FAQs (Data Related) Data Sources
- What is the best source for state arthritis data?
- What are the best sources of data for arthritis prevalence for the nation?
- Where should I go for help when I have a question about using NHIS/BRFSS arthritis data?
- Does the CDC recommend using Medicare and Medicaid data for arthritis surveillance or to generate cost estimates?
- Does the CDC recommend using pharmacy data for arthritis surveillance or to make cost estimates?
- Does the CDC recommend using Veterans Administration (VA) data for arthritis surveillance?
- Does the CDC recommend using hospital discharge data for arthritis surveillance?
- Does the CDC have plans to collect or analyze data on juvenile arthritis?
- How do BRFSS arthritis prevalence estimates compare with those from NHIS?
- Where can I learn more about the Arthritis Conditions Health Effects Survey (ACHES)?
We recommend using data from the Behavioral Risk Factor Surveillance System (BRFSS) for state estimates of the burden of arthritis.
A note about combining years of data.
Different questions and case definitions were used to collect data between 1996–2001 and from 2002 forward. For this reason, it is not valid to look at trends that cross from 2001 into 2002 forward; survey changes occurred starting in 2011. For these reasons, it is not valid to look at trends that cross from 2001 into 2002 or from 2020 into 2011 and later. An overview of BRFSS arthritis question changes is available. Read more about the arthritis case definition.
View the summary of these state-specific 2003, 2005, 2007, and 2009 BRFSS prevalence data. Please see FAQs (Data Related) Data Analysis.
The best source for national arthritis prevalence estimates is the National Health Interview Survey (NHIS), an annual survey conducted by the National Center for Health Statistics. Each year, the NHIS samples U.S. households and gathers information on select adult and child members living in each household. Estimates of health conditions and behaviors from the NHIS are representative of the U.S. civilian, noninstitutionalized population, and are used for many purposes, including Healthy People 2020.
In 2002, new questions (identical to the 2002 BRFSS Arthritis Burden Questions) were included on the NHIS survey to estimate arthritis prevalence in the United States. Cases are defined in the same way as in the BRFSS. Self-reported, doctor-diagnosed arthritis is estimated using the question: “Have you ever been told by a doctor or other health professional that you have some form of arthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia?”
Your first step should be to review the questions and answers on this page. For questions about the mechanics of using the BRFSS, you should check with your state BRFSS coordinator. A list of links and contact information for BRFSS and NHIS are available below.
- BRFSS FAQs
- BRFSS User’s Guide [PDF–986K]
- BRFSSS State Information
- BRFSS State-by-State Listing of How Data Are Used
- National Health Interview Survey (NHIS)
- National Center for Health Statistics
Toll-Free Data Inquiries: 1–866–441–6247
For questions about the arthritis case definition, concepts, and interpretation, contact us at the Arthritis Program at at CDC-Info.
No. Dr. Ken Powell from the Georgia Department of Human Resources explored the use of these data (Powell KE, Diseker RA, Presley RJ, et al. Administrative data as a tool for arthritis surveillance: Estimating prevalence and utilization of services. Journal of Public Health Management and Practice. 2003;9(4):291–298). He found one can obtain different answers, depending on 1) how variables are defined, 2) the number of years included, and 3) how enrollment is defined. He also found that using these data was very labor intensive and was difficult to justify.
In the absence of a clear link to a project, or to your state arthritis plan, we recommend you do not analyze Medicaid or Medicare data, especially if it takes program resources.
In 2013 NCCDHP created a new Chronic Disease Cost Calculator that uses Medicaid, Medicare, and other information to provide state-specific costs estimates for (year/s) for a variety of chronic conditions, including arthritis. Learn more about the Cost Calculator.
No. One problem is that very few arthritis medications are used only for arthritis; most are used for other conditions as well. For example, some of the biologics can cost $15,000 to $20,000 a year and are used for other conditions like inflammatory bowel disease. Another example, methotrexate is a drug used for many types of inflammatory arthritis, but for non-arthritis conditions such as psoriasis as well. This is a problem that is unlikely to change soon.
