Health Statistics
Before starting a search for health statistics, it is helpful to understand how health and medical data is collected and what types of data are most commonly available.
Table of Contents
What are health statistics?
Top 5 challenges to finding data
A note on international data
Other statistical resources by subject at the UC Library
Glossary of Statistical Terms
What are health statistics?
Health and medical statistics incorporate a variety of data types. The most common statistics reported are vital (birth, death, marriage, divorce rates), morbidity (incidence of disease in a population) and mortality (the number of people who die of a certain disease compared with the total number of people). Other common statistical data reported are health care costs, the demographic distribution of disease based on geographic, ethnic, and gender variables, and data on the socioeconomic status and education of health care professionals.
Challenges
to finding data
Seekers of health and medical statistics should be aware of the following
challenges in their quest for data.
First, it is important to know that data collection in the United States
is a fairly recent phenomenon. For example, it was not until 1956 that
Congress enacted legislation to establish the US National Health Survey
in order to collect statistics on disease, injury, impairment, disability,
and other health related topics. Locating historical data (pre 1956)
will be more time consuming and consultation with primary resources
will be necessary unless an analysis has already been published on your
topic.
Second, data collection is decentralized. Numerous federal, state and
local agencies in additional to not-for-profit organizations are involved
in the collection and dissemination of health related data. As a result,
there may be little to no data in some areas and duplication of data
in others. There may also be variations between how data is collected
and described between these various agencies. For example, differences
in time periods covered, geographic areas, and sample sizes used to
tabulate the data will differ. In addition, pay special attention to
the definitions and coverage used by each reporting agency as the terminology
used may have different meanings. It is difficult to obtain a precise
figure when it comes to national samples. In most cases the numbers
reported are estimates. Also, reporting standards and definitions change
over time. If your figures seem higher or lower than expected, check
to see if this change reflects actual trends or if the definition or
reporting of your figure has changed. Finally, you may also see slight
variations in the numbers reported between different agencies. Check
to see whether the figures reported are compiled independently, or if
they were taken from other sources. For example, commercial publishers
often repackage government compiled data. In most cases, the secondary
reporting source has accurately reproduced the original data and conclusions,
but when in doubt, track down the original report.
Third, data collection on a national level takes money and staff to
compile. Only the federal government or a large organization has the
resources to organize and manage large-scale data collection efforts.
Finding data will be easier if you know who is responsible for collecting
data in the area you are studying and the type of data they collect
and publish. Government agencies collect data as part of a federal mandate;
private organizations expend resources as part of their organizational
mission or in response to the needs of the organization's membership.
In either case the collecting agency has predetermined the need and
collection scope for data in a specific area. It is possible that that
data you are looking for has not been collected, or it has not been
analyzed in the manner in which you need. You may need to obtain the
original data and perform your own analysis. This will be easier if
the data was gathered by a government agency, in which case the data
may be available upon request, in a library, or for purchase. However,
if the data was originally compiled by a private organization, it may
not be possible to obtain the original data sets.
Fourth, data collection and analysis at a national level takes time
to compile. Rarely will you find "real time" data except for estimates.
Most statistics are out of date to some degree by the time they are
published. Just because a report is published in 2000 does not mean
the data is from the most recent few years. Current health data is often
based on extrapolations of older data, such as U.S. Census data, compiled
every ten years, or on smaller sample sizes. Again, check to see when
and how the data was collected. In most cases you will probably see
at least a 3-5 year time lag between data collection and a published
report.
Fifth, and compounding the difficulty of locating data at the national or international level, there is a lack of adequate subject access through indexing sources (i.e. Medline). Most health data is compiled and distributed by the federal government. It is notoriously difficult to track down government resources since federal reports etc. are as not well indexed. When possible, see if the reporting agency has the statistics you need. If not, then you will need to search the journal literature in a bibliographic database like MEDLINE or Cancer.gov. It may
be difficult to always determine what type of statistical data is available
from a citation or abstract. When in doubt, read the article to make
sure that you are not missing potential data.
International
Data
Locating international health statistics is usually more problematic than locating U.S. data. Resources, particularly in developing countries, may not be available to collect data as extensively or comprehensively as in the United States. Data collecting efforts therefore vary considerably around the world and comparative data may not be available. The World Health Organization and the
United Nations are two international agencies whose mission and budget allow for international data collecting. Consult the publications of these agencies first when seeking health and medical statistics. See the Library's International Health Research Guide for additional help locating these elusive statistics.
Other statistical resources by subject at the UC Library
Education Statistics
Mathematical Statistics
Economic Data
Need statistical software or actual data?
For access to SAS, SPSS, Stata, etc., contact:
Social Science Research Computing
1155 East 60th Street
Room 049
Chicago, IL 60637
Voice: 773-702-0793
Fax: 773-702-2101
srchelp@src.uchicago.edu
