RAND/UCLA/WHO Workshop On Health Status Measurement In Social Surveys
Santa Monica, October, 1999
Duncan Thomas and Elizabeth Frankenberg
- Introduction
- Health over the life course
- Interview assessments
- Activities of daily living
- Nutrition
- Clinicals and specimens
- Health expenditures and prices
- Summary
- Agenda
- Participants
The idea for
this workshop emerged out of a series of discussions between Dean
Jamison,
Ritu Sadana, Elizabeth Frankenberg and Duncan Thomas. Broadly speaking,
it was intended to provide a forum for a dialogue on the meaning and
interpretation
of health indicators collected in social surveys. In recent years, that
has been a dramatic change in the technology available to measure health
status in a field setting and there has, concomitantly, been increasing
interest in the collection of biomarkers. Several studies have
implemented
new methods for measuring health while others have tested new or revised
instruments that collect health status indicators in interviews. At the
same time, there have been many important advances in our understanding
of
-
the meaning of these health indicators
-
correlations among the health indicators
-
correlations between health status early in life and health status later in life
-
correlations between health and other life course outcomes including productivity.
Recently, there has been nothing short of an explosion in the extent of interest among social scientists in the links between health and various dimensions of socio-economic status and the variation in this relationship over the life course. The overarching goal of the workshop was to bring some of these many threads of current enquiry together drawing on international experts who span the health and medical sciences, social sciences and survey research. The workshop was attended by participants from the United States, Europe, Mexico and several South and Southeast Asian countries.The workshop began with a discussion by Ritu Sadana of some of the key issues in the measurement of health currently facing the World Health Organization, WHO. First, there is a need to improve the conceptual basis used to measure health. This will likely involve the use of both qualitative and quantitative methods as WHO attempts to develop a set of standardized protocols and terminologies that can be applied cross-nationally. It will involve coming up with useful definitions of the boundaries of health, specification of health profiles and identifying a core set of domains of health. From a practical point of view, of particular interest is the use of self-reported measures of health and performance. These measures are widely used in developed countries and have been extensively validated. There are real questions about their validity in lower income countries and on-going work by WHO is attempting to provide some new insights into this issue. Nonetheless, the future is bright.
There are good reasons to anticipate substantial increases in the breadth and quality of data collected on health and socio-economic status (SES) in integrated surveys in the near future. There are also good reasons to anticipate important improvements in the extent and quality of the analyses of those data as they become more widely used.Jennifer Madan then described some of the issues that National Center for Health Statistics (NCHS) faces as it undertakes its responsibilities of collecting, analyzing and disseminating information on health status in the United States. In contrast with many agencies and many researchers, NCHS tends to take health apart rather than try to put it together into a single index. NCHS takes as a given that health should be defined extremely broadly and includes biomedical indicators, risk factors and behaviors, social factors and the mediating influence of SES and demographic characteristics. The National Health and Nutrition Examination Survey (NHANES) and Health Interview Surveys (HIS) are the centerpiece of NCHS survey efforts although they are only one element of a much broader portfolio of health data collection.There can be no doubt that NCHS stands head and shoulders above health data collection agencies in almost any other country. There is an increasing awareness at NCHS of the need to co-ordinate data collection efforts within agencies of the US Government and, particularly, to link survey data with administrative records. Multiple initiatives along these lines are currently underway at NCHS. There are, however, many obstacles -- confidentiality taking center stage in this regard.
There
is also a push at NCHS to develop a greater longitudinal capacity and
discussions
are underway to consider undertaking surveys that more closely resemble
the British Cohort Studies. This is a reflection of a greater interest
among the research and policy constituencies to better understand how
health
flows evolve -- that is, in understanding health transitions -- and to
develop more insights into the causal pathways that culminate in the
stock
of health of an individual, community or population at any point in
time
These remarks
set the stage for the first panel of the workshop. In recent years, there
has been a marked increase in our understanding of the links between
early
life experiences and health and well-being later in life. Important work
along these lines is being conducted by David Barker and his associates
who have placed particular emphasis on intra-uterine health and, more
generally,
the inter-generational transmission of health from mother to child.This
innovative line of research was introduced by Barbara Cohn.
