Suicide is considered the 15th commonest cause of death worldwide. The understanding of suicide risk factors can help detect people at high risk and hence potentially prevent suicide. Despite the recognition of a wide range of risk factors for suicide, the current prediction of suicide remains poor. To help improve suicide prediction, it may be important to examine novel risk factors along with understanding the combined effect of multiple risk factors. Most suicide victims have visited healthcare facilities in their final year which emphasises the importance of these facilities for the detection and potential intervention of suicide if one can recognise the indicators of risk. We believe that the examination of the wealth of healthcare data in CPRD over time linked with other databases like Hospital Episode Statistics, Index of Multiple Deprivation and Office for National Statistics can provide further insight in the understanding of suicide risk indicators.
The overall purpose of this study is to find opportunities for suicide prevention. We will therefore look specifically at several potential risk factors which may be recorded in primary or secondary health data, including:
-Demographic characteristics including ethnicity.
-The pattern of and reason for attendance
-Physical health conditions including self-harm.
-Dementia diagnosis and anti-dementia drugs
-Painkilling medications, including those not related to morphine
We will integrate recognised risk factors and any novel insights obtained from these analyses to attempt to predict suicide more accurately.
Understanding of suicide risk factors can play a central role in tailoring and the targeting of effective suicide prevention strategies. However, currently recognised risk factors of suicide are far from exhaustive and have not been empirically tested together which can partly explain the poor prediction of suicide. Healthcare registries can offer a valuable source of information for studying suicide risk, and since many suicide victims visited healthcare facilities in their final year, they can be a potential place for recognising risk and for intervening. The aim of this study is to consider several potential risk factors, singly and collectively, to predict suicide risk from health data.
We will examine several potential risk factors which may be recorded in primary or secondary health data, including:
-Demographic characteristics including ethnicity.
-The pattern of and reason for attendance
-Physical health conditions including self-harm.
-Dementia diagnosis and anti-dementia drugs
-Analgesic prescribing, including nonopioids medications
As primary and secondary care differ significantly in terms of access, service provision and nature and severity of health conditions managed, we will use the integrated CPRD (for primary care) and Hospital Episode Statistics (for secondary care) databases to help achieve a more holistic understanding of suicide risk factors. For assessing socioeconomic status, linkage with patient Index of Multiple Deprivation will be used. Moreover, we will use linked Office for National Statistics as this improves the validity of identification of suicide death.
We will conduct a population-based case-control study for suicide death. 40 controls will be risk-set sampled for each case. We propose developing (and internally validating) two main types of multivariable regression models and fitting random effects models at the general practice level as appropriate. One type involves examining each main risk factor allowing for confounders and the second involves combining suicide risk factors to predict suicide more accurately.
Suicide death
Timothy Card - Chief Investigator - University of Nottingham
Danah Alothman - Corresponding Applicant - University of Nottingham
Andrew Fogarty - Collaborator - University of Nottingham
Sarah Lewis - Collaborator - University of Nottingham
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation