First Annual Report of Child Deaths in
Greater Manchester,2012/13
introduction
This is the first report which reviews data regarding all child deaths in Greater Manchester (GM). There are four Child Death Overview Panels (CDOPs) within GM, each of which collects information and reviews the deaths of all children aged up to one day under18 years (excluding stillbirths and planned terminations of pregnancy carried out within the law), who normally reside within the geographical boundaries of that CDOP. Each CDOP produces an annual report which summarises the data collected that year, reviews any variations or emerging trends, and makes recommendations for future interventions and data collection. Whilst these highlight patterns and trends within each CDOP, detailed analysis and conclusions are often limited by small numbers, with around 50 to 80 deaths per CDOP per year. Compilation of data from the four CDOPs into a single GM dataset provides larger numbers enabling a greater depth of analysis.
Because this it the first year in which a complete dataset for the whole of GM has been brought together, it is not possible to look at yearly trends. However, now that there is a system in place for this data to be amalgamated, future GM reports will, in time, allow for analysis of trends as they emerge.
This report has two primary aims. The first is to review the data relating to child deaths in GM so that observations about causes and potentially modifiable factors can be made. The second is to review some of the variations in the approaches of the four CDOPs so that greater consistency can be achieved in the future. Whilst they share a common template for collection of data surrounding child deaths, some of the elements of the data collected vary between CDOPs.
Contents
1. Child death overview panels in Greater Manchester / 3
2. Reviews by CDOPs / 5
3. Geographical variations in number of deaths / 8
4. Causes of death / 9
5. Expected versus unexpected deaths / 12
6. Modifiable factors / 13
7. Socio-demographic characteristics / 14
7.1 Sex / 14
7.2 Ethnicity / 15
7.3 Deprivation / 16
8. Neonatal and infant deaths / 18
8.1 Impact of socio-demographic characteristics on infant deaths / 19
8.2 Risk factors identified in deaths of neonates and infants / 20
8.3 Sudden and unexplained deaths in a child (SUDC) / 23
8.4 Day of birth of neonates / 24
9. Consanguineous relationships / 26
10. Mental health of parents of carers / 26
11. Suicide of deliberate self-harm / 26
12. Road traffic collisions / 27
13. Key points / 28
1. Child Death Overview panels in Greater Manchester
Table 1 shows the geographical areas covered by each CDOP and their population sizes.
CDOP / Population sizeSalford, Bolton & Wigan / 828,568
Salford / 233,933
Bolton / 276,786
Wigan / 317,849
Tameside, Trafford & Stockport / 729,177
Tameside / 219,324
Trafford / 226,578
Stockport / 283,275
Bury, Rochdale & Oldham / 621,656
Bury / 185,060
Rochdale / 211,699
Oldham / 224,897
Manchester / 503,127
The socio-economic profiles of the areas are summarised in Table 2.
1
White/ WhiteBritish / Asian/ Asian British / Black/ Black
British / Mixed/ Other / Male Life Expectancy at birth / Female Life Expectancy at birth / Deprivation (% population living most deprived quintile in England)
Salford / 90% / 4% / 3% / 3% / 70.8 / 76.5 / 46%
Bolton / 82% / 14% / 2% / 2% / 72.1 / 77.3 / 40%
Wigan / 97% / 1% / 1% / 1% / 72.4 / 77.4 / 30%
Tameside / 91% / 7% / 1% / 2% / 72.2 / 77.9 / 36%
Trafford / 86% / 8% / 3% / 4% / 73.9 / 78.5 / 11%
Stockport / 92% / 5% / 1% / 2% / 73.5 / 78.9 / 12%
Bury / 89% / 7% / 1% / 3% / 73.0 / 78.4 / 19%
Rochdale / 82% / 15% / 1% / 2% / 71.4 / 76.9 / 44%
Oldham / 78% / 19% / 1% / 2% / 71.2 / 77.0 / 41%
Manchester / 67% / 17% / 9% / 8% / 69.6 / 76.4 / 64%
GM / 84% / 10% / 3% / 3% / Not available / Not available / 37%
1
2. Reviews by CDOPs
The four CDOPs in GM were notified of 239 child deaths between 1st April 2012 and 31st March 2013. In the same year, the panels completed reviews of 269 child deaths.
