|
Epi Wit & Wisdom Articles
Why Epidemiology is
Underutilized as a Tool For Decision-Making in Health Services
By Gilles Dussault
[Editor’s note: Reprinted with
permission from the Epidemiological Bulletin, Pan American Health
Organization, Abridged and revised for The Epidemiology Monitor.]
The usefulness of epidemiology
for decision-making in the management of health services is easy to
demonstrate. Decisions on resource allocation and on the definition of
priorities and objectives cannot be made without data to identify
problems and their distribution in the target population. The problem
arises in understanding why administrators fail to use epidemiology
and how one may foster more appropriate use.
Pressures to change this
situation have been growing in several areas:
1) Increasingly, in countries of
the Americas, policies are being defined according to goals and not
simply in terms of services to be offered and resources to be
mobilized. The objective of such policies is to have a measurable
impact on the health of a given population. This is a very different
objective than trying to meet people’s spontaneous demands for
services.
Such policies require reliable
and pertinent baseline data and acceptable criteria for selecting
priorities, among which one of the most fundamental is to have valid
intervention strategies. Epidemiology can help to validly describe and
explain the state of health and monitor the evolution of problems. It
can also contribute to an evaluation of the effectiveness of different
interventions.
2) Regional and municipal
decentralization and local health systems development favor the
population approach to health problems. The institutional and
professional approaches which focus on the individual may function
without the contribution of epidemiological data (it is enough to
anticipate the demand or, eventually, to foster it), but the
population approach requires such data. Its aims are expressed in
terms of indicators that need to be changed. Without the contribution
of epidemiological data on the evolution of the state of health, there
would be no population approach, unless we wish to consider the
impressions of managers, professionals or representatives of the
population as sufficiently reliable and valid sources of data on
health conditions.
3) The search for a more
equitable distribution of the resources available for health services
also requires an epidemiological contribution at two levels at least:
• Identification of differences
in epidemiological profiles of different regions and population
groups, which is required because equity implies a needs-based
resource allocation
• Evaluation of the impact of
different intervention options in order to avoid wasting limited
resources
If each disbursement represents
an opportunity cost (its equivalent in other disbursements forgone),
valid data on the relative effectiveness of decision options is
essential for the decision-making process, even at the level of
micro-decisions (e.g., to prescribe a drug, to order a test, to admit
a patient). Data on variations in medical practice cannot be ruled
out simply on the argument that no two patients are the same or
because the circumstances of practice have divergent characteristics.
Such a position would be tantamount to accepting that all physicians
are always right. Physicians and other health professionals need to
carry out rigorous analysis on those data in order to determine which
variations are acceptable and which should be eliminated. Every time a
cesarean section, bypass, prostatectomy or tonsillectomy is needlessly
performed, resources are diverted from needs for which effective
interventions might exist; in the context of publicly financed health
services, this is not ethically acceptable.
This position is idealistic to
promote the search for maximum compatibility between interventions and
resource allocations in terms of needs, effectiveness and efficiency.
It is not, however, utopian because we already have the technical
capability to bring us closer to the ideal. Our problem is that we do
not utilize this capability. Why?
Potential vs. Actual Practice
There are two main reasons for
the gap between epidemiology and management:
1) There are few instances in
which governments have adopted a true health policy and have the
authentic will to apply it. It is still rare for planning to be based
on goals formulated in terms of health indicators, and rarer still to
find coherence between organizational strategies and those goals.
Decisions related to resource allocation may be influenced by factors
unrelated to needs. There are economic factors tied to the interests
of equipment manufacturers and producers of other inputs,
pharmaceutical, insurance, and construction companies, consulting
firms, and obviously, politicians. Technical decision criteria—among
which epidemiological criteria would be the most important—play a more
significant role when there is a true commitment from decision-makers
to the objective of changing health conditions. Without this
prerequisite, the notion that epidemiology will have an important
impact on decisions is mere wishful thinking. In addition to producing
pertinent data, epidemiologists should participate in the movement
that sets the promotion of improvement in health conditions as a major
social priority.
Even when the political will
exists to implement a health policy, clear goals cannot always be
defined because of a lack of information.
2) Despite the fact that
circumstances do not always favor the best use of epidemiology in
health service management, one must admit that the behavior of
epidemiologists and managers also contributes to the problem. Both
have different, but not necessarily divergent visions of what is “good
information.” Epidemiologists are concerned with problems related to
the validity of numerators and denominators, the credibility of data
collection tools and the scientific quality of analyses. This is
normal and desirable. Many in the profession are more comfortable with
variables that lend themselves well to quantitative measurements, and
so they tend to reduce health problems to their biophysical
dimensions. That component of epidemiology, tends to give short shrift
to the health perspective and the qualitative methods of the social
sciences. “Hard” epidemiologists who only see disease as the
manifestation of a difference between an observed situation and
professionally defined norms easily dismiss the opinion of
sociologists and anthropologists that health and disease are cultural
products (products of representations) rather than having only a
biophysical basis. Epidemiological development has been occurring more
in relation to academic criteria than to the needs of service systems.
