be analysed for 6707 (95%). We noted no significant differences betwixt the two groups for any of the 2 year outcomes, including the primary outcome, PARCA-R score (table iv). Nearly 6% of children for whom data were available had a major disability. The classification of disability
,
used meant that large numbers of children were assigned a major disability as a result of having poor growth (betwixt 2·8% and iii·0% of all children) and cognitive difficulties (between 1·2% and ane·5% of all children). Other major disabilities such as physical disability, blindness, and deafness were all very uncommon (appendix).
Tabular array 4 Health and development outcomes at 2 years in a sample of surviving infants without the primary outcome selected for follow-up
| Determination back up (due north=3556) | No decision support (n=3510) | Adjusted effect measure (CI) | ||
|---|---|---|---|---|
| PARCA-R composite score | ||||
| Mean (SD) | 98·0 (33·viii) | 97·2 (33·4) | Mean deviation 0·63 (95% CI −0·98 to ii·25) | |
| Median (IQR) | 98 (73–126) | 97 (72–125) | ·· | |
| Unknown | 175 | 184 | ·· | |
| PARCA-R non-verbal cognition scale | ||||
| Hateful (SD) | 27·7 (three·7) | 28·0 (3·half-dozen) | Mean difference −0·22 (99% CI −0·44 to 0·01) | |
| Median (IQR) | 28 (26–30) | 28 (26–31) | ·· | |
| PARCA-R vocabulary sub-calibration | ||||
| Hateful (SD) | 57·iv (27·8) | 56·five (27·seven) | Mean departure 0·82 (99% CI −0·91 to 2·54) | |
| Median (IQR) | 58 (36–81) | 56 (35–80) | ·· | |
| PARCA-R judgement complexity sub-scale | ||||
| Mean (SD) | 12·iv (v·four) | 12·3 (five·iii) | Mean difference 0·07 (99% CI −0·26 to 0·41) | |
| Median (IQR) | 12 (9–xvi) | 12 (9–xvi) | ·· | |
| Cerebral palsy | ||||
| n (%) | four (0·one%) | iv (0·1%) | RR 0·99 (99% CI 0·16 to vi·10) | |
| Unknown (due north) | 111 | 114 | ·· | |
| Non-major or major disability * Disability in any of neuromotor role, seizures, auditory office, communication, visual function, cerebral function, or other concrete disability. | ·· | ·· | RR 1·08 (99% CI 0·98 to i·18) | |
| n (%) | 942 (40·4%) | 840 (37·4%) | ·· | |
| Unknown (north) | 1225 | 1266 | ·· | |
| Major disability * Disability in any of neuromotor function, seizures, auditory function, communication, visual function, cognitive function, or other physical inability. | ||||
| n (%) | 134 (v·vii%) | 135 (6·0%) | RR 0·95 (99% CI 0·lxx to 1·29) | |
| Unknown (due north) | 1225 | 1266 | ·· | |
Missing data are <3% unless otherwise presented; there were no apparent differences in missing data between trial arms. Effect measures were adapted for stratification factors used in the randomisation (centre and twin nativity) and clustering equally a effect of twins and multiple birth episodes. Stratification factors were not adjusted for in the analysis of infant deaths at two years and cerebral palsy because of the pocket-size number of events. Rough issue measures were not presented as identical to one decimal place (2 decimal places for well-nigh outcomes). At 2 years, 29 (0·13%) of 21 508 infants had died in the decision-support group and 35 (0·xvi%) of 21 597 infants had died in the no-decision-support group (adjusted RR 0·83 [99% CI 0·44–1·59])—all deaths were reported to age 2 years excluding stillbirths (i in the decision-support group and two in the no-determination-support group). Data from Republic of ireland were not included in the numerator (n=1 in the decision-support grouping) or denominator (north=1754 in decision-support grouping and due north=1752 in the no-conclusion-support group) considering data for deaths after discharge were not bachelor. CI=confidence interval. PARCA-R=Parent Written report of Children's Abilities—Revised. RR=gamble ratio.
* Disability in whatever of neuromotor function, seizures, auditory function, advice, visual function, cerebral function, or other concrete disability.
- Open table in a new tab
We noted no evidence that the decision-support software performed significantly differently between whatsoever of the pre-specified subgroups for either the main consequence or a range of secondary outcomes (appendix). Furthermore, no differences were noted in the distribution of cord blood pH measurements (appendix). The number of alerts differed significantly by centre, but no other meaning differences were noted by middle (appendix).