Barker's
hypothesis argues that insults during early life -- be it prenatal or in
early infancy -- programs the body so that years later, during adulthood,
health problems emerge. These problems include chronic adulthood diseases
such as coronary heart disease, strokes and hypertension as well as
reproductive,
neurological, cognitive and behavioral functioning. For example, Barker
argues that if a baby born with low birthweight may be indicative of the
foetus being starved of critical nutrients of oxygen which resulted in
stunting development of some parts of the body, relative to others.
Depending on the timing and extent of the intake deficiency, this will result in some organs being less well developed than others. For example, reduced body mass growth is often associated with less developed livers which is associated with higher incidence of coronary heart disease in adulthood.Cohn pointed out that Barker's work is not uncontroversial and that several critics have argued that the work has failed to take into account other confounding factors -- including a fuller conception of maternal health and the role of the environment. She described two important cohort studies in the United States that have the potential to provide further insights into the links between foetal and infant experience and later life outcomes. They are the Child Health and Development Studies (CHDS) and the Collaborative Perinatal Project (CPP). Michael Wadsworth had prepared remarks but, because he had been taken ill, they were presented by Duncan Thomas. Wadsworth introduced the British Cohort Studies which include 3 national longitudinal studies conducted in Britain which share a common design. Each cohort has been surveyed since birth with periodic interviews and assessments of health and cognitive functioning. The first cohort included a sample of all children born in the first week of March, 1946. It has been followed up 22 times with the most recent interviews being at ages 43 (1989) and 51 (1997).
A second cohort followed up all births during the first week of March, 1958, and a third cohort was begun in 1979. Plans are underway for a fourth birth cohort which will parallel the Education and Child Longitudinal Study-Birth Cohort (ECLS-B) which will begin in 2001 in the United States. He also mentioned the Avon longitudinal study of 10,000 children and the Barker follow-up studies. Wadsworth presented a broad overview of some of the results from these and other cohort studies. He began with the effects of early life factors on physical health. He showed that elevated blood pressure at mid-life is associated with poor nutrition and growth in childhood. Risk of later health problems is greatest among low birthweight babies who subsequently are overweight in childhood: a high ponderal index (weight/length 2) is protective. Childhood micro-nutrient intake affects physical development which, in turn, impacts health in adulthood. Elevated levels of Vitamin C intakes in childhood are associated with better lung development and lowers the probability of pulmonary disease (COPD) in adulthood. He proceeded to argue that early life factors influence mental health and cognition drawing on evidence regarding breastfeeding and IQ, emotional disturbance during childhood and physical growth, and the inter-generational transmission of emotional vulnerability. While there was widespread agreement that early life experiences probably do have consequences later in life, there was less agreement about the precise mechanisms. A central concern that emerged in the discussion is the need to parse out the role of phenotype and genotype influences. Demonstrating a correlation between, say, birthweight or breastfeeding and later health or cognitive functioning does not provide the information needed to assess the causal pathways that link poor health early in life with poor health later in life.
There are many confounding factors and an understanding
of those factors is currently missing. That
said, there was also general agreement that this is an extremely exciting
line of inquiryfrom the point of view of those in the biomedical and
social
sciences and, perhaps especially, for those groups whose work transcends
the boundary of both domains. Workshop participants agreed that
investments
in the collection and analysis of high quality long-term panel surveys
that contain comprehensive information on both health and socio-economic
status will surely yield very high pay offs.
