Article Text

Mortality of babies enrolled in a community-based support programme: CONI PLUS (Care of Next Infant Plus)
  1. Alison J Waite1,
  2. Robert C Coombs2,
  3. Angela McKenzie3,
  4. Charlotte Daman-Willems4,
  5. Marta C Cohen5,
  6. Michael J Campbell6,
  7. Robert G Carpenter3
  1. 1Academic Unit of Child Health, University of Sheffield, Sheffield Children's Hospital, Sheffield, UK
  2. 2Sheffield Teaching Hospitals, Sheffield, UK
  3. 3Department of Medical Statistics, London School of Hygiene and Tropical Medicine, UK
  4. 4University Hospital Lewisham, London, UK
  5. 5Department of Histopathology, Sheffield Children's Hospital, Sheffield, UK
  6. 6ScHARR, University of Sheffield, Sheffield, UK
  1. Correspondence to Alison J Waite, Room C1, Stephenson Wing, Academic Unit of Child Health, University of Sheffield, Sheffield Children's Hospital, Western Bank, Sheffield S10 2TH, UK; coni{at}sheffield.ac.uk

Abstract

Objective To report mortality in babies enrolled on a community-based programme, Care of Next Infant Plus (CONI PLUS), which primarily supports parents anxious because of previous sudden unexpected death in infancy (SUDI) in their extended family or following an apparent life threatening event (ALTE) in their baby.

Design Prospective observational study from 1996 to 2010 in the UK.

Results Of 6487 babies enrolled, 37 died (5.7 per 1000). There were 2789 (43.0%) SUDI related babies of whom, six died suddenly and unexpectedly (2.15 per 1000). Four babies were sharing a sofa at night or a bed with parent(s) who smoked or had consumed alcohol. Of the 1882 (29.0%) babies enrolled following an ALTE, five died suddenly and unexpectedly (2.66 per 1000): four unexplained and one due to infection. None occurred while sharing a sleep surface, and at least three died during the day. The remaining 1816 (28%) babies were enrolled for other reasons. Seven died suddenly and unexpectedly (3.85 per 1000), two were unexplained and none associated with bed sharing.

Conclusions The number of SUDI deaths in babies enrolled on CONI PLUS is higher than expected from UK averages. Deaths in babies enrolled because of family history of SUDI were mostly associated with inappropriate sharing of a sleep surface at night and mostly outside the peak age range for sudden infant death. The opposite is true for those enrolled following an ALTE. The number of deaths is small but findings suggest a different mechanism for death in these two groups.

  • Comm Child Health
  • Mortality
  • SIDS

Statistics from Altmetric.com

What is already known on this topic?

  • The Care of Next Infant programme is widely used in the UK and highly valued by parents.

  • Sharing a sleep surface with a parent who smokes is associated with an increased risk for sudden infant death syndrome (SIDS).

  • There are wide differences in the reported mortality in babies following an apparent life threatening event (ALTE).

What this study adds?

  • Deaths in babies enrolled because of sudden unexpected death in infancy in a relative showed avoidable risk factors mostly associated with smoking and sleep place.

  • This is the largest follow-up study of infants following an ALTE.

  • Mortality from SIDS following idiopathic ALTE was 4/1883 (2.13 per 1000) and was not associated with bed sharing.

Introduction

The Care of Next Infant (CONI) programme1 is widely used in the UK to support families with children born following a sudden unexpected death in infancy (SUDI), that is, a death that is unexpected in the previous 24 h.2 CONI addresses the anxieties bereaved parents have about subsequent children3 and recognises that these children are at increased risk of death.4–6 Mortality on CONI has been reported previously.7 CONI offers intensive health visiting support and tools to assess infant health. The success of CONI prompted clinicians to suggest the programme be extended to support other groups of anxious parents. CONI PLUS includes babies with:

  • SUDI in the extended family (aunt/uncle/cousin)

  • previous sibling infant death (not SUDI)

  • following an apparent life threatening event (ALTE).8 Full definitions for SUDI, ALTE and SIDS are available in the online supplementary appendices.

In practice, babies have been enrolled for additional reasons and selection is left to the local team. CONI PLUS is offered until the baby is 6 months or for 6 months following an ALTE and includes:

  • Regular health visitor (usually weekly) home visits

  • Movement (apnoea) monitor

  • Symptom diary

  • Sheffield weight chart (enables weight gain assessment over 2- and 8-week periods)9

  • Room thermometer

  • Baby check (19 signs and symptoms scored to determine severity of illness)10

  • Training in resuscitation.

In most areas, parents are offered direct access to a paediatrician when their child is unwell.

CONI PLUS was set up in 1996 and by 31 December 2010 had recruited 6487 babies from 175 centres. This report describes the population enrolled in CONI PLUS and deaths under 12 months.