Discussion
In this trial of more than 46 000 women, we found no evidence that the use of conclusion-support software in conjunction with cardiotocography reduced the likelihood of poor neonatal outcomes compared with cardiotocography alone.
In another randomised trial, which likewise recruited in the UK, the employ of decision support was assessed in women monitored during labour with fetal electrocardiographic monitoring.
27
- Nunes I
- Ayres-de-Campos D
- Ugwumadu A
- et al.
Central fetal monitoring with and without computer analysis. A randomised controlled trial.
- Google Scholar
This study too showed no evidence that conclusion support improved the primary effect of cord blood metabolic acidosis in 7730 women.
27
- Nunes I
- Ayres-de-Campos D
- Ugwumadu A
- et al.
Central fetal monitoring with and without figurer analysis. A randomised controlled trial.
- Google Scholar
The results of 1 small trial
28
- Ignatov P
- Atanasov B
Indirect standard cardiotocography plus fetal claret sampling versus indirect quantitative cardiotocography— a randomised comparative written report in intrapartum monitoring.
- PubMed
- Google Scholar
of 220 women in Bulgaria have suggested that decision support is associated with benefits with respect to cord claret metabolic acidosis. In the United kingdom continuous electronic fetal eye-rate monitoring is not routine, making generalisability to settings in which it is routine less certain.
The strengths of this study are its contemporaneous information drove and size. Potential weaknesses include the challenges of utilize of a blended primary outcome mensurate, the potential for staff to learn from exposure to the decision-back up arm of the trial, resulting in improved outcomes in the control arm, and the effect of bookkeeping for multiple comparisons.
Use of a blended primary outcome might not always exist helpful if different components of the effect respond differently to the intervention.
Nosotros initially hypothesised that the components of the event would have similar incidences, with each component probable to contribute effectually a 3rd to the composite. Estimates of the incidence of these components for eligible women were difficult to notice before the trial began.
11
- Brocklehurst P
A study of an intelligent system to support determination making in the management of labour using the cardiotocograph—the Infant study protocol.
- Crossref
- PubMed
- Scopus (19)
- Google Scholar
The perinatal bloodshed in our study (thirteen per 46 614 babies [0·3 per m]) was lower than the previous estimate (1·05 per k), and the incidence of neonatal encephalopathy requiring cooling was also lower than previous estimates (0·eight per grand vs one·3 per yard).
eleven
- Brocklehurst P
A study of an intelligent system to support determination making in the management of labour using the cardiotocograph—the INFANT study protocol.
- Crossref
- PubMed
- Scopus (xix)
- Google Scholar
However, the incidence of prolonged admission to neonatal units with show of compromise at birth, for which we had no good data when planning the trial, occurred more than oft (291 per 46 614 [6 per thousand]) and contributed essentially more than to the higher-than-anticipated overall primary event rate of seven per thousand compared with our estimated 3 per 1000. This frequency afforded us power to detect smaller differences in the composite upshot than we had originally planned.
The potential weakness of staff learning from exposure to the decision-support system was identified when planning the trial. Nosotros acknowledged that passive learning from the determination-support system would be possible. However, function of our previous hypothesis was that, although some poor cardiotocographic interpretation is due to a lack of grooming, some clinicians have poor intrinsic design-recognition abilities, which, past definition, would not be afflicted past training, and the performance of such clinicians would be especially improved past help from automatic interpretation. We collected a range of process outcomes to measure the touch on clinician behaviour during the trial, and these information suggested some evidence for behaviour change in the decision-support arm: fetal blood sampling was more than frequent and the incidence of repeated yellowish alerts lower than in the command grouping. Mayhap different activity was taken in response to the alerts in the decision-support arm of the trial—eg, clinicians might take reduced the dose of an oxytocin infusion in women having their labour augmented if the infusion was leading to very frequent contractions. Such deportment could have prevented farther yellowish alerts, leading to a decrease in this grouping, although we practise non accept any straight evidence for this scenario. Even if this effect did occur, information technology did non result in whatsoever significant change in clinical outcomes. Although the median time from last level of red concern to nascence might seem lengthy (58 min), some red levels of concern did non prompt immediate commitment—eg, the cardiotocographic monitor picking up the maternal heart rate.
Nosotros accounted for multiple comparisons in the trial by using 99% CIs for all secondary outcomes. Significant findings in secondary outcomes require careful interpretation irrespective of the level of significance, and factors such equally the force of the finding and plausibility need to be taken into business relationship.
