Whitty et al (1) identified MLCs are not just linked to older citizens; socio economic factors and life events come into play. In order to treat patients effectively predictable clusters of diseases can be identified, this will inform thinking into who to better tackle management of coexisting physical and mental health problems, which once identified can be embedded into medical training and continuous professional development, including for specialists.
Our objective is to firstly estimate the prevalence of MLCs (2+,3+ considering mental and physical health combined and separately) in England, at a regional level, and by local areas within it by sex and age group.
Secondly common MLCs patterns will be identified using summary statistics and cluster analysis considering MLCs and socio-demographic factors. The completeness of weight and smoking status will be assessed for consideration for inclusion.
The study population of interest will include all active registered patients in England from the CPRD Aurum dataset.
Public Health England (PHE) will identify MLC based on the Cambridge code list (2). This list identifies 38 long-term conditions from electronic health record data.
Long term conditions will be classed as present or not based on the Cambridge code list, the list considers time frames of conditions to allow for recovery of chronic conditions such as cancer.
Using the Index of multiple deprivation (IMD), sex, age and MLC at person level, MLC prevalence estimations will be calculated by matching and scaling based on ONS mid-year population estimates of IMD, sex and age.
Patterns of common MLCs will be identified between the 38 conditions from cluster analysis.
This research will help gauge the needs of people with MLC to improve delivery of care and help develop models of care for people with complex and MLC, in primary care including access to person-centred approaches.
1) Estimation of the prevalence of multi-morbidity and physical and mental health comorbidity;
This involves the estimation of the prevalence of 2+, 3+ MLCs regardless of the type of condition captured by each cut-off
a) Sex, and area (region, Integrated Care Systems (ICS) and Local Authorities (LA))
b) Sex, age group (0-24, 25-44, 45-64, 65-84, 85+) and area (region, ICS and LA)
Please refer to section : Describe the study population in terms of key inclusions and see Expected prevalence can be calculated as follows paragraph for details on how sub regional estimates are calculated
2) Estimation of prevalence by ‘prevalence type’ ;
To provide further insight into the patterns in prevalence of multi-morbidity in (1) above, the prevalence of multiple LTCs will be estimated based on ‘prevalence type’ i.e. physical conditions only, mental health conditions only, and physical/mental health conditions only.
Estimated prevalence will be calculated by calculating the number of people with
1. Two or more physical (only) long term conditions;
2. Two or more mental health (only) long term conditions;
3. Two or more physical and mental health long term conditions (at least one in each category);
4. Three or more physical (only) long term conditions;
5. Three or more mental health (only) long term conditions;
6. Three or more physical and mental health long term conditions (at least one in each category);
c) Sex and area (region, Integrated Care Systems (ICS) and Local Authorities (LA)).
d) Sex, age group (0-24, 25-44, 45-64, 65-84, 85+) and area (region, ICS and LA).
Please refer to section : Describe the study population in terms of key inclusions and see Expected prevalence can be calculated as follows paragraph for details on how sub regional estimates are calculated.
3) Prevalence of selected conditions and combinations of conditions
This involves the estimation of prevalence of:
a) Selected conditions (hypertension, asthma, coronary heart disease, cancer, chronic obstructive pulmonary disease, stroke and mental health (other than dementia)) alone;
b) Selected conditions and the single condition they most commonly occur with;
c) Selected conditions plus additional conditions up to 4+ conditions i.e. selected conditions plus another condition, selected condition plus 2 other conditions, selected condition plus 3 conditions and selected condition plus 4 or more conditions;
d) Exploratory: Conduct an exploratory cluster analysis across all 38 conditions to identify common clusters ;
e) Exploratory: Perform multi-nominal analysis to see if the number of MLC’s can be predicted from socio-demographic and lifestyle factors;
Laura Potts - Chief Investigator - Public Health England
Laura Potts - Corresponding Applicant - Public Health England
Holly Townsend - Collaborator - Public Health England
Ian Wan - Collaborator - Public Health England