Abstract
Background There is a lack of effective scoring systems to identify populations at risk of Sudden Infant Death Syndrome (SIDS) with the published scoring systems limited by the infrequency of SIDS and reliance on retrospective data. The Sheffield Birth Score (SBS) is a risk-based score previously calculated at birth using 8 factors (such as birth weight and maternal smoking) which allocated targeted Health Visiting if the numerical total score was over 500. With SIDS rates in Sheffield above the national rate, can the SBS be expanded to improve identification of possibly preventable deaths and therefore focus intervention programmes?
Objectives The objective of the study is to assess whether the inclusion of active social care involvement as a factor can improve the predictive power of the current SBS model to better identify populations at risk of SIDS.
Methods Firstly data on birth scores from 7,321 births in Sheffield was matched with social care involvement 9 months before the birth to 6 months after. The data was then ranked in quintiles to assess summary data on the high risk vs. the low risk groups.
Secondly birth score data on a cohort of deaths classed as SIDS over a 10 year period was matched to social care involvement (n=44). The risk for SIDS was modelled using logistical regression.
Results The overall picture of births showed that 8.4% of births scored 500 or above. When matched to social care data, 51.3% of social care contacts were in the top quintile.
Looking at the cohort of 44 SIDS cases the average birth score was 485 - higher than over 80% of the total births, and 30 of the 44 cases were in quintile 5. A cut off of 500 would identify 52% of these cases. Using a lower cut off birth score of over 347 would have identified an additional 12 deaths, 5 of whom had social care involvement (41%). Using a score of over 452 (highest quintile) would have identified 30 deaths, 17 of whom had social care involvement (56%). All babies with social care involvement are 11.4 times more likely to die of SIDS compared to those with no social care involvement. A baby scoring in the lower four quintiles with social care involvement is 27.6 times more likely to die than a baby in the lower four quintiles with no social care involvement.
Conclusions The study presented confirms the higher the birth score the greater the likelihood of social care involvement. Social care as a factor is a useful predictor of SIDS, and may actually be more useful in the lower quintile birth score groups – previously thought to be low risk. The link of social care to high SBS and SIDS suggests the SBS does identify families where there is greater risk by encompassing multiple factors within that particular home and dynamic. However, due to the small numbers of cases of SIDS in the time frame reviewed it was impossible to determine whether the inclusion of social care involvement to the SBS would improve the statistical model overall.