In this trial, the only two significant findings relate to behaviour alter in clinicians favouring decision support. These findings were in the expected direction of result and are mutually supportive, suggestive of a real effect.
Detection of abnormalities in the fetal heart rate can improve outcome only if caregivers respond accordingly to the alerts. A review of all severe agin outcomes in the trial showed no evidence of differences in suboptimal care between the 2 groups. Therefore our hypothesis that substandard care is largely related to failure to identify pathological fetal centre-rate patterns is not supported. Almost adverse outcomes associated with preventable substandard intendance seemed to involve failure to have appropriate management decisions one time the cardiotocographic abnormality had been recognised. This aspect will be reported in detail in a follow-up paper. Our hypothesis that unnecessary intervention would be reduced was also not supported.
The conclusion-back up software used in this trial identifies fetal heart-rate abnormalities.
6
- Keith RDF
- Beckley Southward
- Garibaldi JM
- Westgate JA
- Ifeachor EC
- Greene KR
A multicentre comparative report of 17 experts and an intelligent figurer system for managing labour using the cardiotocogram.
- Crossref
- PubMed
- Scopus (119)
- Google Scholar
,
vii
Skinner JF, Harris G, Greene KR. Computerised decision support for managing labour using the cardiotocogram: 500 cases with the range of abnormality. 28th British Congress of Obstetrics and Gynaecology; Harrogate, UK; June thirty–July 4, 1998.
- Google Scholar
,
8
- Keith RDF
- Greene KR
Evolution, evaluation and validation of an intelligent conclusion support tool for the management of labour.
- Google Scholar
However, the alerts practice not take into account other information about the labour, such as duration of labour, the rate of labour progress, and presence of meconium, all of which could modify the way a clinician interprets the fetal center rate and acts on this information. Further development of determination-support software could improve the quality of feedback that the system provides to clinicians to brand a difference to outcomes. In view of the importance for parents, clinicians, and wellness services of the consequences of intrapartum hypoxia, identification of signs of early compromise during labour then that timely intervention can exist used to reduce poor outcomes is an urgent unmet need.
Correspondence to: Prof Peter Brocklehurst, Birmingham Clinical Trials Unit, Institute of Applied Health Research, College of Medical and Dental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK [e-mail protected]
Contributors
All members of the writing committee contributed to the development of the protocol and management and running of the trial. LL did the analyses. PB wrote the Article, and revised it with input from all authors. All authors read and approved the terminal Article.
Writing committee
Peter Brocklehurst, David Field, Keith Greene, Ed Juszczak, Sara Kenyon, Louise Linsell, Christopher Mabey, Mary Newburn, Rachel Plachcinski, Maria Quigley, Elizabeth Schroeder, Philip Steer.
Co-investigator group
Peter Brocklehurst, David Field, Keith Greene, Ed Juszczak, Robert Keith, Sara Kenyon, Louise Linsell, Christopher Mabey, Mary Newburn, Rachel Plachcinski, Maria Quigley, Elizabeth Schroeder, Philip Steer.
Recruiting centres
Birmingham Women's Infirmary (Nina Johns, Tracey Johnston, Gemma Barnfield, Karen Davies), Chelsea and Westminster Hospital (Marking Johnson, Holly Patterson), Derriford Infirmary, Plymouth (Imogen Montague, Sally Watmore, Alison Stolton), Homerton University Hospital (Maryam Parisaei, Natasha McGhee, Silvia Segovia), Lancashire Women & Newborn Center (Elizabeth Martindale, Hilary Jackson, Josephine Holleran), Liverpool Women'southward Hospital (Devender Roberts, Siobhan Holt), Northwick Park Hospital (Bosko Dragovic, Miriam Willmott-Powell, Laura Hutchinson, Benedek Toth, Gemma Chandler, Suzanne Ridley), Nottingham Metropolis Hospital (George Bugg, Anna Molnar, Denise Lochrie), Princess Anne Infirmary (Jillian Connor, David Howe, Katie Head, Sue Wellstead), Princess Purple Hospital (Alan Mathers, Laura Walker, Isobel Crawford), Queen Alexandra Hospital (David Davies, Zoe Garner, Lucy Galloway), Queen's Medical Centre (George Bugg, Yvette Davies, Carys Smith, Gill Perkins), Rotunda Infirmary (Mike Geary, Fiona Walsh, Ursula Nagle), Royal Blackburn Hospital (Elizabeth Martindale, Hilary Jackson, Louise O'Malley), Majestic Bolton Hospital (Narmada Katakam, Heather White, Emma Tanton), Royal Derby Hospital (Rosie Hamilton, Hilary Glanowski, Ethel Forde), Southern General Infirmary (Alan Mathers, Christina MacDonald, Lorna McKay), St Mary's Infirmary (Leroy Edoziern, Paula Doran, Julie Dillon, Cara Taylor, Paula Evans), Stoke Mandeville Infirmary (Veronica Miller, Christopher Wayne, Julie Tebbutt, Ellie Hendy), University College London Hospital (Patrick O'Brien, Seni Subair, Helen Dent, Camille Mallet), University Hospital of Coventry and Warwickshire (Siobhan Quenby, Jane Hillen), University Hospital of North Staffordshire (Peter Young, Tracey Harrison, Louise Wood), Warrington Infirmary (Rita Arya, Lindsay Roughley), Warwick Hospital (Olanrewaju Sorinola, Carole Rogers, Janet Phipps).
