This course provides an introduction to systems thinking and systems models in public health. Problems in public health and health policy tend to be complex with many actors, institutions and risk factors involved. If an outcome depends on many interacting and adaptive parts and actors the outcome cannot be analyzed or predicted with traditional statistical methods. Systems thinking is a core skill in public health and helps health policymakers build programs and policies that are aware of and prepared for unintended consequences. An important part of systems thinking is the practice to integrate multiple perspectives and synthesize them into a framework or model that can describe and predict the various ways in which a system might react to policy change. Systems thinking and systems models devise strategies to account for real world complexities. This work was coordinated by the Alliance for Health Policy and Systems Research, the World Health Organization, with the aid of a grant from the International Development Research Centre, Ottawa, Canada. Additional support was provided by the Department for International Development (DFID) through a grant (PO5467) to Future Health Systems research consortium. © World Health Organization 2014 All rights reserved. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters. All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. Johns Hopkins University Bloomberg School of Public Health has a non-exclusive license to use and reproduce the material.
Lecture 1B: Health Systems Applications of CAS: Health Systems Frameworks and the Problem of Implementation
Loading...
来自 Johns Hopkins University 的课程
Systems Thinking In Public Health
369 个评分
从本节课中
Introduction to Systems Thinking and Complex Adaptive Systems
This module introduces the basics of systems thinking and the rationale for using a systems approach to solve public health problems in complex adaptive systems.
与讲师见面
David Bishai, MD, PhDProfessor
Population, Family and Reproductive Health
Ligia Paina, PhDAssistant Scientist
International Health
So, welcome back.
In this next section, we're going to be talking about health systems,
applications of complex adaptive systems, and
particularly looking at some basic frameworks of health systems.
And using the problem of implementation to Illustrate the point of complexity.
So, this example, this is the World Health Organization example of the building
blocks for health systems strengthening.
It is perhaps the most commonly used framework for a health system today.
It originated around 2006, and published in 2007.
But you should know that there are well over 50 different kinds of health
system frameworks that are out there and the number is growing all the time.
Usually to represent particular uses of having a health system framework.
But what you see here is what are commonly known as these building blocks
of a health system that come together to improve access, coverage, quality, and
safety to some overall goals or outcomes of a health system that are shown there.
But let's stop here for a moment and ask,
is there something wrong with this model of a health system, for
something that is used for many, many purposes for what it wasn't intended.
And I should say that its original original intention was not actually
to describe a health system for analytic purposes, but
really to show Ministers of Finance and Prime Ministers about what
is it where they should invest their money in order to strengthen health systems.
So hence, you get this emphasis really on these so-called building blocks.
What's missing in this model?
And here I have the picture of the model in the bottom right.
And think about it before you move on to the slides because it's most difficult to
find things that aren't there, and that's really what's the problem with this model.
So the first thing that's wrong with this is that, or lacking I should say
is that it's missing critical parts of the system and it's missing the linkages.
There are no people in this system which,
if you're an economist you'll be thinking this is the demand side of the equation,
but you might also say, from a sociological perspective that well,
actually people are at the heart of the health system so why aren't they there?
But they're not the only things that aren't there.
The key organizations or actors, the key parts of a health system,
are actually missing.
And when you look at linkages, how things are connected,
we're not actually seeing the institutions and incentives that are there, although
perhaps they're intended to be when you talk about governance and leadership.
But it's not really displayed in this kind of system.
This sort of system, although it has arrows leading from the building blocks to
outcomes, it really does ignore the dynamic nature of the system
of how parts interact with each other, and as a consequence it really doesn't
address sort of the ability to show the importance of context.
It has a sort of limited view of interventions that are rather fixed.
A set of intended outcomes as opposed to things that really happened that are both
positive, negative, and unanticipated.
And really, there's no possibility for learning an adaptation in this model.
It's almost shown as a deterministic type of model or
mechanistic model where things lead to the other.
And it really has been used and applied as a basis for basically, you know,
you put something in place and you repeat it and repeat many times.
And that's the model for how things are scaled up or
enlarged in sort of a replication type of fashion.
