Vitalnet BRFSS
Query & Mapping Software

A. Introduction

BRFSS data are valuable: From the CDC BRFSS web site: "The Behavioral Risk Factor Surveillance System (BRFSS) is the world's largest, on-going telephone health survey system, tracking health conditions and risk behaviors in the United States yearly since 1984." The BRFSS is widely recognized to be a rich source of public health data. The BRFSS produces powerful data for tracking and comparing health risks.

BRFSS data are underutilized: As explained below, BRFSS data have many complexities. Thus, analyzing BRFSS data by hand using generic stats software is tedious, error-prone and expensive. And the complexity has mostly prevented the availability of adequate BRFSS query systems, despite large amounts of time and money spent on the effort. Basic results may readily be found in reports. However, going further, such as age-adjusted results, most subpopulation results, and getting publication-ready output, is quite difficult. The result is that BRFSS data are currently greatly underutilized.

How to maximize use of BRFSS data: Vitalnet BRFSS Query & Mapping Software (VB-QMS) is the Vitalnet module for analyzing BRFSS data. VB-QMS correctly handles the complexities of BRFSS data. It makes just about any map, table, or chart needed. VB-QMS provides the needed combination of flexibility, ease-of-use, and attractive output, in a professional software system. VB-QMS makes it much easier and more reliable to analyze BRFSS data, and lowers costs.

Use VitalWeb browser platforms

B. Handling BRFSS Complexities

1) Weighted data: BRFSS data are of necessity weighted. Therefore, analysis of the data must take the weighting into account. That rules out "normal" methods of data analysis. The solution: Vitalnet BRFSS Query & Mapping Software (VB-QMS) correctly analyzes weighted BRFSS data, the same as certain other software such as SAS is able to do. However, VB-QMS is much more than a generic stats software: VB-QMS "knows" all the details about BRFSS variables, and takes the needed steps to ensure correct results.

2) Many questions: There are hundreds of BRFSS questions. This is confusing to the analyst trying to analyze BRFSS data by hand with generic stats software. And it makes developing home-grown analysis software much more difficult. The solution: The VB-QMS internal data dictionary stores and integrates data from the many questions, enabling VB-QMS to correctly analyze the large number of questions.

3) Question combinations: Several questions may need to be combined for analysis, or several responses to a question may need to combined. So again, analysis by hand using generic stats software is tedious and error-prone. And again, designing home-grown analysis software is much more difficult. The solution: The VB-QMS internal data dictionary "knows about" BRFSS variables and how they interact, so the data are correctly imported and analyzed.

4) Wording changes: Question wording can (and does) change over the years. This makes BRFSS results possibly subject to misinterpretation, unless the output is clearly documented. This is tedious and error-prone to do by hand, or to incorporate within a home-grown BRFSS query system. The solution: The VB-QMS internal data dictionary keeps track of any question wording changes, and VB-QMS output is documented to show how question wording changes over time.

5) Sub-populations: Many BRFSS questions are only asked of a sub-population. Misinterpretation may result if the output is not clearly documented. This again makes analysis by hand with generic stats software tedious and error-prone. And again, this complicates the development of home-grown BRFSS query software. The solution: The VB-QMS internal data dictionary keeps track of sub-population questions. So Vitalnet correctly imports and analyzes the data. And VB-QMS output is documented to show the sub-population information.

6) Questions not asked: Many questions are not asked some years. Again, this makes analysis with generic stats software tedious and confusing. And it leads to awkward home-grown query systems. To analyze data for a different year or variable, the user often has to "back out", and start over again. Also, the system may not let the user combine data years, or do a time trend analysis. The solution: The VB-QMS internal data dictionary keeps track of skipped years. The VB-QMS interface lets the user easily combine years, never requires "backing out", and does time trend analyses.

7) Similar questions: Some questions can be confusingly similar. This can easily lead to analysis errors when the data are analyzed by hand, and again makes development of home-grown software more difficult. The solution: The VB-QMS internal data dictionary keeps track of all the questions, including similar questions. To prevent any confusion, VB-QMS output documents the actual question and variable name.

8) Split surveys: Split surveys lead to multiple weights for each record. This can make analysis by hand very tedious and error-prone. It also greatly complicates the development of any home-grown automated system. The solution: The VB-QMS internal data dictionary keeps track of which weight applies to which variable, so VB-QMS automatically produces correct results.

9) State-added questions: Each State may add State-added questions. These questions can be confusingly similar to CDC questions, and add even more questions to the number available. This is again confusing for analyzing the data by hand, and the results for State-added questions are normally not available in national reports. The solution: VB-QMS handles State-added questions like any other question. Any State may include State-added questions in their VB-QMS system.

10) Data dictionaries: Analyzing BRFSS data with generic stats software requires the user to pore through a data dictionary. However, that is tedious and error-prone. The solution: VB-QMS automatically produces correct results. The user never needs to refer to the data dictionary. The VB-QMS output fully documents the analysis.

11) Age-adjusted weighted percents (AAWP): AAWP are used to make comparisons between populations with different age structures. However, calculating AAWP by hand using generic stats software is tedious and error-prone, and again greatly complicates creation of home-grown software. The solution: VB-QMS automatically produces correct age-adjusted weighted percents.

