Moves into residential care and sheltered housing in later life

1
Moves into residential care and sheltered housing
in later life
ESRC Centre for Population Change
Jane Falkingham, Maria Evandrou, Olga
Maslovskaya, James Robards and Athina Vlachantoni
Centre for Research on Ageing, Centre for
Population Change CLC, University of Southampton

• Workshop on UK Population Change Housing across
  the Life Course
• Tuesday 16 June Wednesday 17 June 2015
• University of St Andrews

2
Outline

• Background and research aims
• Data methods
• Results
• Discussion conclusions
• Acknowledgements

3
Background

• The recent White Paper (2012) highlighted the
  complexity of organising social care for older
  people.
• Dilnot report (2011) emphasised the importance of
  older people having choices regarding housing
  arrangements in later life.
• At the 2011 Census 16.4 of the population was
  over the age of 65 years compared to 15.9 in
  2001.
• Population ageing and changing family structures
  increase pressures on the current system.
• Increasing demand for social care poses
  challenges for policy-makers and families alike.
• Other research has considered housing transitions
  in relation to palliative caring (Abarshi et al.,
  2010).
• Higher mortality among persons transitioning to
  long-term care institutions has been identified
  using census data (Grundy, 2010).

4
Aim and research questions

• Overarching aim is to explore patterns of older
  peoples transition into different types of
  accommodation in later life
• I to understand the factors associated with a
  persons transition into different types of
  accommodation (eg. demographic, health,
  socio-economic).
• Are the factors associated with an older persons
  transition into residential care different to the
  factors associated with an older persons
  transition into sheltered accommodation?
• II to investigate mortality in relation to
  residential and sheltered housing transitions
• Which housing pathways are the strongest
  predictors of subsequent mortality?
• What is the duration in each housing type in
  relation to subsequent mortality?
• What are the key socio-economic and gendered
  differences for housing transitions?

5
Defining residential care and sheltered
Accommodation

• Residential care is a type of living arrangement
  where older persons with physical and/or mental
  frailty move into a residential home providing
  board and personal care 24/7, following the
  assessment of their needs.
• Sheltered housing is mainly for older people and
  usually takes the form of a group of small
  bungalows or flats supervised by a Scheme
  Manager, who can offer some help and support in
  an emergency.

6
Data and Methodology I Moves into residential
sheltered accommodation

• BHPS (1991-2008), focus on 65 in England Wales
• Paired-year records to increase number of
  transitions
• Two models with separate outcome variables
  Transitions between t0 and t1 into a) residential
  care (N113) and b) sheltered accommodation
  (N175)
• Binary logistic regression, stepwise model,
  interactions, robust standard errors
• Categories of explanatory variables Demographic
  characteristics Health and well-being Use of
  formal care services Socio-economic
  characteristics Informal care receipt Time

7
Conceptualising transitions into long-term care
8
Data and Methodology II Mortality and transition
into residential sheltered accommodation

• British Household Panel Survey (BHPS) waves
  1993-2008. Pooled years. Data for household and
  individual.
• BHPS members aged 65 years plus at 1993 and
  living in a private household selected. N24,227.
• Binary logistic regression of mortality three
  waves after baseline.

9
Data and Methodology II Mortality and transition
into residential sheltered accommodation

• Cross wave housing trajectory variable created
  for persons in private housing at baseline.

