Public health data – from basic disease statistics to large datasets of de-identified data – are a vital resource to researchers, clinicians, and even patients.  However, these resources are not found in a single place but are spread out over a wide variety of government, international, and non-profit organizations.

Here’s where to locate and get access to some of the most useful health data.

General Demographic Statistics:

• US Census Bureau QuickFacts
  
  • QuickFacts provides statistics for all states and counties, and for cities and towns with a population of 5,000 or more.
• Population Reference Bureau (PRB) Data
  • Provides U.S. and international indicators as well as world population data

General Health Statistics:

• National Center for Health Statistics (NCHS)
  • Primary US government health statistics website, compiles data from a variety of government resources
    • FastStats provides quick access to statistics on topics of public health importance
    • Population surveys such as the National Health and Nutrition Survey (NHANES)
    • Certain restricted data (requires a proposal of why it is needed)
• CDC Wide-ranging Online Data for Epidemiologic Research (WONDER)
  • CDC WONDER is the access point for the CDC’s public data in an easy-to-use online query system
• Centers for Medicare & Medicaid Services (CMS)Research & Statistics
  
  • CMS Data Navigator provides a easy-to-use web interface to search for CMS statistics
  • Medicare Claims Public Use Files (PUFs) provides the raw de-identified Medicare claim data, from inpatient, outpatient procedures, prescriptions, and more
• WHO Global Health Observatory (GHO)
  
  • Data repository for the World Health Organization’s health statistics for all 194 member states, includes over 1,000 indicators

Cancer Statistics:

• American Cancer Society Cancer Facts & Statistics
  • Cancer Statistics 2020 (annual article from CA: A Cancer Journal for Clinicians)
  • Cancer Statistics Center is an interactive tool using ACS’s data
• CDC’s Cancer FastStats
• National Cancer Institute’s SEER (Surveillance, Epidemiology, and End Results)
  • SEER*Explorer is an interactive program to create individualized datasets using SEER data.
  • Cancer Query Systems include data from SEER on incidence, mortality, survival, prevalence, and probability
• CDC United States Cancer Statistics
• IACR Global Cancer Observatory (GCO)
  •  Part of the International Agency for Cancer Research at the WHO
  • The GCO is an interactive platform of cancer statistics from across the world

New York(City/State) Health Data:

• Health Data New York
  • A subset of OpenNY that provides access to health data from across New York State
• NYC Health Data
  • Community Health Profiles details health and other demographic information for all 59 community districts across NYC
  • EpiQuery uses data from a variety of surveys that can be customized to a variety of demographics

“We should look to the mind, and not to the outward appearance” said Aesop. Yet, outward appearance sometimes matters.

In publishing, the title and abstract of an article in a database (e.g. Pubmed, Embase, Web of Science), as well as some other information, such as listed your affiliation, represents the “face” of your article. It’s very important to make your title/abstract an adequate representative of your article, and many science authors do a good job putting a lot of essential information in the abstracts. Yet, an important aspect may be overlooked,  not just the information itself, but what makes it “findable,” or more exposed to the world.

Although some biomedical databases such as Pubmed and Embase assign controlled vocabulary (subject headings such as MeSH or Emtree terms) to the references, which aids others in finding your article, not all databases do so, and not for articles in any journal. In any case, even if your article is assigned subject headings in some databases, the text words in the database record, especially in the title, abstract and author supplied keywords matter a lot. They matter for finding your article by a database searcher, and they matter for how your article record is sorted in relevance ranking in search results. It’s in your power, as an author, to make your article findable easily via a database.

Here are some tips on how to make your publication findable via a database search:

• Use both abbreviations adopted in the field and the corresponding full names in the title, abstract and author keywords. Example: long non-coding RNA (lncRNA)
• Use special characters with caution as many of them can’t be used as search terms (e.g. plus and minus characters). It is best to use full words instead. For example, if a database user types HER2+ in the PubMed search box, the user will retrieve all articles on HER2, both positive and negative, as the “+” character is ignored by PubMed. At the same time, if a database user types HER2 positive in the search box, the user may never find your article if you only used HER2+ in the information that goes into a database record. If you used the words HER2 positive in the title, abstract or keywords instead or in addition to, your work would be easily found in PubMed by a searcher using the word positive when searching for HER2+ cancers.
• Some databases recognize Greek characters (e.g, if a searcher pastes the title of your article where you used the Greek character for Beta in the PubMed search box, they will find your article but if they paste this title in full in the Web of Science search box, they won’t find your article unless they type beta as a word.
• Use author keywords options offered by many databases to supply well established synonyms, alternatively spelled words and important additional terms not already used in the title or abstract to increase the chance of a database user finding your article if they search by terms not used by you in the title/abstract.
• Use the full name of your institution as your affiliation, not just the institution abbreviation, department or address.

Key takeaways:

Always think about the ways your article can be searched for and found in different major databases when you name your article, write the article abstract, supply author keywords, list your affiliation, list the sponsor/grant, etc.

The new PubMed is coming in 2020, and we are here to help you navigate the new interface. Here are five highlights:

• Improved mobile experience: You can now enjoy all PubMed’s functionality and features on your tablet or mobile device.
• Sort Order: “Best Match” is the default sort order for search results. It uses an algorithm to show the most relevant results first. PubMed lets you easily switch to “Most Recent”. Your browser will remember your sorting preference for any following web sessions (except if you clear your browser cache). As a personal preference, I like to toggle back and forth between both options, relevancy is important, but I also don’t want to miss any new studies.
• Citation: A great new feature allows you to copy and paste the citation of an article using the action “Cite”. You can choose from four citation styles: AMA, MLA, APA, or NLM.
• Enhanced retrieval of synonyms, plural forms, and British/American spelling: Please note that while this is a very convenient feature, PubMed may find keywords that are irrelevant to your search. For instance, when searching for the keyword “Nurse”, PubMed will automatically search for the keywords nurse, nurses, nursing, but also the synonyms breastfeeding(s), breastfeeder(s), lactation(s), lactating(s), and lactators. If your search retrieves too many irrelevant results, we will be happy to meet with you to demonstrate how to turn off synonyms retrieval. Come to one of our classes or schedule a consultation with one of us.
  
• Enhanced data discovery: PubMed added a new filter “Associated Data”, allowing you to retrieve articles with supplemental data from the following data repositories: clinicaltrials.gov, GenBank, figshare, and Dryad.

The new PubMed will become the default in spring 2020. The legacy version will run for several months in parallel before being fully retired. If you would like to familiarize yourself with the new PubMed interface, join us in one of our classes.