The workshop
discussed some of the interview-based methods of collecting information
on health status. Bob Wallace
plays a key role in the design of
the Health and Retirement Study (HRS) and the Assets and Health Dynamics
of the Oldest Old (AHEAD). The former is a random sample of households
with one member born between 1931 and 1941; the latter sampled households
with a respondent born prior to 1923. He described the scientific issues
that were part of the original motivation of the surveys and discussed
how those issues had been modified as the survey progressed. He discussed
how issues that had arisen had been resolved and how new issues had
emerged
during the course of the survey. HRS
and AHEAD are done on the telephone and so rely on self-reported health
indicators. Wallace's discussion of their limitations included concerns
about the validity of illness reports in an environment in which there
are important limits to personal knowledge about current health
conditions.
There are special challenges that are particularly germane when working
with older respondents revolving around cognitive functioning, the
complexity
of health states and illnesses because of co-morbidities, poly-therapies
and decreased functioning with age of many of the body systems. Several
of these factors are likely to be associated with a respondent's
willingness
to participate in a study of this nature: participation tends to decline
with age, reduced cognitive functioning and elevated health problems.
This
choice-based sample poses special challenges for research that seeks to
understand the behaviors of the population. It was noted that the fact
that respondents enter these surveys only after they have reached age 50
means that many of the life events that will have affected their health
will have already occurred prior to the first interview. Whereas Wallace
is very skeptical that attempts to reconstruct key events in an
individual's
health life history, others in the audience felt that remained an open
question and one that should be tested empirically.
This seems right particularly
in view of the life course approach to examining health outcomes
discussed
in the previous session of the workshop. Jane
Gentleman followed up with a discussion of the Health Interview
Survey
(HIS). It is a very large scale annual survey that serves both as a key
source of information on prevalence of health problems and also as a
sampling
frame for more focussed surveys such as the Medical Expenditures Panel
Survey (MEPS). The multiple demands on HIS has dictated a number of
design
compromises. For example, it is important to include a large number of
households to build the MEPS sampling frame; that precludes interviewing
more than one adult and one child in each household.The discussion moved
on to the construction and interpretation of health status scales. John
Ware
describes health status as being characterized by a series of
concentric circles. Biological functioning lies at the center; the next
layer involves physical and mental health; the next layer is concerned
with social and role participation; finally, there is health-related
quality
of life. When measuring health, one needs to be clear about which of
these
levels one is interested in. Moreover, he notes that it is important to
distinguish between objective functioning, distress and well-being and
a respondent's articulation of their perception of their own health.That
said, Ware advocates rigorous standardization of health measurement to
provide comparability of indicators across time and space using a common
metric or ruler. He described the International Quality of Life
Assessment
(IQOLA) Project which seeks to translate and adapt the Medical Outcomes
Study SF-36 so that those question can be validated, normed and
documented
for use in international comparisons of health. SF-36 and SF-12
translations
and adaptions have been successful in 46 countries. The distinction
between
physical and mental health plays a key role in the SF-36 based scales.
Ware noted that there appear to be different correlations among the
individual
items among Japanese respondents relative to respondents in the United
States and Europe. These differences indicate that the physical and
mental
health do not the mean the same thing to a Japanese and American person.
Ware concluded with some observations on some of the exciting
opportunities
that the internet affords.
These
discussions
have all raised, either implicitly or explicitly, the central issue of
how one should interpret self-reported health status indicators. As an
example, consider the first question in SF-36. "In general, would you say
you health as (1) excellent, (2) very good (3) good (4) poor or (5) very
poor?" Does this question mean the same thing to all respondents? Would
the answer from someone who is at the peak of good health in his or her
twenties be comparable with the response from someone one who is slowing
down in his or her sixties or seventies? The question does not provide
an explicit comparison group. Should you compare your health now with
health
at some other time and, if so, what other time? Should it be the recent
past, the distant past or even the future? Or should you compare yourself
with someone else in the population? If so, who is that person? The
average
health of someone in your age group, in your socio-economic group, in
your
circle of friends or simply your perception of the average of the
population
in general. Clearly,
measures of general health status (GHS) and morbidities capture not only
underlying health status but also knowledge about health status and that
knowledge tends to vary in systematic ways with SES. In a relatively
homogenous
population where variation in knowledge is relatively minor, this will
be a second order problem.