No. VA data would be relevant only to the special population using the VA, and cannot be generalized to the whole population.
In general the answer is no. The data are not useful for community prevalence estimates because the group of people with arthritis who are hospitalized as a skewed (and therefore biased) sample of people with arthritis.
At the national level, we have determined arthritis impact on hospitalizations by looking at the National Hospital Discharge Survey. At the state level, similar impact studies could be done if there were a complete source of hospital discharges. However, states are discouraged from pursuing these activities unless there is a direct link to program activities and they are consistent with state arthritis plans.
Estimating arthritis prevalence and impact among children and teens is very difficult. Juvenile arthritis is relatively uncommon and it is not sensible to try to collect information through the usual self-report surveys because it is so rare. After considerable effort, working with experts in this contentious field, we have an approach that provides prevalence estimates for a variety of arthritis-related conditions including juvenile arthritis. This approach can also monitor national prevalence over time. Sacks JJ, Helmick CG, Luo YH, Ilowite NT, Bowyers S. Prevalence of an annual ambulatory health core visits for pediatric arthritis and other rheumatic conditions—United States, 2001-2004. Arthritis Care Res 2007;57(8):1439-1445.
An in-house CDC analysis of the two surveys, which used identical questions in 2002 and 2003, generated different national estimates. Some of the reasons may be:
- Different survey modes: NHIS is administered in person, for the most part, while BRFSS is a phone survey.
- Additional, interspersed questions in NHIS: Compared with the BRFSS, the NHIS in 2002-2003 interspersed some additional questions about arthritis that might influence respondents’ answers.
- Cumulative error: from combining 50 independent BRFSS surveys.
For this reason, we recommend using the NHIS, which is designed to generate national estimates, for making national estimates, and using the BRFSS, which is designed to generate state-specific estimates, for making state estimates.
The Arthritis Conditions Health Effects Survey (ACHES), developed and funded by the Centers for Disease Control and Prevention, surveyed 2,238 persons with arthritis or chronic joint symptoms ages 45 years or older found by using a random digit dialing technique between June 2005 and March 2006. Data were gathered on symptoms, limitations, levels of physical functioning, effects of arthritis on work, knowledge and attitudes about arthritis, self-management of arthritis, physical activity, anxiety and depression, and demographics. Data from ACHES are now available for public use (see procedures below). Note: Although the ACHES questions are provided for you, it is important to read and understand the disclaimers regarding the use of certain questions in the survey. ACHES Screener, Consent and Questionnaire [PDF–361K]. See disclaimers at the end of this answer.
Procedure for using ACHES data: ACHES Data Documentation [PDF - 601KB]. Proposed research activities involving the use of ACHES data are reviewed and approved by the Arthritis Team. A special use dataset will be generated from the ACHES dataset. Data may be restricted based on completeness, quality and applicability to requested use, or sensitivity of release. To obtain access to ACHES survey documentation (screening and consent script, questionnaire (Spanish), code book, ACHES final report volume one and/or ACHES final report volume two), email us at firstname.lastname@example.org and specify the documents you require. To obtain access to the data applicants must complete the 1) project summary form [PDF–16K] and 2) data sharing use agreement [PDF–20K] and return them to:
Chief, Arthritis Program
Division of Population Health, Arthritis Team
NCCDPHP, Centers for Disease Control and Prevention
4770 Buford Hwy, N.E., Mailstop F-78
Atlanta, GA 30341-3724
Phone: (770) 488-5114 Fax: (770) 488-- 5964
Questions SF 1–10 are copyrighted questions from the SF36. They may only be used after obtaining a licensing agreement (see http://www.qualitymetric.com/products/SFSurveys.shtml).
Questions MOOD 1–12 are from the AIMS2 and according to Dr. Robert Meenan may be used free by government and academic researchers (http://web.archive.org/web/20080107032627/http://www.qolid.org/).