And that's not the model itself.
That's sort of how you interpret the model and how you use it.
So again, it's sort of the problem of using any model for
things that it's not intended for.
But it is important, you know, it is an illustrative in terms of
trying to understand what actually is a health system, in that it is more than
building blocks, which are both a combination of both parts,
as well as connections or functions in a health system.
So what are some alternatives?
I'm not going to go through the 50 plus types of systems that are out there.
But one way that you can look at it, is, look at basic things have a health system,
in terms of who are some of the actors involved, what are some of the functions
or connections, and then to what kind of outcomes or purposes of a health system.
So here I've illustrated three basic types of actors.
There's the people that are supposed to benefit from the health system,
as well as be engaged and manage their own health.
There are providers.
Those could be public, private, informally trained, preventive,
curative, and then there's the state, which involves policy makers, politicians.
The state may also have providers as well, but
this are regulators and many other different actors,
but I've tried to simplify it in terms of these three broad categories of actors.
And of course, there are hundreds of different actors in a health system.
Then there are these four broad categories of functions that
relate to each of these different actors.
From financing, how you raise money, how you pool it,
how you spend it, to how you manage the resources coming into it.
Many of these are building blocks, the input management of people,
knowledge, drugs, capital.
There's the service delivery itself, which many of us focus on, which can be curative
care, it could be public health services, it can be inpatient care, outpatient care.
And then there's that oversight or governance type of function that has and
includes policy, setting, regulation, information, and strategic partnerships.
Key functions that link the different actors in a health system.
And these work towards, say in this case three broad categories of outcomes.
One related to health status.
How sick you are and obviously, you can look at average levels or
distributions if you're interested in equity.
There's the issue of protection from health impoverishment or
becoming poor either because of illness or because of having to pay for health care.
And then there's sort of this more reflexive type of outcome
related to trust or satisfaction in the system itself.
Now you might say well, we haven't actually shown the connections between
these things, although I've tried to describe how the yellow circles or
ovals are really functions that connect the parts, but
you can make these kind of connections if you bother showing all these arrows,
as is shown here in this, but it becomes a little bit complicated.
And the more complicated it becomes,
the less useful it becomes as a systems diagram.
But this is just to illustrate the point that, in any system,
you need to have purpose, you need to have elements, and
you need to have connections between them that lead towards this purpose.
So I'm going to shift gears now, and
talk about implementation or evidence-based policy and planning.
And really as a basis for why we need to apply systems thinking in health,
not just to come up with these nice pretty diagrams that we've had, or
at least I hope you think they're pretty.
And the question we have is why can't health systems,
in this case I'm talking about national health systems that provide population,
public health, preventive, curative care.
Why can't they deliver consistently and
at a large scale in particular to the vulnerable populations?
Here I'm showing three books that I'm going to be taking some of the examples
and thinking from the first the World Health Report, 2000.
The second is a book that we wrote a few years ago about improving service
delivery in middle and low income countries and always learn from there.
And the latter being a guide on implementation research,
which is looking at how do you understand and intervene in health systems.
And which is really about the science of delivery,
to try and, and the book was really intended to link both policy makers and
researchers together with some common language and approaches.
Now, I'm going to caricaturize the current situation.
It basically it follows a typical formula for how do you improve and
how do you scale up health services.
Again, I could be talking about preventive services or curative services.
For this example, it doesn't really matter.
But you start off with the notion that you use to choose the right health
interventions, which is typically what's meant by cost effective and Preferably
cost-effective for an intervention that's affecting a disease of large burden.
And here's two examples of books that have said some ground-breaking work in
terms of defining what are cost-effective interventions, in both cases,
low and middle-income countries being the focus.
The second step is to set ambitious and common targets and here you see the MDGs
which are about to be replaced by another set of goals in 2015.
Then they approach us to fund them, and here you see some funders,
all of which have really emerged in the last 15 years.
As major funders of interventions to improve health in low income countries.
And the last step is that you implement these interventions as designed.
And this is where we come into trouble with this cookie cutter approach.
One size fits all,
the blueprint design problem where we failed in terms of systems thinking.