Method

The babies are registered with the national CONI office in Sheffield. Registration details include: sex, mode of delivery, birth weight, gestation, initial feeding method, age of mother, age of mother at first live birth, parity, smoking in parents, parents’ ethnic origin, mother's marital status, father's employment status and reason for enrolment.

Infants recruited on to CONI PLUS have been categorised into nine groups (table 1), but this report concentrates on the two largest: Group 1 (SUDI in extended family) and Group 2 (Following ALTE).

Table 1

Reason for enrolment and mortality under 1 year

Deaths of CONI PLUS babies under 1-year-old are notified to the CONI office by the local coordinator. Information on the cause of death is obtained from the local paediatrician. SUDI is subject to a Coroner's enquiry and, since 2008 in England, a review by Child Death Overview Panels.2 Following these enquiries and when deaths are unexplained, the parents are offered a home visit by a paediatrician from the CONI steering group. A detailed family and clinical history is taken and further information collected by interview with their health visitor and general practitioner. The postmortem report and available sections and slides are reviewed by an experienced paediatric pathologist.

Ethics

We are reporting a long-standing service evaluation. The protocol for the follow-up of families following a death was approved by South Sheffield Ethics committee in 1994 and regularly reviewed. In 2009, the National Research Ethics Service advised that ethical approval was not required and this was accepted by the local committee. All families signed a consent form to give permission for registration data to be shared with CONI. Following a death, parents are approached by the local team prior to our home visit and a further consent obtained to give permission to share information about their baby's death.

Statistical methods

The χ2 test was used to compare proportions between the groups and the F test to compare means. We used an overall mortality rate of 5.2/1000 live births, a mortality rate for sudden infant death syndrome (SIDS) and unascertained of 0.45/1000 live births and a SUDI rate of 0.90/1000 live births as averages over the time period covered from ONS data for England and Wales to get an approximate expected number of deaths in each group and compute the standardised mortality ratio (SMR).11 ,12 A more exact method would be to apply the prevailing rate for each year, but SUDI rates were not available before 2004, and the average over the years available gives a reasonably robust estimate of the annual rate over the study period since the rates do not vary greatly from year to year. The CIs for the SMR were found using an exact method in Open Epi.13

Results

Table 1 shows the reason for enrolment in CONI PLUS and the number of deaths.

Group 1: history of SUDI in the extended family

A total of 2789 babies were enrolled from 2374 families with a history of SUDI in the extended family; 311 (13%) families used the programme at least twice. Data on the relationship to the baby who died were missing for 15 cases and more distant than aunt/uncle/cousin in 252 cases (number of deaths in family ranged from 1 to 8). Of the remaining 2107 families, 1868 (89%) had a family history of one SUDI in aunt, uncle or cousin and 239 (11%) two or more SUDI including at least one aunt, uncle or cousin.

Group 2: Following ALTE

A total of 1882 babies from 1864 families were enrolled due to ALTE before their first birthday but after discharge from hospital following birth.

Data on hospital admission following the ALTE were available for 1801 (96%): 1587 (88%) babies were admitted, 179 (10%) were not admitted and 35 (2%) were already in hospital for other reasons prior to the event.

The attributed cause for the ALTE was provided by the local team. These have been classified into subgroups (table 2). In 56% of babies, a possible diagnosis was identified while in 44% of babies, no explanation for the ALTE was offered. Terms such as ‘apnoea’, ‘cyanosis’ or ‘floppy’ are included in the unexplained group. The exact age at first ALTE is known for 1643 (87%) of the babies. The mean age was 7.6 weeks and median 5.9 weeks (corrected for gestational age 4.1 and 2.9 weeks); 36% of ALTE babies were born at less than 37 weeks gestation and 30% had a birth weight less than 2500 g. Babies in the infection group had the youngest postconceptual age at ALTE (mean 42.9 weeks) and the lowest birth weights (mean 2501 g).

Table 2

Attributed cause for apparent life threatening event

Groups 3–9: various reasons for enrolment

At total of 543 babies from 468 families were enrolled in Group 3 (Following a sibling death other than SUDI). Smaller numbers of babies were enrolled due to prematurity, ALTE in sibling, ill at birth, congenital anomaly, poor obstetric history or for multiple reasons (Groups 4–9, see table 1).

Population data

Characteristics of the parents and babies are available in the online supplementary appendix tables A1 and A2. In Group 1 (SUDI in extended family), 22% of the mothers were <20 years at the birth of the CONI PLUS baby compared with 10% in Group 2 (Following ALTE) (p<0.01). For 56% in Group 1, this was their first child compared with 36% in Group 2 (p<0.001). Groups 1 and 2 had the highest levels of maternal and paternal smoking: compared with the other groups: 43% and 51% in Group 1 and 36% and 44% in Group 2. Group 1 had the highest proportion of maternal and paternal White British ethnicity (94% and 92%), fewest married mothers (26%) and lowest levels of employment in mother's partner (73%). Group 1 babies were also more likely to be delivered normally (69%), singleton (96%), have birth weights ≥2500 g (88%) and be born after 37 weeks gestation (89%). This group had the lowest proportion of initially breast fed babies (40%).