However, validation of these sorts of questions in international settings suggests that in many low income settings, the issue is not negligible. It is, for example, not uncommon to find that health, as measured by GHS, is worse among the better off in a society relative to the poorest when health. A similar finding has been reported for morbidities, especially those that are based on diagnoses which are often known only after some interaction with a health care professional. Use of health care is positively correlated with income and SES in most setting and this correlation is particularly strong in low income countries with health care access is not universal. Thus, it is not surprising that knowledge of ailments is greater among higher income groups.Activities of daily living (ADLs) and Instrumental ADLs (IADLs) are thought to ameliorate some of these shortcomings.
John Strauss explained that ADLs ask about difficulties with specific tasks that respondents are likely to undertake on a daily basis -- such as difficulty walking a certain distance, climbing stairs, dressing. While there are likely to be differences in perceptions about what is meant by "difficulty", there is less scope for heterogeneity in interpreting "walking 100 metres". Strauss observed that the correlations between difficulties with many ADLs and indicators of SES tend to indicate the better off are in better health although he noted that does not mean ADLs do a good job capturing underlying health -- rather they do not appear to have as grave problems in interpretation as GHS and morbidity information. Nonetheless, they do suffer from two serious shortcomings as they are currently conceived in most surveys. First, the ADLs that have been extensively used tend to do a poor job of discriminating among reasonably healthy people unless they are fairly old; few prime age adults have difficulty dressing or bathing themselves. Strauss encouraged thinking about identifying activities that might help yield information about health of these people. Second, as John Ware point out, ADLs are context sensitive. It makes no sense to ask a respondent whether he or she has difficulty climbing stairs if stairs are not part of the person's normal routine. It is important, therefore, to design the questions so that they are relevant for the respondent.It was also noted that direct observation might provide useful information that complements responses to ADL questions.
For example, tests of
mobility
such as time to walk a distance or time to stand from a sitting position,
may yield insights into the links between this array of indicators,
underlying
health status and SES -- and how those links vary across age, gender
and ethnicity. Raynard Kington noted that both ADLs and mobility tests are
included in NHANES; they are also included in the MHSS and there is one
mobility test in IFLS. Duncan Thomas noted that even timed tests should
not be interpreted without caution; they will not only capture health
status
(or difficulties with certain activities) but they are likely to pick
up enthusiasm and exuberance as well.
-
Itemized expenses:overnight hospital stays, nursing home stays, doctor visits, outpatient surgery, dentist visits, home health care, special facilities or services
-
Total out-of-pocket expenses for three categories:
-
overnight stays (hospital and nursing homes)
-
health center/hospital visits and outpatient services
-
home health care and special facilities or services
-
(using
brackets
for "don't know" or refusals in each case to reduce the incidence of
missing
values). MEPS
has a long battery of questions that are thought to provide the "gold" standard in this literature. The astonishing result is the fact that the
abbreviated set of questions in HRS yield estimates of expenditures that
are remarkably close to those in MEPS. This suggests that even relatively
simple questions do a very good job of capturing the bulk of individual
health expenditures. Moreover,
Hurd pointed out that the correlation between health expenditures, use
of health care and health status and follow the kinds of patterns one
might
expect -- both for total expenditures (including insurance) and for
out-of-pocket
expenditures. The patterns reported in the RAND Health Insurance
Experiment
are largely reflected in the HRS. Contrasts
between the United States and low income societies were brought out by
Elizabeth
Frankenberg
. Drawing on her experience collecting data on health expenditures
and prices at both the household level and at the facility level in
Indonesia
she noted that the "price" of health care encompasses substantially more
than the pecuniary costs associated with care but also involve the costs
of getting to health care providers and the lost income associated with
seeking out care. In
low income settings, health insurance remains relatively rare. Thus,
health
plans (or health insurance coverage) is not the primary determinant of
care seeking behaviors. Indeed, it is typical to observe people using
multiple
care options either sequentially or possibly even simultaneously.