10
Results I Determinants of moves into
residential care and sheltered housing in later
life
11
Figure 1 Proportion of older personstransitioni
ng into (a) residential and (b) sheltered
accommodation, England Wales, 1991-2008
(1a)
(1b)
Source Authors analysis of BHPS, 1993-2008.
12
Table 2 Variables associated withtransition
into residential/ sheltered care
  Transitions into residential care Transitions into sheltered accom.
Variables Chi-sq (df) Chi-sq (df)
Age group at t0 352.996 (4) 104.485 (4)
Marital status at t0 120.457 (3) 61.466 (3)
Region at t0 8.141 (1) 26.048 (3)
Access to washing machine at t0 170.746 (1) 443.709 (1)
Waves at t1 57.755 (16) 30.182 (15)
Self-reported health status at t0 56.065 (3)  
Hospitals inpatient days at t0 139.424 (3)  
Use of social worker at t0 92.879 (1)  
Change in use of social worker between waves t0 and t1 83.955 (1)  
Hospital in-patient days Use of social worker 92.306 (3)  
Use of home help at t0   42.803 (1)
Highest educational qualif. at t0   16.326 (2)
Household type at t0   85.947 (1)
Housing tenure at t0   263.200 (3)
Access to car at t0   87.901 (2)
13
Table 3 Determinants of moving into residential
care (I)
Variables Odds ratio
Age group at t0 Age group at t0
65-74 (ref) 1
75-79 2.188
80-84 6.662
85-89 12.554
90 17.841
Self-reported general health at t0 Self-reported general health at t0
Excellent (ref) 1
Good or very good 0.740
Fair 2.874
Marital status at to Marital status at to
Married or living as a couple (ref) 1
Widowed 3.277
Divorced or separated 3.494
Single never married 3.812
14
Table 4 Determinants of moving into residential
care (II)
Variables Odds ratio
Hospitals inpatient days at t0
None (ref) 1
Under a week to 2 weeks 0.59
2-5 weeks 1.52
5 weeks to a year 5.54
Use of social worker at t0  
No (ref) 1
Yes 2.61
Change in use of social worker between waves t0 and t1  
Otherwise (ref) 1
Started such use 3.89
Other variables included in the model access to
washing machine, interaction effect
15
Table 5 Determinants of moving into sheltered
accommodation (I)
Variables Odds ratio
Age group at t0
65 to 74 (ref) 1
75 to 79 1.91
80 to 84 2.11
85 to 89 2.62
90 2.08
Marital status at t0  
Married or living as a couple (ref) 1
Widowed 0.55
Divorced or separated 0.29
Single never married 0.36
Household type at t0  
Living alone (ref) 1
Living with other people 2.38
16
Table 6 Determinants of moving into sheltered
accommodation (II)
Variables Odds ratio
Access to car at t0
Yes (ref) 1
No 1.82
Does not drive 1.90
Highest educational qualification at t0  
No qualification (ref) 1
None of mentioned qualifications or O-levels/A-levels 0.97
Degree, teaching, nursing or other higher qualification 1.84
Housing tenure at t0  
Own outright (ref) 1
Own with mortgage 1.68
Rented from Local Authority or Housing Association 6.71
Rented privately or other rented 1.55
Access to washing machine at t0  
Yes (ref) 1
No 5.09
17
Figure 2 Predicted probabilities of moving into
residential care for individuals by time spent in
hospital during the previous year, use of social
worker and self-reported health status
17
Source Authors analysis of BHPS, 1991-2008.
18
Summary of results I

• Moves into residential care and in sheltered
  accommodation are associated with different sets
  of factors, indicating different pathways into
  long-term care in later life.
• For the move into residential care, an older
  persons age and marital status, as well as
  variables associated with their health status and
  use of support services are the main predictors.
• By contrast, indicators of an older persons
  socio-economic circumstances appear to have
  strong effect on the risk of moving into
  sheltered accommodation, although demographic
  characteristics (age, marital status, living
  circumstances) and region are also part of this
  story.

19
Discussion I

• What could explain such differences?
• - The move into residential care is more likely
  to take place at a later point in ones life
  course than the move into sheltered
  accommodation, hence health is an important
  determinant of this type of move.
• - The move into sheltered accommodation is more
  likely to be determined by the older persons (or
  couples) financial resources, rather than their
  need for care, hence socio-economic factors are
  more important with this type of move.

20
Policy implications I

• Particular types of long-term care housing may be
  more appropriate for particular stages of the
  life course, associated both with ones age and
  health status.
• Socio-economic resources are at least as
  important as demographic characteristics and
  social resources in the form of available
  informal support.
• Policies which target specific groups of the
  older population also need to take into account
  cohort differences in both financial and social
  resources, as well as the rise of different
  expectations of older people (eg. on
  independence, consumption, technology).
• Key role of health and social care professionals
  in facilitating the transition into long-term
  care.

21
Results II Housing and care transitions at
older ages in relation to subsequent mortality
22
Results (IIa)
Residential housing transitions T2-T3 are crucial
to predicting mortality
Table 7 Binary logistic regression outcome is
death T2-T3

• Other explanatory variables included model
• - BHPS wave.
• - Marital status.
• - Employment category.
• - Access to a car.
• - Health status.

