Finding medical / health care statistics online
To browse other articles on a range of HSL topics, see the A-Z index.
Finding (and using) medical and health care statistics is a major almost-daily challenge health librarians contend with in their professional work. Health statistics is defined as "...numerical data that characterize(s) the health of a population and the influences that affect human health" (Friedman, 2005). Health researchers often use health statistics to design, implement, monitor and evaluate specific clinical research outcomes, health programs and services. Health librarians must know how to find these statistics, and to teach health researchers how to extract data accordingly (or refer to a relevant expert as required).
Searching for health statistics "...is arduous, painstaking, multi-channel, overly-Googlized" and may be due to the fact that health statistics can refer to datasets, types and formats, and a range of epidemiological, health utilization and personnel statistics. Many health statistics worldwide are now open access (see also open data) and some are findable via Google or Bing search engines. Nonetheless, expert skills are required to know which websites are "trawled" for statistics and which databases must be searched in their native interfaces, especially when materials are stored in the deep web. For easier intelligibility, health statistics may be presented visually using tables and graphs, or accompanying narratives (sometimes narrative is covered separately). The most common statistical health information requested is disease prevalence in populations, hospitalizations and procedures, and vital statistics such as births, deaths, marriages and divorce rates. Disease incidence within populations includes flu outbreaks (see Influenza A(H7N9) virus), pandemics and morbidity and mortality statistics (number of people who die of certain diseases compared to total population). Other common statistical information are the economics and costs of health care, demographic distribution of diseases based on ethnic background, gender, socioeconomic status and education.
In public health, there is some demand for information that is created right now on the web and in social media spaces. To read a bit more about this trend, see Clinical surveillance technologies & mashups in public health. In 1947, Brodman et al published Brodman E, Pheulpin FJ, Deutschberger J. Some statistical methods useful to the medical librarian. Bull Med Libr Assoc. 1947 Jan;35(1):7-57. which is still relevant today.
Reference interviews involving health statistics
Online courses & pathfinders
Cool statistical aggregating tools
BC health statistics sources
Canadian health statistics sources
In Canada, we have a Canadian Federal Depository Services Program but several organizations collect and aggregate statistics in health notably Statistics Canada, provincial health ministries and organizations such as the Canadian Cancer Society and the Canadian Institute for Health Information (CIHI). The Data Liberation Initiative is a rich source of health statistics but may only be accessible from universities. For a more detailed, specialized view of statistical sources in Canada, see Finding medical / health care statistics online II.
Key StatsCan health statistics
Increasingly, the Canadian government is placing its materials online. One of the places they are collecting government documents is http://publications.gc.ca. At this site, you may search and read fulltext online.
Some key American & International Health Statistics sites
Locating international health statistics is more difficult than finding data at a provincial or national level. Adequate expertise, particularly in developing countries, may not be available to collect health data as comprehensively as needed to produce reliable numbers. Data collecting varies considerably around the world, and comparative data is not always available. The World Health Organization and the United Nations are two important agencies whose missions and budgets allow for collection of international data. When looking for accurate health and medical statistics, you may want to make an appointment with a medical librarian.
Background on findability
A major challenge for all librarians, not only health librarians, is locating the most accurate, relevant statistics to answer reference questions.
Often statistical information is not in a form that is optimum for our users. Health statistics are widely scattered in the published literature and in many web locations; the challenge for anyone seeking the right answer is to know where to begin. The specific statistic, number, answer or information needed may be located deep within a report or a table, perhaps even in the deep web. Finding relevant personalized statistics, therefore, in a timely manner and in usable formats are major challenges. It could be said that all librarians need to know more about how health statistics are gathered and published in order to find them, and that strategic searching (planning, even) is paramount in finding the best numbers. It may also be important to establish whether the information needed is usually tracked by municipal, provincial or federal government bodies, and whether, say in Canada, Statistics Canada or the Canadian Institute for Health Information (two of Canada's major health statistical organizations), are likely to produce what your user needs. Further consideration must be given to whether or not the information will be freely-available or whether a university-based subscription is required. A lot of government information and data have moved from their print origins to electronic and born-digital formats. Formerly 'fixed' print publications are now almost exclusively available via websites that feature customizable self-service functionality. Occasionally these statistics sites will aggregate health statistics for a fee. Although health statistics can be customized for a fee, some statistics cannot be accessed except by those with high-level access to the required databases. Some of the most difficult to find health statistics are in the deep web. In those cases, health librarians can look for answers in the grey literature.
Browsing by call numbers
Recent statistics on specific topics are listed in the * 16 classification section of each subject area
For many health and hospital librarians, finding health statistics is made easier by being strategic and 'targeted' about where to look for the best numbers. Generally speaking, search strategies that take advantage of the “natural structure” of health statistics are most effective. (For example, look at the table at the top right of this entry.) Targeted, faceted browsing for specific statistics can be achieved by going to the most relevant government and health websites. For example, for statistics about cancer in Canada? Go to the Canadian Cancer Society statistics page.
In the literature, Berinstein (1998) suggests a three-step strategy to finding statistics: 1) selecting the right place to look, 2) devising an effective search strategy, and 3) evaluating the data. Some basic knowledge of the structure of statistics gathering in health and medicine is recommended for all health librarians. It's important to know, for example, that while countries gather statistics, Canada, the United States and various European countries as well as the World Health Organization produce comprehensive health statistics. Many other countries aim to do as well but do so in a piecemeal fashion. Consequently, the availability of statistics on particular health conditions is extremely variable unless it is linked somehow to major diseases and syndromes. It is possible to find statistics by searching PubMed, Google, journal articles or by directly contacting organizations connected with specific diseases. Some governments track vital statistics, infectious diseases, conditions requiring hospitalization and hospital admissions, etc. Statistics are often findable by using search engines - see diabetes or by consulting relevant websites likely to publish them. Remember some chronic conditions that do not require hospitalization are not tracked by federal or provincial organizations. Statistics on these conditions are difficult to locate since data collection and analysis are time-consuming and expensive.
Some things to remember about health statistics
Data collection in health care and health services utilization is often uncoordinated, fragmented, decentralized, slow and expensive. Health statistics are consequently a challenge to find; however, remember that:
Evaluating & citing statistics