A review is completed when enough information about the death is available to enable a review by the panel and this has taken place. Collection of background information can often be a lengthy process, particularly if the nature of the death means that an inquest into its cause is required. There is some variation in the information which it is deemed necessary to have before a panel will review a death. This, together with the varying nature of deaths in each CDOP, means that the average length of time between a death and a review differs between CDOPs
(Table 3).
CDOP / Mean length of time from death to review (days) / Maximum length of time from death to review (days)Manchester / 192 / 651
Salford, Bolton & Wigan / 249 / 1394
Bury, Rochdale & Oldham / 310 / 1104
Tameside, Trafford & Stockport / 319 / 846
The overall average length of time between death and review for all deaths reviewed in the year 2012/13 was 267 days. As well as variations between CDOPs, the length of time for a review differs according to cause of death (Table 4).
Malignancy / 162 / 420
Chromosomal, genetic and congenital anomalies / 197 / 801
Perinatal/neonatal event / 229 / 1394
Chronic medical condition / 332 / 846
Acute medical or surgical condition / 339 / 763
Sudden unexpected, unexplained death / 339 / 881
Infection / 370 / 930
Deliberately inflicted injury, abuse or neglect / 436 / 650
Trauma and other external factors / 437 / 796
Suicide or deliberate self-harm / 506 / 822
The period of time between a death and its review createsan issue when deciding which deaths to include in this report. There are three possible ways in which eligibility for inclusion in this report could be determined.
- Date of death within the year 2012/13
- Date that death was reported to CDOPs within the year 2012/13
- Completion of review of a death within the year 2012/13.
There are drawbacks for each of these options. If either of the first two were selected, there would be deaths included in the review for which key data was missing, such as categorisation of cause of death and the presence of modifiable factors, which are determined at the panel review. This would limit the analysis. An additional and serious issue with the first option is that there is a potential for deaths which are reported to a CDOP a long time after the death occurred being missed from inclusion in a report. There are occasions where cases are brought to the attention of CDOPs many months after the date of death. In this instance, they may be too late for inclusion in the appropriate report but would not be included in a subsequent report because the date of death would lie in the previous year. Finally the third option has the limitation that reviews are often completed two or three years after the death has occurred. Therefore annual trends will be less apparent because deaths included in the report will span a number of years, and yearly rates cannot be calculated. However, the latter option does ensure that no deaths are missed and that data collection has been completed by the time a death is included in a report. In addition, this option enables discussion about delays in reviews and missing data so that greater consistency can be achieve across CDOPs.
This is a report of all the child deaths for which a review was completed during the period 1st April 2012 to 31st March 2013. A review has not been completed for 113 of the deaths notified in 2012/13 and these are therefore not included in this report.
Table 5 shows the number of deaths outstanding from each year for which a review was completed in the year 2012/13 and are thus included in this report.
Year death notified / 2008/9 / 2009/10 / 2010/11 / 2011/12 / 2012/13 / TotalSalford, Bolton & Wigan / 1 / 2 / 7 / 33 / 44 / 87
Bury, Rochdale & Oldham / 1 / 12 / 30 / 28 / 71
Manchester / 25 / 31 / 56
Tameside, Trafford & Stockport / 10 / 22 / 23 / 55
Total / 1 / 3 / 29 / 110 / 126 / 269
3. Geographical Variations in Number of Deaths
Table 6 shows the number of deaths included in this report for each area. Areas are listed in order of their population size.