Trial steering committee
Bob Arndtz, Denis Azzopardi, Zoe Chivers, Andy Cole, Max Parmar (chair until November, 2015), Tracy Roberts, Julia Sanders, Derek Tuffnell (chair from November, 2015).
Information-monitoring committee
Deborah Ashby (chair), Jane Norman, Andy Shennan, Helen Spiby, Win Tin.
Announcement of interests
KG is a founding shareholder of K2 Medical Systems, and was clinical director for development of the Babe and Guardian systems until 2005. CM is an employee and shareholder of K2 Medical Systems. All other authors declare no competing interests.
Acknowledgments
This study was funded past the United kingdom of great britain and northern ireland National Institute for Health Research (NIHR) Health Technology Cess plan (project number 06.38.01). SK was partly funded by the NIHR Collaboration for Leadership in Applied Health Enquiry and Intendance West Midlands. The quality-of-care review panel consisted of an obstetrician (PS), neonatologist (Ilya Kovar), and midwife (Christina McKenzie) for all sessions except 1, when CM was replaced by Mary Griffin. Previous funding from the Uk Medical Research Council enabled the initial development of Guardian and the Infant software. The views and opinions expressed past authors in this publication are those of the authors and do not reflect those of the UK NHS, NIHR, the United kingdom of great britain and northern ireland Medical Research Quango, Central Commissioning Facility, NIHR Evaluation, Trials and Studies Coordinating Heart, the Health Technology Cess program, or the Section of Health.
Supplementary Material
References
- 1.
Continuous cardiotocography (CTG) equally a form of electronic fetal monitoring (EFM) for fetal assessment during labour.
Cochrane Database Syst Rev. 2013; 5 () - 2.
Birth asphyxia and the intrapartum cardiotocograph.
Br J Obstet Gynaecol. 1990; 97 : 470-479 - 3.
Obstetric accidents: a review of 64 cases.
BMJ. 1990; 300 : 1365-1367 - 4.
Case-control study of intrapartum intendance, cerebral palsy, and perinatal death.
BMJ. 1994; 308 : 743-750 - 5.
K2MS Guardian: full electronic capture of patient information during childbirth.
- half dozen.
A multicentre comparative study of 17 experts and an intelligent calculator system for managing labour using the cardiotocogram.
Br J Obstet Gynaecol. 1995; 102 : 688-700 - 7.
Skinner JF, Harris One thousand, Greene KR. Computerised decision support for managing labour using the cardiotocogram: 500 cases with the range of abnormality. 28th British Congress of Obstetrics and Gynaecology; Harrogate, UK; June 30–July 4, 1998.
- 8.
Development, evaluation and validation of an intelligent decision support tool for the management of labour.
in: Chang AMZ Rogers MS Assay of complex data in obstetrics. Bailliere Tindall, London 1994: 583-605 - 9.
Fourth dimension to optimise and enforce training in interpretation of intrapartum cardiotocograph.
BJOG. 2016; 123 : 866-869 - 10. Intrapartum care for healthy women and babies. RCOG Press, London 2014
- 11.
A study of an intelligent system to support conclusion making in the management of labour using the cardiotocograph—the Babe study protocol.
BMC Pregnancy Childbirth. 2016; 16 : 10 - 12. Perinatal mortality surveillance, 2004: England, Wales and Northern Ireland. Confidential Enquiry into Maternal and Child Health, London 2006
- 13.
Validation of a parent report measure of cognitive development in very preterm infants.