Mortality in CONI PLUS babies

A total of 37 babies died in their first year (table 3). The expected number of deaths based on national rates is about 33.7, showing a slightly raised mortality, but which is not statistically significant (SMR 1.1, 95% CI 0.78 to 1.50). Of 11 families offered a home visit by a CONI paediatrician following a SUDI, six accepted (three in Group1, three in Group 2).

Table 3

Deaths aged less than 1 year

Nineteen babies died following deterioration in a pre-existing condition. Eighteen deaths presented as SUDI: seven attributed to SIDS, four unascertained, four to infections, one to upper respiratory tract obstruction, one to hypoxic ischaemic encephalopathy and one to congenital anomaly and infection. From national death rates over this period, we would only have expected about six SUDIs (SMR 3, 95% CI 1.8 to 4.6) and three SIDS (SMR 2.3, 95% CI 1.0 to 4.6).

Mortality in Group 1 (SUDI in extended family)

All six deaths in Group 1 presented as SUDI. One was attributed to pneumonitis at 3 weeks. Of the remaining five deaths attributed to SIDS or unascertained, one was aged 3 weeks, one aged 8 weeks and three aged over 26 weeks. We would have expected about 1.25 SIDS or unascertained in this cohort (SMR 4.0, 95% CI 1.47 to 8.87).

Mortality in Group 2 (Following ALTE)

Of 14 deaths in Group 2, nine were due to pre-existing conditions and five presented as SUDI. One SUDI was attributed to systemic cytomegalovirus infection in a baby with residual aortic coarctation following corrective surgery. Three deaths were attributed to SIDS and for these no cause for the ALTE was given. One death was unascertained with the ALTE attributed to convulsions. No SUDI occurred in the babies whose ALTE was attributed to gastro-oesophageal reflux, feeding related problems or infection. We would have expected about 0.85 SIDS or unascertained in this cohort (SMR 5.89, 95% CI 2.16 to 13.0).

Mortality in Groups 3–9

Of 17 deaths, 10 were in babies with pre-existing conditions with poor outlook and seven presented as SUDI including two unascertained.

Circumstances and risk factors present for SUDI, n=18

Details of the circumstances and demographics for the 18 babies who died suddenly and unexpectedly are shown in table 4. At least five of the six deaths in Group 1 (SUDI in extended family) are known to have occurred in babies sharing a sleep surface at night: two sleeping on a settee, two sharing a bed with a parent who smoked and had taken alcohol or cannabis the previous evening and one sleeping in a playpen at night with a twin sibling.

Table 4

Sudden unexpected deaths: cause of death, place of death and other factors

None of the five SUDI in Group 2 (Following ALTE) babies occurred while sharing a sleeping surface. Three of the deaths occurred during the daytime, two in ‘bouncy’ chairs and one sleeping alone on a settee.

Overall, 11 of the 18 (61%) mothers smoked cigarettes and 9 (56%) of the mother's partners were known to smoke. Of the 11 babies classified as SIDS or unascertained, five were aged outside the peak age at risk for SIDS (one aged under 4 weeks and four aged over 26 weeks) and four of these five were in Group 1.

Two babies in Group 2 and one baby in Group 4 (Prematurity) had a monitor in use when the baby was found collapsed but resuscitation was unsuccessful. It appears one mother slept through the alarm or the alarm did not ring. Of the remaining 15 babies, in two monitor use had stopped, in six the monitor was not in use and in seven babies use is not known (one was bed sharing and four died during the daytime).

There were no known cases of infanticide/homicide.

Discussion

This report documents the outcome of over 6000 babies enrolled on the CONI PLUS programme since 1996. Local clinicians decide which families are offered CONI PLUS, and hence parents with a close family history of SUDI, but who are not anxious and who can implement advice on reducing the risk of SIDS, may not be enrolled. Overall, the number of deaths was small, but slightly greater than expected. However, our data suggest the population has a higher than average risk for SIDS.