Area / Number of deaths reviewed in 2012/13Manchester / 56
Wigan / 18
Stockport / 19
Bolton / 42
Salford / 27
Trafford / 16
Oldham / 27
Tameside / 20
Rochdale / 25
Bury / 19
Most striking is the large number of deaths in Bolton, relative to its population size. Whilst this may reflect a larger number of recent cases in Bolton having been reviewed in a timely fashion so that a large proportion of them have been included in this report, subsequent GM reports should continue to review this. Wigan and Stockport have amongst the lowest number of child deaths, despite these areas being amongst the most populated areas. This is likely, at least in part, to be due to the lower levels of deprivation in these areas and lower numbers of ethic minority groups who experience higher levels of child deaths (see Section 7).
4. Causes of death
For each child death, panels are required to decide which category the cause of death falls into. This ensures a certain level of consistency in reporting and enables comparison acrossCDOPs. The list of categories shown in Table 7 is hierarchical. If a death falls into more than one category, it is allocated to the category highest up the ranking. However, this is not always a straightforward task and the decision ultimately lies at the discretion of each panel. In order to achieve greater consistency, the chairs of the four GM CDOPs meet regularly and discuss a small number of cases.
The categoriesand numbersof child deaths in GM are shown in Table 7 and Figure 1.
Cause of death / Number reviewed in 2012/13 (percent)1. Deliberately inflicted injury, abuse or neglect / 3 (1%)
2. Suicide or deliberate self-harm / 11 (4%)
3. Trauma and other external factors / 10 (4%)
4. Malignancy / 12 (4%)
5. Acute medical or surgical condition / 16 (6%)
6. Chronic medical condition / 11 (4%)
7. Chromosomal, genetic and congenital anomalies / 70 (26%)
8. Perinatal/neonatal event / 97 (37%)
9. Infection / 18 (7%)
10. Sudden unexpected, unexplained death / 20 (7%)
Not known / 1 (0%)
The majority of child deaths occur in early life and result from events around the time of birth or from conditions which pre-date birth such as genetic and congenital anomalies (Figure 1). This single death for which the cause was unknown occurred whilst the child was abroad and there was no information available to determine the cause of death.
Table 8 shows the number and proportion of deaths within each category according to geographical area.
1
Cause of death / Salford / Bolton / Wigan / Bury / Rochdale / Oldham / Manchester / Tameside / Trafford / Stockport / GM1. Deliberately inflicted injury, abuse or neglect / 2 (5%) / 1 (5%) / 3 (1%)
2. Suicide or deliberate self-harm / 1 (4%) / 1 (2%) / 1 (6%) / 2 (11%) / 1 (4%) / 3 (5%) / 1 (6%) / 1 (5%) / 11 (4%)
3. Trauma and other external factors / 2 (7%) / 2 (5%) / 1 (4%) / 2 (4%) / 1 (5%) / 1 (6%) / 1 (5%) / 10 (4%)
4. Malignancy / 4 (10%) / 1 (6%) / 1 (2%) / 3 (19%) / 3 (16%) / 12 (4%)
5. Acute medical or surgical condition / 1 (4%) / 2 (5%) / 1 (5%) / 1 (4%) / 3 (11%) / 4 (7%) / 2 (10%) / 2 (11%) / 16 (6%)
6. Chronic medical condition / 1 (6%) / 1 (5%) / 1 (4%) / 2 (7%) / 1 (2%) / 2 (10%) / 3 (19%) / 11 (4%)
7. Chromosomal, genetic and congenital anomalies / 6 (22%) / 14 (33%) / 2 (11%) / 2 (11%) / 9 (36%) / 5 (19%) / 16 (29%) / 5 (25%) / 6 (38%) / 5 (26%) / 70 (26%)
8. Perinatal/neonatal event / 11 (41%) / 11 (26%) / 7 (39%) / 9 (47%) / 7 (28%) / 14 (52%) / 23 (41%) / 6 (30%) / 2 (13%) / 7 (37%) / 97 (37%)
9. Infection / 2 (7%) / 1 (2%) / 4 (22%) / 3 (12%) / 2 (7%) / 3 (5%) / 3 (15%) / 18 (7%)
10. Sudden unexpected, unexplained death / 4 (15%) / 4 (10%) / 2 (11%) / 3 (16%) / 2 (8%) / 1 (4%) / 3 (5%) / 1 (5%) / 20 (7%)
Not known / 1(2%) / 1 (0%)
Total / 27 / 42 / 18 / 19 / 25 / 27 / 56 / 20 / 16 / 19 / 269