Moreover,
it is often difficult to classify providers into well-defined sub-groups
-- many providers take on multiple roles. In addition, providers of
traditional
care are an important part of the health care delivery system in many
countries,
and they are often excluded from surveys. Probing all sources of care is
critical for any survey interested in learning about use of health care
in low income settings. Collection
of information on health care expenditures is complicated by two facts.
First, payment schemes are often complicated and can involve payments
over
extended periods of time, can involve payments that are conditional on
outcomes (such as being cured) or subsidies for particular types of care
during a limited time interval. In Indonesia, for example, in some areas
it is possible to obtain a health card which provides free health care
for a period of time; the mechanism whereby those cards are allocated is
not entirely clear and so it is key that surveys collect information on
these sorts of institutions in order to interpret reported health
expenditures.
Probing for expenditures on different types of care is key: spending on
traditional care is seldom perceived as expenditure on health care per
se. Second,
part of the cost of health care is the cost of getting that care. In many
low income settings, the travel costs and time costs of getting to a
health
center are large -- and dominate the costs of getting care once at the
health center. Waiting times are often long at public health facilities
-- and they should be included in the (expected) time costs associated
with getting care. Moreover, the majority of the population in most low
income countries do not have jobs that include benefits such as sick
leave.
For these people, time spent at a health center is time away from earning
income. These
non-pecuniary costs are individual-idiosyncratic and depend on the
distance
from the respondents home (or place of work) to the health facility and
the value of that person's time. While computation of those costs is far
from straightforward, ignoring them is likely to yield results that
provide
little by way of insights into the determinants of health-seeking
behaviors.
Note also that these costs also likely affect the choice of type of
health
care and whether or not someone will seek health care at all in
systematic
ways. A
related but more subtle issue revolves around measurement of the quality
of health services. In principle, the price of health care should depend
on the quality of that care; the best studies of the effects of prices
on use of care (and health outcomes) will adjust prices for quality. This
is not easy. It is, however, potentially critically important given the
tremendous heterogeneity in quality of care that is commonplace in most
developing countries. Frankenberg
described the methods used in IFLS2 to collect information on health
expenditures,
prices and perceptions of quality of health services at the individual
and household level. She described how IFLS interviewed community
informants
to obtain information on the availability and quality of health services
in the community. She briefly described how health facilities were
sampled
(randomly from a listing of all facilities that are in the catchment area
for respondents in the enumeration area based on the lists provided by
household and community informants) and the information that was gathered
at each of the health facilities that was visited. This included prices
of services, availability of services and a broad array of
characteristics
that are likely to be indicative of quality. An
important element of health status is nutrition. This suggests that the
relative price of foods will influence health outcomes. Frankenberg noted
that food prices are collected at both the household and community level
in IFLS -- drawing on up to 6 different informants at the community
level.
She also pointed out that data on prices are typically readily available
from central statistical agencies and those data might productively be
matched with household survey data. It
is abundantly clear that collection of "prices" of health care services
is very difficult -- whether it be in a low income or higher income
environment.
Frankenberg proceeded to highlight one of the advantages of collecting
longitudinal data on household and facilities: it may be feasible to
measure
changes in "prices" better than measuring levels of prices. For example,
during the hiatus between the 1993 and 1997 waves of the IFLS, there was
a very large increase in investments in a program of village midwives who
are intended to serve the needs of reproductive-age women and their
children.