12) Confidence intervals (CI): Confidence intervals are very useful. However, calculating CI by hand using generic stats software is tedious and error-prone. The solution: VB-QMS automatically and quickly produces correct confidence intervals, based on jack-knife replication.

C. Some Key Advantages of VB-QMS

Standards-Based: Before programming the more complex BRFSS data, we had years of experience analyzing and displaying other public health data, and years of experience using BRFSS data. As a first step for developing the Vitalnet BRFSS module, we carefully reviewed how the CDC and State report BRFSS results, and carefully read the extensive technical documentation on the CDC BRFSS web site. We also spoke with CDC and State staff.

Optimized database: We designed a novel record-level database architecture specially optimized for BRFSS data. This was a key step. The special database design results in low disk space requirements, rapid analyses, and retention of record-level detail. Note that VB-QMS analyzes record-level data, not some set of precalculated results.

Internal BRFSS data dictionary: Vitalnet BRFSS Query & Mapping Software has its own special internal data dictionary, in database format. It includes question wordings, responses, file layouts, and many other details. It includes all the essential information from BRFSS codebooks, and much more. It includes all the metadata needed to make VB-QMS run correctly and prevent misinterpretation.

BRFSS database engine: Vitalnet BRFSS Query & Mapping Software has a special internal "database engine" to correctly read the database and automatically produce the correct results, including weighting, based on user selections.

Data importing included: Vitalnet data importing routines were modified, to rapidly and reliably import BRFSS data into the VB-QMS data warehouse. We import any new data for you. Importing new data into an existing system normally takes about one week, mostly to modify the internal data dictionary to reflect the new data. Checking the results to ensure correctness takes a few more days.

Special BRFSS interface: VB-QMS uses the same "look and feel" as other Vitalnet modules, making it easy for users to switch between data sets. The interface is customized to handle special features of BRFSS data, such as the many questions and the survey-related outcome measures.

Best mapping capabilities: Vitalnet makes png, pdf and other map formats. Each map is attractive and publication-ready. Each map can be easily and extensively customized by the user.

Over 100 BRFSS questions: Currently, over 150 BRFSS variables and "overall questions" are incorporated into the Vitalnet BRFSS module. State-added variables can be added, as you request.

Age-adjusted weighted percents: VB-QMS easily and correctly does age-adjusted outcome measures, which historically have been under-utilized with BRFSS results.

Validated to be correct: Vitalnet BRFSS Query & Mapping Software has been stable and used for years with reliability and correct operation. The program has been extensively verified and tested, both 1) for internal consistency and 2) in comparison with many hundreds of key data points from the CDC BRFSS web site, State web sites, and CDC WEAT. We will validate your VB-QMS system to make sure that results agree with those that you produce internally and in reports.

Customizable: Vitalnet BRFSS Query & Mapping Software can be customized to add other capabilities or analyses. Let us know if anything you wish to have added.

Use VitalWeb browser platforms

D. Outcome Measures within VB-QMS

From a simple menu, select whatever outcome measure you might need to analyze BRFSS data.

Unweighted counts simply tabulate the number of responses. You can select [Unweighted # of valid interviews] or [Unweighted # 'yes' for a question] or [Unweighted # 'no' for a question].

Weighted counts adjust the number of responses to reflect the actual population size. You can select [Weighted # of valid interviews] or [Weighted # 'yes' for a question] or [Weighted # 'no' for a question].

Weighted percents compare adjusted response proportions taking into account BRFSS weighting. You can select [Weighted % 'yes' for a question] or [Weighted % 'no' for a question].

Age-adjusted weighted percents prevent misleading results if an age effect (e.g., more diabetes in older people) and a difference in age structure between two populations (e.g., UT younger than FL). VB-QMS lets the user easily and reliably produce age-adjusted output. You can select [AA weighted % 'yes' for a question] or [AA weighted % 'no' for a question].

E. Examples of VB-QMS output

The whole point of VB-QMS is to let the user easily and reliably make tables, maps, and charts. Vitalnet output produces correct BRFSS numerical results, and includes BRFSS-specific documentation. Below are links to examples of output produced by the Vitalnet BRFSS module. Each example took a minute or less to design and produce. The examples below are just a sample of what is possible with VB-QMS. If you'd like us to add something you find useful, let us know.

Tables: The fundamental Vitalnet output type is cross-tabulations. Essentially, Vitalnet lets you compare "anything with anything", make the rows and columns exactly as needed, select any of eight or so "outcome measures", analyze the subpopulation needed, and prevent invalid analyses.

Time trends: A key use of BRFSS data is to tell whether a problem is getting better or worse. Alternatively, we may wish to track increased prevalance of a healthy behavior. VB-QMS allows this to be done quickly and easily. VB-QMS does a statistical analysis to determine if a significant trend is present.

Charts: To better understand and visualize comparisons, charts are generally easier and more effective than just numbers. VB-QMS automatically makes easily customized bar charts, line charts, and pie charts. Importantly, VB-QMS prevents users from making useless or epidemiologically invalid charts.