1
5. Expected versus unexpected deaths
All deaths reviewed by CDOPsare classified as either expected or unexpected. Of the 269 deaths reviewed during the year 2012/13, 157 (58%) were expected. The proportion of deaths which are expected/unexpectedvaries between age groups. The majority of deaths within the first year of life are due to either complications relating to prematurity or due to a chromosomal, genetic or congenital abnormality. Many of these deaths are likely to be expected because the health issuemay be known about before the death occurs. Conversely, in older children deaths from unexpected causes such as road traffic collisions are more likely. Data on whether the death was expected or not was missing from 12 cases, 11 of which occurred within the Bury, Rochdale and Oldham CDOP, and one within Bolton, Salford and Wigan.
6. Modifiable Factors
During review, each CDOP panel uses the evidence available to decide whether or not there were modifiable factors which may have contributed to the vulnerability or death of the child. Modifiable factors are those which, by means of locally or nationally achievable interventions, could be modified to reduce the risk of future child deaths.
Across GM, it was felt that modifiable factors may have contributed to 69 of the child deaths (26%). There is a level of discretion in terms making the decision of whether modifiable factors are present or not. Table 9 shows the proportion of deaths in each CDOP to which modifiable factors were felt to have possibly contributed.
CDOP / % of deaths in which modifiable factors were presentSalford, Bolton & Wigan / 39%
Bury, Rochdale & Oldham / 21%
Manchester / 29%
Tameside, Trafford & Stockport / 18%
There is likely to be a significant element of genuine variation in the number of deaths with modifiable factors between CDOPs, either due to differences in area demographics or because of random annual variation. However, it is worth noting that Salford, Bolton and Wigan CDOP has more than twice the number of deaths with modifiable factors than Tameside, Trafford and Stockport CDOP. This should be reviewed in subsequent GM CDOP reports to look for patterns which might suggest genuine variation or differences in approach between CDOPs. Ultimately, CDOP chairs should decide on appropriate guidance on this issueto ensure consistency across GM.
7. Socio-demographic characteristics
A summary of the demographic characteristics of child deaths in GM is given in Table 10. Of the 269 deaths reviewed, 41% were neonatal deaths (babies who died within 28 days of birth), and an additional 25% were deaths were amongst babies between 28 and 364 days old (Figure 3). 53% were male, and 46% were female (the sex of one child was indeterminate).
Characteristic / Number / Proportion of child deaths (numbers in brackets are the percent of the GM population)Age
0-27 days / 111 / 41%
28-364 days / 67 / 25%
1-4 years / 30 / 11%
5-9 years / 17 / 6%
10-14 years / 23 / 9%
15-17 years / 21 / 8%
Sex
Male / 143 / 53% (49%)
Female / 125 / 46% (51%)
Indeterminate / 1 / <1%
Ethnicity
White/White British / 168 / 62% (84%)
Asian/Asian British / 54 / 20% (10%)
Black/Black British / 20 / 7% (3%)
Mixed/Other / 23 / 9% (3%)
Not Known/Not Input / 4 / 1%
Deprivation Quintile (1 = most deprived)
1 / 120 / 45% (37%)
2 / 35 / 13% (21%)
3 / 21 / 8% (16%)
4 / 17 / 6% (14%)
5 / 10 / 4% (11%)
Data not available due to processing error / 66 / 25%
Although baseline population data for some of these factors is available for the 0-15 year-old population specifically, this was not available on a GM-level for all factors. Therefore data for the all-age GM population is presented for consistency (although it is recognised that this may not reflect the child population exactly).