Dev Med Kid Neurol. 2004; 46 : 389-397 - xiv.
Developmental assessment of preterm infants at two years: validity of parent reports.
Dev Med Child Neurol. 2008; 50 : 58-62 - 15.
Urgency of caesarean section: a new classification.
J Roy Soc Med. 2000; 93 : 346-350 - 16.
Report to the Patient Condom Enquiry Programme (Policy Research Programme of the Department of Health). Monitoring the incidence of neonatal encephalopathy—what next?.
- 17. The Neonatal Survey and Yorkshire Neonatal Network Report 2005. University of Leicester, Leceister 2005
- eighteen.
Plymouth randomised trial of CTG only versus ST waveform plus CTG for intrapartum monitoring; 2400 cases.
Am J Obstet Gynecol. 1993; 169 : 1151-1160 - nineteen.
Umbilical cord blood gas analysis at delivery; a time for quality data.
Br J Obstet Gynaecol. 1994; 101 : 1054-1063 - 20.
Pathologic fetal acidemia.
Obstet Gynecol. 1991; 78 : 1103-1107 - 21.
Percentiles for gas values in human umbilical cord claret.
Eur J Obstet Gynecol Reprod Biol. 1983; 14 : 341-346 - 22.
Relative take chances estimation in randomized controlled trials: a comparing of methods for independent observations.
Int J Biostat. 2011; vii : 5 - 23.
Follow-upwardly studies: a example for a standard minimum data set.
Arch Dis Child. 1997; 76 : F61-F63 - 24. Disability and perinatal care: measurement of health status at ii years. A report of ii working groups convened past the National Perinatal Epidemiology Unit and the former Oxford Regional Health Authority. National Perinatal Epidemiology Unit, Oxford 1994
- 25.
When (not) to finish a clinical trial for do good.
JAMA. 2005; 294 : 2228-2230 - 26.
Severe asphyxia due to delivery-related malpractice in Sweden 1990–2005.
BJOG. 2008; 115 : 316-323 - 27.
Central fetal monitoring with and without computer analysis. A randomised controlled trial.
Am J Obstet Gyecol. 2017; 129 : 83-ninety - 28.
Indirect standard cardiotocography plus fetal claret sampling versus indirect quantitative cardiotocography— a randomised comparative report in intrapartum monitoring.
Akusherstvo I Ginekologia. 2012; 51 : iii-10 - 29.
Composite outcomes in randomized trials: greater precision but with greater incertitude?.
JAMA. 2003; 289 : 2554-2559 - 30.
The main outcome fails—what side by side?.
N Engl J Med. 2016; 375 : 861-970
Article Info
Publication History
Published: March 21, 2017
Identification
DOI: https://doi.org/10.1016/S0140-6736(17)30568-8
Copyright
© 2017 The Author(south). Published by Elsevier Ltd.
User License
Creative Commons Attribution (CC BY 4.0) |ScienceDirect
Access this article on ScienceDirect- View Large Image
- Download Hi-res image
Linked Manufactures
- Computerised cardiotocography—study design hampers findings
-
Cardiotocography is a diagnostic tool, not a therapeutic modality. Diagnostic tests are unlikely to improve outcomes unless followed past specific and effective therapeutic interventions. When the management of highly specific abnormalities is not specified in a written report protocol but is instead left to the individual discretion of many providers with very different management approaches, the chance that all of them will react in the same way is small.
- Full-Text
-
- The INFANT report—a flawed design foreseen
-
During the INFANT Study1 (March 21) awarding process in 2006–07 and since, concerns were formally raised regarding study design weaknesses past myself and some of the clinical investigators, clinical collaborators, and the National Establish for Wellness Research (NIHR) referees.2 The source of the concerns were two-fold.
- Full-Text
-
- The Babe trial
-
Robert Keith1 (April 28, p 1697) suggests that the Infant trial2 is flawed past design. Previous reports3,4 suggested that a key element in substandard intrapartum care is the failure of clinicians to recognise an abnormal fetal heart-charge per unit pattern, so the trial was prepare to investigate whether decision support that detects and highlights abnormality of the fetal centre charge per unit could meliorate outcomes. Nosotros compared decision support with no decision support. All other aspects of care were kept abiding, including the utilise of the Guardian platform, which is the electronic data collection system in which the decision-support software operates.
- Full-Text
-
Source: https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)30568-8/fulltext
0 Response to "Why Does a Baby's Heart Rate Drop During Labor"
Post a Comment