Frankenberg suggested that contrasting changes in health care use and
health
outcomes of reproductive age women between 1993 and 1997 in areas where
a village midwife was introduced with women in areas where no village
midwife
existed could provide some insights into the effect of the associated
change
in the "price" of health care. Moreover, to control for other changes
that
are occurring (such as changes in investments in health infrastructure
more generally), one could contrast this difference with the difference
in health status of older women (who would not be the focus of the
village
midwife) and also of men. Similarly,
with the economic crisis in Indonesia, the price of public health
services
has increased dramatically between 1997 and 1998 while the quality of
services
have declined. In the private sector, prices increased less and there is
little evidence of quality changes. These provide arguably exogenous
changes
in the "price" of health services and they can potentially serve to
identify
the effects of prices on health care use as well as health outcomes.In
order to exploit these sorts of "shocks" -- which are all too common
nowadays
-- we need to have on-going large scale surveys in place prior to the
policy
change (or economic shock) as well as follow-on surveys after the policy
change (or economic shock). The surveys need to collect a very broad
array
of indicators of health service availability, prices and quality at the
facility and community level along with matched household surveys that
collect information on use of health care, health status, health
expenditures
and socio-demographic characteristics of respondents. This is not easy
-- but without a concerted attempt to build this sort of capacity,
researchers
and policy makers will continue to handicap
themselves.
While we
cannot
possibly do justice to the richness of new ideas that emerged, the range
of points of views expressed and the quality of the discussion that took
place during the workshop, we have sought to provide an overview of some
of the main themes that emerged. Wrapping up this phase of the workshop,
James
P. Smith
opened the bidding by identifying the key unresolved issues
before us. Much
has been learnt in recent years and there has been a spectacular increase
in the quantity and quality of scientific data on health and socio
economic
status in the last decade. However, Smith argued, much remains to be
learned
and we will need to seriously re-think our strategies for the collection
of data if we are going to make progress in some key dimensions in which
our understanding of the processes underlying health outcomes is
seriously
lacking. It
has been argued that health is a stock which reflects the cumulative
effect
of life events; there are important feedbacks between life events and
health
outcomes and between health status and life choices. If this is true,
then
the evolution of a person's health depends critically on its past
trajectory.
Knowing something about a person's health at a single point of time
cannot
possibly provide us with the sort of information that we need to reliably
predict the future trajectory. Smith
called, therefore, for increased investments in longitudinal surveys;
echoing
Jennifer Madan's introductory comments about discussions within NCHS
about
the feasibility of conducting cohort studies along the lines adopted in
Britain, he applauded the work of Barker, Wadsworth and Marmot. He
suggested
that longitudinal surveys should not be the monopoly of higher income
countries
but that those which are started now in low income settings would likely
have a tremendous pay off as it is in those countries that there is
spectacular
change on-going. We
cannot wait 40 years for these studies to yield the kind of cohort data
that is available in the United Kingdom. Smith thus also called for
experimenting
with the collection of retrospective data on health. He suggested that
the survey research literature on recall error indicates that it is
possible
to collect high quality information on salient life time events: these
might include hospitalizations, major illness during childhood or as an
adult, the health and survivorship of siblings and parents. Incorporating
this sort of information into cross-section surveys could substantially
enhance their value in developing models of health transitions and
understanding
behaviors of individuals. Many
are skeptical that this sort of information can be collected. Smith
argued
that skepticism is largely based on ignorance: there have been very few
careful attempts to collect this sort of data and even fewer evaluations
of the data. The question should not be can we collect this information?
Rather it should be what information can we collect retrospectively? And
how should that information be interpreted? It would be naive to think
that prior health status will be measured accurately -- but will measures
that are correlated with prior health status be measured sufficiently
accurately
to be useful for analysis and policy making?Smith noted that there has
been a very substantial increase in the health content in many
household-based
socio-economic surveys. This is reflected in the HRS/AHEAD, several
European
surveys and in many of the surveys conducted in developing countries. It
is clear from the workshop that there are forces in place that will
likely
result in this trend continuing. In contrast, however, there has been
rather
little movement towards collecting indicators of socio-economic status
in health surveys. He suggested that in those surveys, the marginal
product
of yet another health measure is likely to be much, much smaller than the
marginal product of measures of resources such as income and wealth. The
links between health and SES are surely at the center of much of the
debate
in health care -- and health surveys are not being left out in the cold
because they simply do not have adequate reliable information on SES.