So why do we need systems thinking?
Well, what's wrong with the cookie cutter approach?
With this picture here is actually a simplification where each line represents
a country, and it's the linear trend of a country in terms of changing
skilled birth attendance.
The 90% is the current target, line with diamonds on it is the so
called average, and every other country fits on itself.
And what you can see [INAUDIBLE] demonstrates itself is
its own line of progression.
And what you see is that although there are few countries at the top that
are very similar in terms of maintaining high level of skilled birth attendance.
Otherwise countries have very divergent patterns.
Some going down over time and it looks a bit like a pick up sticks drawing.
But the notion of having an average country really doesn't apply to anybody.
Well there might be one or two countries that are close to average.
But otherwise, no such thing as an average country.
If you then take, this is one of the MDG target services,
if you take one of the MDG goals, which is child mortality,
you see a similar type of picture except that this time the cluster of countries
that are the same is at the bottom, those that have low mortality.
Everybody else is changing at different rates, and if you do the statistics
behind it, you find actually that it's the individual country intercept and the slope
that is the most defining characteristic of the model, so suggesting that countries
are more different than they are alike in terms of how they reduce child mortality.
And this is just already generalizing whole countries comparing to each other,
not reflecting the wide levels of diversity within a country which
would make this even larger, the kind of diversity.
So it suggests that when you have similar goals even similar kinds of interventions,
with different context, different ways of implementing
the results are vastly different from one place to the other.
In the next slide, we actually look at how the service coverage
affect different parts of a population.
And so here we're looking at the poor rich ratios for
six sets of primary care services around the world, in this case the aggregation is
even greater Its at regional level so grouping countries together and
what we've done is compare the rates of services delivery where you'd
expect everybody to be a 100% coverage and this is a spider diagram.
So if the rich and the poor is 20% each had equal levels of service coverage,
that number would be at one.
It would be that outer band all the way across.
And if the poor had higher service coverage than the rich,
the number would be larger than one.
So the first thing you see is that nowhere in the world on average
do the poor have higher service coverage than the rich.
But also, there's not a whole lot similar in terms of the different services and
the degree to which the poor benefit compared to the rich.
If one thing you could say is perhaps that contraceptive prevalence and
ORS tends to be more equitable than other services, and the least equities
tends to be medical service delivery, in other words, institutional delivery.
But otherwise, across regions, there's no fixed pattern.
So again, it tells you that, in reality, there's no fixed ways of becoming
pro-poor, which you know is one of the main targets and
when you're working in public health as to try and
reach ensure that services get to the most disadvantage.
So again, there's no fix types of outcomes that you can get.
So that requires you to think systematically and
outside of that cookie cutter that you otherwise might like to use.
So this next slide I'm going to go through an example from Afghanistan where we look
at implementation and in this case it's about implementation of building up their
health sector, primary health care, after the Taliban left in around 2002, but this
actually started this assessment in 2004, and we're using a balanced scorecard,
in this case we're going to look at eight years of a balanced scorecard.
Now after the Taliban had left and an interim government came in place,
they put a standard package for primary care, this basic package of care for
primary care, that they wanted implemented across the country.
No matter which NGO or funding agency was put in place, and
they basically follow that.
And they came up with the notion that we'd follow this balance score card,
which is this sheet here with the red, green, and yellow signs on it.
An agreement around basically 29 different indicators and six different domains.
And they were measured every year by an independent party,
which happened to be John's Hopkins with Indian Institute of
Health Management Research, in collaboration with the Ministry of Health.
And we did it in every province of the 34 provinces.
We did it every year, and it was used for annual priority setting, budget setting.
In many cases, there were contracts and contract bonuses were given or
decisions about maintaining them.
And as you can see from this, I know for each indicator and
province this is the baseline in here, that you could at a glance look at
which province had which problems which were the red coloured ones,
which were basically the bottom quintile of performance.
Which ones were doing really well which was the green.
And in the middle was everything else,
the middle three quintiles of performance with benchmarks being Afghanistan.
So this is an approach that showed huge diversity
in terms of how the different pieces of primary healthcare were put together.