7.1 Sex
Unusually, there was a slight female predominance amongst the neonatal deaths (Figure 3). This is discussed further in Section 8.
7.2 Ethnicity
The ethnic groups of child deaths are shown in Figure 4.
Whilst 63% of deaths were amongst White/White British children, this group is underrepresented compared to numbers in the population. All other ethnicities are over-represented, particularly the Mixed/Other ethnic group and Black/Black British children (Figure 5). This is seen across the four CDOPs and is consistent with findings from previous years’ CDOP reports. A number of possible explanations for this exist; however, deprivation is likely to be a significant factor.
7.3 Deprivation
There is a consistent trend across deprivation quintiles, with the number of child deaths increasing with increasing deprivation (Figure 6). Furthermore, the most deprived quintile is over-represented. In GM, 37% of the population are within the most deprived quintile, whilst 45% of the child deaths in GM are from this quintile.
Comparison of the percent of child deaths in each area which were amongst the most deprived quintile, and the percent of the population in each area living in the most deprived quintiles, it is clear the this quintile is consistently over-represented, often dramatically so. Unfortunately, due to a processing error during the transfer of information to the GM database, data on the deprivation quintile of child deaths in Bury, Rochdale and Oldhamis unavailable.
8. Neonatal and Infant Deaths
The majority (66%) of deaths reviewed occurred in the first year of life. Neonatal deaths are those which occur within the first 28 days of life, and infant deaths are all those which occur under the age of one year. The causes of deaths of children under one year are shown in Table 11 and Figure 8.
It should be noted that the large number of sudden unexpected/unexplained deaths in infants (SUDCs) span a three year period. Four of the SUDCs reviewed in 2012/13 occurred in 2010/11, whilst 12 occurred in 2011/12, and two in 2012/13. SUDCs are discussed further below.
Salford, Bolton & Wigan / Bury, Oldham & Rochdale / Manchester / Tameside, Trafford & Stockport / TotalAcute medical or surgical condition / 1 / 1
Chromosomal, genetic and congenital anomalies / 17 / 14 / 10 / 10 / 51
Chronic medical condition / 1 / 2 / 3
Deliberately inflicted injury, abuse or neglect / 1 / 1
Infection / 3 / 2 / 1 / 1 / 7
Malignancy / 1 / 1 / 2
Perinatal/neonatal event / 29 / 30 / 22 / 14 / 95
Sudden unexpected/ unexplained death / 7 / 6 / 3 / 1 / 17
Suicide or deliberate self-harm / 1 / 1
The majority of the deaths (80%) lie within two categories: chromosomal, congenital and genetic, and perinatal/neonatal event. As discussed in Section 4, the decision of which category to classify a death is not an exact science and there is likely to be some variation between panels on this issue. One of the purposes of this report is to review differences in approach of the four CDOPs so that greater consistency can be achieved in the future. Therefore a small number of the deaths which lie in the less-populated categories are discussed further below.
The single death listed as an acute medical or surgical condition was the result of a known complication of prematurityand was also linked to a congenital abnormality, which was listed as a contributing factor on the death certificate. However, the ‘perinatal/neonatal event’ category (which prematurity falls into), and the ‘congenital anomaly’ category both fall below the ‘acute medical or surgical’ category on the hierarchy. This case demonstrates the complexity of panels’ decisions about categorising deaths.
All three deaths categorised as chronic conditions also met the criteria for other categories. One was the direct result of a congenital/chromosomal anomaly, one was related to extreme prematurity, and one was due to an acuteinfection.
Given the age at death, initial review indicated that the single death listed as suicide or deliberate self-harm wasan error. Indeed, further investigation revealed that this was an inputting error and the death was actually classified as a sudden/unexplained death.