Smith
called for bringing people from outside the health discipline to the
table
when designing health surveys. Kington responded by noting that NHANES
has taken that criticism seriously and is currently piloting an SES
module
which is estimated to take about 20 minutes. The
workshop moved on to discuss the specifics of several surveys conducted
in low income settings in an effort to provide a sense of the practical
lessons learned from the fielding of those surveys. Greg Pappas and
Omar
Kahn
described the Pakistan National Health Survey; Kelly Hallman
described
the work conducted by IFPRI in Bangladesh; Randall Kuhn
provided
an overview of the Matlab Health and Socioeconomic Survey and Elizabeth
Frankenberg described the Indonesian Family Life Survey. Kathleen
Beegle
and
Bondan
Sikoki
reviewed some of the ideas that have been suggested for the
next wave of the Indonesian survey and they were given extremely valuable
feedback by the experts in the workshop. Dean
Jamison
ended the workshop and called for the development of an informal
working group that would provide a forum to share both ideas and
experience
with regard to the collection and analysis of survey data that would
subsequently
inform research and policy making. The goals of this forum would be
three-fold:
-
To facilitate enhancing the quality of research in the field by bringing together people from disparate fields; these people would not ordinarily interact with each other and yet, given the complexity of the issues at hand, there would be considerable benefit to everyone involved.
-
To encourage new and creative thinking that spans disciplinary and locational boundaries and foster an environment that supports risk-taking in research studies. This would involve taking calculated risks in both the collection of new, innovative survey data and in the analysis of those data. He challenged the workshop participants to think about how the richness of survey data could be brought to bear on the issues that are germane in the macroeconomic growth literature.
-
To help disseminate the work that is on-going and the lessons that have been learned both in terms of feasibility and desirability of fieldwork strategies and in terms of scientific results that are of substantive interest to the research and policy-making forum.
RAND/UCLA/WHO
WORKSHOP
on
HEALTH
STATUS
MEASUREMENT IN SOCIAL SURVEYS
AGENDA
RAND, Santa
Monica
October 16-18,
1999
PART I: HEALTH STATUS MEASUREMENT AND INTERPRETATION, SATURDAY OCT 16
INTRODUCTION
AND OVERVIEW
09.00-09.30
Duncan Thomas
- Introduction and welcome
Ritu Sadana,
WHO - Goals of the workshop
Jennifer
Madans,
NCHS - NCHS approach to collection of health
indicators
HEALTH
OVER THE LIFE COURSE: MEASUREMENT AND
INTERPRETATION
09.30-11.00
Chair:
Elizabeth
Frankenberg
Barbara Cohn
- Evidence from US
Michael
Wadsworth
- Evidence from Europe
General
discussion
INTERVIEW
BASED ASSESSMENTS
11.00-13.15
Chair:
Elizabeth
Frankenberg
Bob Wallace
- Health status in the HRS
Jane Gentleman
- Health in the HIS
John Ware -
Use of health scales
General
discussion
ADLs
14.15-15.15
Chair:John
Strauss
- ADLs and physical assessments
General
discussion
NUTRITION
15.30-17.30
Jean-Pierre Habicht - Nutrition and
health
General
discussion
18.30
Reception at
the home of Kin Bing Wu and Dean Jamison
SUNDAY OCT
17
CLINICALS & SPECIMENS
9.00-12.30
Chair: Duncan
Thomas
Theresa Seeman
- Experience in US: MaCarthur project
Raynard
Kington
- Experience in US: NHANES
Greg Pappas
- Prospects, problems and prerequisites in LDCs
General
discussion
HEALTH
EXPENDITURES AND HEALTH PRICES
13.30-15.00
Chair: Ritu
Sadana
Dana Goldman
- Experience with NMCES and Admin. data
Mike Hurd -
Experience in HRS
Elizabeth
Frankenberg
- Experience in LDCs
Jim Smith -
Critical unresolved issues
General
discussion
PART II:
MEASUREMENT
OF HEALTH: RECENT AND PLANNED SURVEYS IN LOW INCOME
POPULATIONS
The second part of the workshop will focus on practical issues that are germane to surveys in low income populations. A goal is to bring together the information that has been laid out during the prior day and half along with recent field experience with health measurement and use that knowledge to inform the next generation of health surveys in LDCs.