Looking at patient perspectives,
looking at staff perspectives, looking at the ability to provide services.
A lot to turn around the observation of the quality of care and
then financial systems.
And the overall vision was really about whether they were reaching the poor and
whether women were benefitting, at least as compared to men.
So that was the set up of the Balanced Scorecard.
Trying to put the different pieces of the primary care system together and
then showing it transparently in a way that was used for decision making.
So that's the scorecard in year one and we did if for
each province that you could follow over time and
it was used over the next ten years and it continues to be used as we talk.
This slide here, which is a bunch of lines going in different directions
tries to show you how over from 2004 to 2013,
each line represents a province with just one of those indicators.
It's how well did health providers do a proper history and physical exam using
a checklist that we used through a random selections of patients going through.
And as you can see, each province had a very different outcome from year to year.
And what stands out I guess, is that in 2008 one province that had
had a really bad year, and then picked it up next year.
But for the most part you can see that there is a lot of variability.
So again, when you're thinking about implementation, you have
to recognize that there is variability in performance, and therefore you really need
to have a good framework to be able to follow how well things are going.
If you put them together, you get a different perspective, so
here we've simplified now, each line.
Represents an average of each of those six domains and
how they work nationally from year to year.
And we did this over five years here because we changed the indicators
substantially in 2009.
But what you see here is actually all of them showed significant improvement
over the course of the five years.
So what you see here is that at an average level, putting them together,
you can see positive trends on a national basis that,
again, is a good sign in itself, but also recognizing that there's large variation,
and you have to think about how provinces affect each other differently.
And this was actually used in order to
highlight how they could learn from provinces that were doing better.
How they could focus efforts on either on indicators or
domains that were doing poorly, or on provinces that were doing poorly.
This gives you the parameters to be able to do that.
So it's a systematic way of looking at how things fit together, used for
regular kind of management decision making.
If you put this together, again,
using that book on looking across long delinquent countries,
there's some common findings about implementation that is important.
And again, a real reason why you need to use systems thinking when you're thinking
about how you implement things.
One is that there are many interventions that work in pilot projects or
in research conditions.
Most publications are about
studies that are well controlled that actually show an improvement.
But implementation faults are nonetheless very common in real world conditions,
less common in the published world, but in the real world in implementation,
they happen all the time.
We do know that there are many different types of strategies that can succeed, but
they're not actually replicable even in much detail,
even when they have the same level.
If you do integration management of childhood illness program,
the INCI program, you might have the same training materials,
you might even have the same guidance, but how they work
actually is implemented in very very different ways from place to place.
A user fee policy dictated by central decision makers
might get implemented very differently from region to region, or
even from health worker to health worker, when they have that kind of flexibility.
So that's the third point, that policy makers often define strategies, but
they often have limited influence in how they're actually implemented in practice
because it's these front line health workers, or street level
bureaucrats they're sometimes called, have this capability of changing things.
We often find that strategies often achieve their objectives, particularly if
you're reading the literature, because that's what gets published.
But they also produce many unintended and unpredictable consequences.
And we'll come to that a bit later.
But this is an important part about implementation, is that you can get
something that you're focused on to work, and yet that may have dire consequences
for something else in the health system, or it may have positive influence.
If you do something, for example, that might improve
people's willingness to come in because you offer a good quality service for, say,
antenatal care, maybe they'll stick around for that institutional delivery.
But often times you might find that,
well they'll invest a lot of money in a particular program.
Let's say an HIV program.
All the health worker attention and the focus, the funding gets put on HIV.
And the other services are left behind and
they don't do as much, be it reproductive health or chronic diseases.
And that's a common kind of finding, particularly in low resource settings,
when you put a lot of resources in a single area.
Lastly, I can say that there are many health policies that intend to serve
the poor and vulnerable populations, but when they're actually implemented,
they rarely measure and show how they improve services for disadvantaged people.
So it's not really clear that the policies are being implemented to actually
accomplish what they're intended to do.
So in fact, that describes sort of some of the basic rationales for
wanting a health system you want to be looking at complexity
because of the nature of how things work in a health system.