Confidence intervals (CI): Along with the number of responses to a question, confidence intervals assess the statistical reliability of results. A narrow interval is more reliable than a wide interval. VB-QMS automatically and quickly produces correct confidence intervals.

Maps: Along with the tables and charts described above, VB-QMS has a complete mapping system. The Vitalnet mapping flexibility and output are much better than any alternative.

F. Comparison with Generic Stat Packages

Vitalnet BRFSS Query & Mapping Software provides much better performance, usability and value than the generic stat packages that have historically been used to analyze BRFSS data.

Motivation for GSP - BRFSS data have historically been analyzed using certain generic stat packages (GSP), because there was no alternative. "Generic stat packages" include SAS, SPSS, SUDAAN, and Stata.

GSP shortcomings - There will always be usefulness for GSP analysis in the hands of an expert user, for special studies not otherwise possible, and for double-checking results. However, using a GSP is far beyond the capability of the great majority of users, and is ill-suited for doing routine analyses. Even in expert hands, using a GSP is quite complex, error-prone, awkward, and time-consuming, especially for complex survey data such as BRFSS data. The user needs to understand many details about BRFSS file layouts, BRFSS variables, and how to use the complex software. Also, a GSP is typically not cheap, can easily cost thousands of dollars. Also, the needed training is expensive and time-consuming.

Vitalnet provides much better value - Vitalnet is a new and fundamentally different kind of stats package. Vitalnet is specifically and totally customized for analyzing BRFSS data (or some other data set of interest to you), for a particular jurisdiction. Vitalnet "knows" all about the data set.

Vitalnet is much easier, more reliable, and more useful. Instead of the confusing array of statistical tests offered by a GSP, including myriad not-needed options and leaving out some needed options, Vitalnet offers exactly the options the user needs, in a fully menu-driven format. Instead of the arduous, error-prone task of setting up a GSP analysis, you merely choose options from self-explanatory menus and press "Go". Vitalnet training is mostly focused on understanding data analysis in general, instead of struggling with the mind-numbing details of using a particular GSP.

Here are some ways Vitalnet is much better than a GSP for analyzing BRFSS data:

Use VitalWeb browser platforms

G. Comparison with Home-Grown Systems

Vitalnet BRFSS Query & Mapping Software provides much better performance, usability and value than current home-grown BRFSS analysis systems.

Motivation for HGS - Most realize that a general statistics package (GSP) is totally impractical for the great majority of users to analyze BRFSS and other public health data. Therefore, web-based home-grown systems (HGS), as developed by the CDC and some States, have been produced with the goal to make BRFSS and other data easier to access and analyze.

HGS shortcomings - However, in practice the HGS have mostly failed to meet the need. Developing and maintaining data warehouse software is very complex. For the most part, the home-grown systems lack needed capabilities, have awkward interfaces, and cost too much to develop and maintain.

Vitalnet provides much better value - Vitalnet provides the needed combination of analytical flexibility, ease-of-use, and output capabilities, in a professional software system. With Vitalnet, government agencies can get out of trying to be in the software development business, and get back to their public health mission. Vitalnet allows government agencies to smoothly and efficiently make best use of their data, at much lower cost.

Each HGS analyzing BRFSS data is somewhat different, with its own strengths, weaknesses, and quirks. Almost any HGS is better than using a GSP to analyze BRFSS data. However, none of them adequately meet user needs for analyzing and visualizing BRFSS data. Here are some ways Vitalnet differs from typical home-grown systems we have tested:

H. Additional Information

Cost comparison - The direct and indirect costs for a "home-grown" system add up to large amounts. Many thousands per year for software licenses and hardware costs. Many thousands per year for personnel costs (system analysts, programmers, documentation writers, managers). The actual costs will vary from State to State, but can easily be in the hundreds of thousands, or much more. And that's assuming that the "home-grown" system works well. Using generic stats software is also expensive, and with low productivity. We guarantee a better solution, at a lower cost.

Headache comparison - Do you really want to be in the software development business? Consider that 1) development takes years, 2) "home-grown" costs in succeeding years do not decrease that much in practice, 3) very difficult to hire and retain required level of programmers, 4) significant risk of project failure trying to develop complex software, and 5) "technology transfer" has logical flaws, and has not worked well in practice. We guarantee a much better solution, with no headaches / hassles.

General information - Vitalnet is a professional data warehouse solution, the product of over 20 years of development. Besides BRFSS data, Vitalnet can analyze any data set of your choosing, including birth, death, pregnancy, hospital discharge, and many more. Vitalnet runs on a PC desktop (VitalPro) and over the internet in a web browser (VitalWeb).

If you would like more information about Vitalnet BRFSS Query & Mapping Software, please fill out feedback form or call us. Vitalnet BRFSS Query & Mapping Software (VB-QMS) can be rapidly and reliably implemented for use by your State. We will customize your VB-QMS system as required to make best use of your BRFSS data. We want to help you accomplish your goals, succeed in your career, and enjoy your work. Please let us know if any questions or suggestions.

Use VitalWeb browser platforms