LESSONS
FROM THE FIELD: feasibility, costs and benefits of different
approaches
15.30-17.30
Chair: John
Strauss
Elizabeth
Frankenberg
- experience from Indonesia
Kelly Hallman
Bangladesh
Omar Khan (by
speaker phone) Pakistan
Randall Kuhn
Bangladesh
General
discussion
MONDAY 18
OCT
LOOKING INTO THE FUTURE
09.00-12.00
Review of what
has been learned from the workshop with specific reference to forthcoming
surveys in low income populations. Informal discussion about the big
questions
that need to be addressed and the approaches that should be considered
in surveys currently being planned in Bangladesh, Indonesia, Malaysia and
Mexico. Financial
support for the workshop was provided by the World Health Organization,
the National Institute on Aging (through the Aging Center at RAND) and
the National Institute Child and Human Development (through the
Population
Research Center and the Center for Research on the Family and Development
Center at RAND).
RAND/UCLA/WHO
Workshop
on
Health
Status
Measurement in Social Surveys
Participants
-
Victoria Beard, RAND
-
Kathleen Beegle, RAND
-
Alok Bhargava, University of Houston
-
Jay Bhattacharya, RAND
-
Virginia Cain, National Institutes of Health
-
Barbara Cohn, Public Health Institute, Berkeley, California
-
Eileen Crimmins, University of Southern California
-
Julie DaVanzo, RAND
-
Elizabeth Frankenberg, RAND
-
Jane Gentleman, National Center for Health Statistics/CDC
-
Dana Goldman, RAND
-
Jean Pierre Habicht, Cornell Unviersity
-
Kelly Hallman, International Food Policy Research Institute
-
Paula Hamilton, RAND
-
Michael Hurd, RAND
-
Dean Jamison, UCLA and World Health Organisation
-
Omar A. Khan, Johns Hopkins University
-
Raynard Kington, National Center for Health Statistics/CDC
-
Randall Kuhn, RAND
-
David Kurth, RAND
-
Jennifer Madans, National Center for Health Statistics/CDC
-
Joyce Mann, RAND
-
Gregory Pappas, Office of the Assistant Secretary for Public Health and Science
-
Endang Pudjani, Lembaga Demografi
-
Iip Rifai, RAND
-
Luis Rubalcava, Centro de Investigacion y Docencia Economicas (CIDE), Mexico City
-
Ritu Sadana, World Health Organization
-
Muda Saputra, (Ciput) Lembaga Demografi
-
Sondi Sararaks, Public Health Institute, Kuala Lumpur, Malaysia
-
Narayan Sastry, RAND
-
Teresa Seeman, School of Medicine, UCLA
-
Bondan Sikoki, RAND and University of Port Harcourt, Nigeria
-
James P. Smith, RAND
-
John Strauss, Michigan State University
-
Paramita Sudharto, Ministry of Health, Jakarta, Indonesia
-
Sukria Sumantri, (Cecep) Lembaga Demografi
-
Wayan Suriastini, Lembaga Demografi
-
Graciela Teruel, Universidad Iberoamericana, Mexico City
-
Duncan Thomas, RAND and UCLA
-
Michael Wadsworth, Univesity College, London
-
Robert Wallace, University of Iowa
-
John Ware, New England Medical Center
-
David Weir, University of Chicago
-
Robert Willis, University of Michigan
-
Kin Bing Wu, The World Bank
-
Peter Yau, UCLA
-
Julie Zissimopoulos, UCLA
Top