We do not have a control group but when comparing population characteristics we find Group 1 (SUDI in extended family) and Group 2 (Following ALTE) differ from each other in a number of ways and also from average UK statistics. Smoking in mothers in Group 1 (43%) and Group 2 (36%), and in her partner (51% and 44%) is higher than the average 30% reported for women and 34% for men for 20–34 year olds in the UK between 1998 and 2010.14 Breast feeding at 1 week is low at 40% in Group 1 and 43% in Group 2 compared with average UK figures of 55% in 2005 and 69% in 2010.15 Unemployment is higher than the UK rate which varied between 5% and 8% in 1996–2010, especially in Group 1 (SUDI in extended family) where it was 27%.16 Smoking, bottle feeding and unemployment are all associated with an increased risk for SIDS.17

Eighteen SUDI occurred and while numbers are small, the circumstances of these deaths are notably different across the enrolment groups. Of six SUDI in Group 1 (SUDI in extended family), five were at night while sharing a sleep surface, four with a parent plus known contra-indications (settee or bed plus alcohol/smoking/drugs). One was sleeping in a playpen with a twin sibling. There are many case-control studies that show an associated risk for SIDS when babies bed share with parents who smoke or have consumed >2 units of alcohol,18–20 and cosleeping on a settee is particularly high risk.17

In contrast, of five SUDI in the Group 2 (Following ALTE), three died during the day, none were sharing a sleep surface and two were in ‘bouncy chairs’. Occlusion of the airways in a ‘sitting’ position due to flexion of the head has been suggested as a trigger for ALTE.21 Among babies enrolled for other reasons (Groups 3–9), there were seven SUDI, three died or collapsed during the day and none died while bed sharing. SIDS and unascertained deaths in Group 1 appear largely associated with recognised high risk sleeping behaviours whereas these are notably absent in the other groups.

When Group 2 (Following ALTE) deaths are considered by the diagnoses made following the ALTE, no deaths occurred in 682 (36%) babies whose ALTE was attributed to gastro-oesophageal reflux or infection while three deaths (all SIDS) occurred in 835 (56%) babies with unexplained ALTE (3.6 per 1000, SMR 7.99, 95% CI 2.03 to.21.76).

With over 1800 infants, this is the largest study of infants following an ALTE. The babies are cared for by their local paediatric services and details of investigations were not collected. Almost certainly there is variation in the level of investigation and reasoning behind the clinicians’ decision to offer CONI PLUS; however, age at ALTE in our cohort is similar to other published studies.22 There is no consensus on the mortality of babies given the diagnosis of ALTE. Bonkowsky et al allocated a diagnosis of ALTE to infants fulfilling the standard American criteria.23 In their cohort of 471 infants admitted over 4 years to a single hospital, two subsequently died, both older than 1 year and both with apparent developmental and respiratory problems. McGovern and Smith in a systematic review found only two sudden deaths in 355 cases (5.6 per 1000) from eight studies.22 A prospective study from West Virginia, USA, reports 19 SIDS in 765 cases (24.8 per 1000).24

Two European studies report an incidence for ALTE of 0.46 per 1000 live births in Sweden,25 and 0.58 per 100026 in the Netherlands. If these figures are representative for the UK, over 14 years CONI PLUS has probably recruited a third of all the UK ALTE babies. In this group, there have only been five sudden and unexpected deaths (5 per 1882, 2.7 per 1000) suggesting this is a rare event.

There are limitations to this study. We do not have complete ascertainment because enrolment is determined by the local teams and there is wide variation in uptake. Reporting of deaths is voluntary and may be incomplete. Although mortality was higher than expected from national rates, it was not markedly higher given the number of risk factors present in this group. The lack of a control group prohibits the comparison of mortality in similar populations not offered CONI PLUS.

However, some implications for public health are clear. In Group 1 (SUDI in extended family) and Group 2 (Following ALTE), 45% of deaths were sudden and unexplained. For the deaths registered as SIDS or unascertained, there are notable differences in age, place and time of death between those babies enrolled due to a family history of SUDI and those enrolled following an ALTE or for other reasons. Some of the former are dying in inappropriate sleep locations and this suggests they are potentially preventable. Emphasis needs to be placed on ensuring a safe sleep environment for all babies and particularly in families with the risk factors of smoking and alcohol. That smoking is a major risk factor for both ALTE27 and SIDS underlines the need to discourage maternal smoking in pregnancy. Our finding that SIDS following ALTE is not associated with night-time or bed sharing needs further investigation and, although the numbers are small, possibly suggests a different mechanism for death.

Acknowledgments

We would like to thank the families and health professionals involved with CONI PLUS.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors AJW designed the study, collected and collated data and wrote the first draft. RCC and CD-W designed the study and collected clinical data. AMcK collated data and performed the statistical analysis. MCC analysed and interpreted data. MJC analysed data. RGC designed the study and analysed data. All authors contributed to the interpretation of the data and revised the draft paper.

  • Funding Alison Waite and Angela McKenzie are funded by the Lullaby Trust, grant no 98.

  • Competing interests None.

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

  • Ethics approval South Sheffield Ethics Committee.

  • Patient consent Parental consent obtained.

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.