In Latin America, epidemiology
has been more concerned with the sociocultural aspects of health
problems. But my impression is that the field has had little contact
with health services management.
The result is that if
epidemiologists were to assume responsibility for management they
might be astonished to discover the differences between the type of
information they need to make management decisions and the information
that they actually produce. On the other hand, one finds scant
familiarity among managers with the contributions that epidemiology
can make to management. In North America, the great majority of
managers know little about the potential contributions of
epidemiology.
For managers, useful information
is that which covers the population serviced, is quickly accessible,
easy to interpret, and inexpensive. Managers have little awareness of
the difficulties that hamper or often prevent the production of such
data. They have a more institutional than populational perspective
which explains their limited interest in epidemiological data.
Consequently, epidemiologists have an educational task to fulfill.
They need to explain the limits of what is possible, the problems of
incompatibility among data sources, validity requirements, and
methodological difficulties in measuring certain variables.
Municipal and regional
decentralization favors the horizontal and vertical integration of
institutions that provide services (an integration that requires a
very strong political commitment). As resource allocations are linked
to the distribution of needs among the population, managers will begin
to become concerned about health information and to voice demands for
epidemiological expertise. Consequently, they will not be satisfied
with traditional indicators; they will also want information on
problems that epidemiologists disregard: mental and social problems,
family and urban violence, drug abuse, and the effectiveness of
intervention options. Managers have the responsibility to better
define their needs and explain to epidemiologists the role that
socio-health information plays in the decision-making process. For
their part, epidemiologists should be prepared to respond to these
requirements if they want to maintain their professional standing.
Linking Epidemiology and
Management
Decisions in the health sector
are being based on numerous factors, such as the requirements and
preferences of users, professionals and managers (not necessarily in
this order). Furthermore, other factors come into play such as
political and economic interests, power relationships among
participants in the decision-making process, the costs and
availability of resources, perceived needs and measures. Our concern
is to conceive strategies to broaden the relative contribution of
health and social data to decision-making.
Quebec’s attempts to guide
resource allocations in this way are recent but they demonstrate that
actors in the health service system have quickly appreciated the need
for epidemiological data in order to justify their resource
requirements. Professionals and managers are seeking relevant data,
now that they understand that the rules of the game require
documentation of the need for and utility (relevance) of proposed
health services.
It would be most helpful to
strengthen the education of management in the discipline of
epidemiology and that of epidemiologists in the field of management.
Administrators should know the language of epidemiology, how it works,
and what its limitations are. The mistake that must be avoided is to
teach epidemiology to managers as if they were to be trained to become
epidemiologists. Epidemiology should be taught, instead, as a
management tool, as a decision-making aid. It should foster managers’
adoption of a population approach to health needs, and it should
provide strategies for identifying those needs. On the other hand,
epidemiologists should be sensitive to management needs and produce
pertinent and useful data, presented in a form that increases the
likelihood of it being used. Therefore, they should learn to
communicate information.
To my mind, the current problem
is not one of scarcity of resources, but the poor utilization of
resources. Managers complain that they do not have access to data they
would like to use and that the data they do receive are not relevant.
This characterization may be a bit overstated, but it does reflect the
differences in perception between administrators and epidemiologists
over what is relevant. Inclusion of epidemiologists on management
teams can contribute a great deal to the process of defining the
needs, priorities and strategies of intervention and evaluation. Such
a proposal does not mean that epidemiologists should be subordinated
to management requirements and act only as data-supply technicians.
Epidemiology should conserve its role of critical analysis of policies
and decisions in the health sector; managers would also benefit from
recognizing the role that should be fulfilled in evaluation
activities.
Conclusion
Epidemiology is not, nor will it
become, a substitute for decision-making. Its role is to introduce
more rationality into the process. It has numerous potential areas of
influence:
1) in public health policies,
helping to define priorities, objectives and strategies
2) in the reconfiguration of
services, examining the consequences of decentralization, out-patient
surgery, reducing admissions, and integrating services into programs
3) in the professional
practices, studying variations in effectiveness and efficiency
4) in management practices
5) in research priorities
These contributions are
necessary both in the context of declining available resources,
characteristic of rich countries, as well as in the context of
increasing investments in the health sector, which is occurring in the
Latin American countries that have been controlling inflation and have
undergone growth. Consequently, the challenge for both epidemiologists
and administrators is to achieve the type of alliance that produces
policies and strategies that have a greater impact on the well-being
of populations.
Published December 1995 v
|