Source Authors own analysis of BHPS.
23
Results (IIb)
Health and marital status are key in accounting
for housing transitions and mortality in later
life
Figure 3 Odds ratio of death (T2-T3) reference
is no transition from private housing
MODEL 1 wave, age and sex
MODEL 2 wave, age, sex and marital status

• MODEL 3 wave, age, sex, marital status and
  health status
• MODEL 4 wave, age, sex, marital status, health
  status and socioeconomic variables

Source Authors own analysis of BHPS.
24
Results (IIc)
Male mortality is higher across all housing
transition types
Figure 4 Predicted probabilities of death for
males and females by housing transition
Source Authors own analysis of BHPS.
Source Authors own analysis of BHPS.
T01994, 90 years, never married, poor / v poor
health, manual and unskilled occupation, missing
for access to a car.
25
Summary of results II

• Key explanatory variable combines (sheltered and
  residential) housing trajectories and duration of
  stay to examine mortality in later life.
• Annual transitions more realistic for the study
  of housing moves, differs to past research.
• Transitions into sheltered and residential
  housing have been compared side-by-side.

26
Summary of results II

• Transitions to residential care 12 months before
  any mortality show the highest mortality risk.
• Health and marital status are important in
  accounting for housing transitions and higher
  odds of mortality.
• Men show higher odds of mortality across the
  range of housing transitions.

27
Discussion policy implications II

• Understanding the mortality risk associated with
  different housing pathways among older people
  is crucial in planning for the future demand for
  long-term care housing and services, and
  important in a policy context which provides
  alternatives between different types of long-term
  care accommodation.

28
Discussion policy implications II

• Government policy encourages older people to
  remain in their private home for as long as
  possible this paper shows that older people's
  move away from home is associated with a
  significant mortality risk.
• The results in this paper highlight that it is
  not only the type of accommodation which affects
  an older persons mortality risk, but also the
  length of time they spend there.
• Future research can further investigate the
  mortality risk for older people who move from
  their own private home to a relative's private
  home in later life.

29
References

• Robards, J. Vlachantoni, A. Evandrou, M. and
  Falkingham, J. (2014) Mortality at older ages
  and moves into residential and sheltered housing
  evidence from the United Kingdom Journal of
  Epidemiology and Community Health 68(6)524-529.
• Vlachantoni, A. Maslovskaya, O. Evandrou, M.
  Falkingham, J. (2015) The determinants of
  transitions into residential accommodation in
  later life in England and Wales CPC Briefing
  Paper
• Vlachantoni, A. Maslovskaya, O. Evandrou, M.
  Falkingham, J. (revise resubmit) The
  determinants of transitions into sheltered
  accommodation in later life in England and
  Wales Journal of Epidemiology Community Health

30
ESRC Centre for Population Change
Find out more and contact us
Web www.cpc.ac.uk Email cpc_at_southampton.ac.uk Te
l 44 (0)2380 592 579

• Twitter _at_CPCpopulation
• Facebook /CPCpopulation
• Mendeley Centre-for-population-change
• Scoop.It Centre-for-population-change

31
Multivariate analysis Variables in the final
model(main effects only)
Demographic Age at t0 Marital Status at t0 Region at t0
Health and well-being Health Status at t0 Change in Health Status Hospital in-patient Days at t0
Use of services Use of Social Worker at t0 Change in Use of Social Worker
Socio-economic status Financial Situation at t0 Income at t1 Washing Machine at t0 Access to Car or Van at t0
And Time
32
Multivariate analysis Variables not
statisticallysignificant in the final model
Demographic Sex Ethnicity Education SES Number of children Living arrangements (all at t0)
Informal care receipt Cared by household member at t0
Health and well-being GHQ score at t0 Change in GHQ score Disability Status at t0 Change in Disability Status
Use of services Use of Health Visitor at t0 Change in h.visitor use Use of Home Help at t0 Change in home help use Use of Meals on wheels at t0 Change in meals on wheels use Number of visits to GP at to
Socio-economic status Tenure at t0 Central heating at t0 Access to Car or Van at t0 Income support at to Disability Allowance at to Attendance Allowance at to