Transcript Needs Assessment

The Epidemiology of Complexity
and Needs Assessment
Nigel Monaghan
Environment
Agents
Agents
Evolution of Needs Assessment
• Third Party (Stevens and Raftery)
to
• Participatory
Needs Assessment
• “a systematic method...:
– for reviewing the health issues facing a
population, leading to agreed priorities and
resource allocation that will improve health
and reduce inequalities”
Health Development Agency 2005
– of identifying unmet health and health care
needs of a population and making changes to
meet these unmet needs’
Wright and Kyle, in Pencheon 2006
Objective of needs assessment
• Inform process of change
• Engage community on shared view of
problems and agree actions
Move to participatory
• Recognises that change is often needed within
a system (not to a process)
• Increases the probability that most/all will be
supporting the change
• Accepts that change achieved may be
compromised
Philosophy vs Practice
• While the philosophy of needs assessment has
evolved
• Our use of data has not
• Most needs assessments report on incidence
and/or prevalence of determinants, disease
and impact
Simple Model
Determinants + Primary Prevention
↓
Early Disease + Secondary Prevention
↓
Late disease + Treatment + Rehabiliatation
↓
Outcomes
Simple Needs Assessment
• Population demography and “determinants of
health data”, incidence
• Services available: primary prevention,
secondary prevention, care services
• Service/treatment effectiveness
• Models of care
• Outcome measures and targets
• Information and research priorities
Traditional epidemiology
• Built on simple relationships
• Determinants increase probability of disease
• Incidence/prevalence defines pockets of high
need
Too Simple?
• Dahlgren and Whitehead model is a simple
model of complex interplays and interactions
contributing to chronic disease
• A simple deterministic approach is used to
summarise the epidemiology of a series of
nested complex systems
• Beware over- simplification!
For Infectious Disease
In health protection:
• We ask about environments visited
• We ask about interactions over periods of risk
of transmission
• We ask about the nature of those interactions
For infectious disease
• We consider the context (diagnosis)
• We consider the nature of relationships and
contacts
• The model is not of Brownian motion but of
agents (carrying disease or at risk) interacting
Epidemiology of Health Protection
• Is based upon a simple model of disease –
source, pathways, susceptible receptor
• Is complicated by uncertainty as to upstream
sources or previously exposed downstream
recipients
• Places value on interactions with environments
and between agents and the quality of those
interactions
Epidemiology of Health Protection
Paradigm
• If we understand the pathways and
relationships (i.e. the interactions and their
quality) we can identify the source and
potential receptors and use this information to
manage the disease
Do relationships influence health?
•
•
•
•
Do smokers associate with other smokers?
Do drinkers drink together?
Do some people take exercise together?
Do families eat the same food?
Is obesity contagious?
• Is weight gain in one person associated with
weight gain in friends, siblings, spouse or
neighbours
(Christakis and Fowler 2007)
• Increased risk of obesity associated with:
• friend 57%
• sibling 40%
• spouse 37%
• Same sex = stronger association
Is obesity contagious?
• Obviously obesity is not an infectious disease
• Reasons for obesity could include:
– Social network effects (as per Christakis and Fowler 2007)
– Choice of friends based on weight factors
– Common context i.e confounding factors (3 listed by
Cohen Cole and Fletcher 2008)
Is obesity contagious?
• Cohen-Cole and Fowler suggest that the
context, specifically the environment explains
obesity (with some possibility of a role for
person-person interactions)
• Does obesity emerge from interactions
between people and interactions with their
environment?
• Do we understand what is happening?
Epidemiology of obesity
• When we undertake needs assessments...
• Do we look for features of an obesogenic
environment in the various places people live
their lives?
• Do we consider who people interact with?
• Do we collect enough qualitative information?
e.g Smoking Cessation
• Do we focus on the prevalence of smoking?
• Do we understand the networks within which
some people smoke (family, friends, work)?
• Do we consider intervening at family level or
seek to work with groups of people?
e.g Smoking Cessation
Do we understand...
• as much as tobacco marketers do about
smokers?
• why people start smoking?
• where they smoke?
• who they are with when they smoke?
e.g. Teenage pregnancy
• In addition to reporting rates...
– Do we understand why some children become
pregnant?
– Or why others in the same community do not?
– Do we understand why some choose to become
parents?
• When one girl becomes pregnant...
• ...do friends become pregnant too?
Teenage pregnancy
• In addition to the quantitative data
• Have we collected qualitative data?
• Have we understood the different factors
contributing to teen pregnancy?
• And if not, how effective are morning after pill
and long acting contraception likely to be?
Teenage pregnancy
• E.g. For teenage pregnancy do we understand:
– Reasons why some teenagers become pregnant?
– Reasons why others do not?
– Interplay between factors which have led to a
pregnancy?
– Interplay between factors which have led to a
decision to continue an unplanned pregnancy?
• Potential for a good research bid?
The epidemiology of complexity
• Environment (physical and psycho-social)
• Agents
• 2 way interactions (environment with agents
and agents with agents)
• Emergent properties
If the interactions...
• ...of a complex adaptive system differ from
those of simple systems
• Then maybe the needs assessment process
needs to reflect those differences
Implications for needs assessment
• Information about the environment, e.g.
Determinant of health type data
√
• Information about the agents, e.g.
Demography of the wider population and of at
risk groups
√
Implications for needs assessment
• Information about interactions between
agents and with the environment e.g.
qualitative information to understand range of
factors leading to teenage parenthood
X
• Attempt to understand how agents and
environment are interacting (consider
modelling?) leading to emergence of public
health problems
X
Implications for needs assessment
• Developing understanding (and possibly
models of problems) which are conceptually
closer to spirit of Dahlgren and Whitehead
model than a deterministic upstream/
downstream, cause/effect relationship. X
Through a complexity lens
Violence and pubs
• Changing the environment can help – remove
glass = fewer glassings
• Understanding the scale of the problem
meant collecting data from A&E
• Intervening when people are ready to help
meant brief intervention at follow up in A&E
Through a complexity lens
• Have we considered why the same few pubs
seem to attract violence – implications for relicensing?
• Do we understand how and why violence
emerges and what antecedents it emerges
from?
• This should be a topic for another day
For complex public health problems...
• ...should we consider the nature and quality of
interactions as important
• Should we put more effort into understanding
these interactions?
• Before we design and implement
interventions?
For complex public health problems...
• ...and what about the role of physical and
psycho-social (cultural) environments
• Are there lessons from evolution (and
breeding) for interventions
• To inform design and implementation of
interventions?
Thanks for listening
• The content of this presentation probably has
implications for evaluation too.
• Again, a topic for another day.
Co-learning
• Needs assessment was originally designed for
“specialist hospital care in controlled
environments”, and as designed is probably
still adequate for purpose
• There has been “mission creep” with needs
assessment applied to more complex health
and social care situations
Co-learning
• In complex health and social care situations if
it is appropriate to assess need then the
approach should seek to capture
understanding the nature and quality of
agent/agent and agent/environment
interactions