Skip all navigation Skip to page navigation

DHHS Home | A-Z Site Map | Divisions | About Us | Contacts

NC Department of Health and Human Services
N.C. State Center for Health Statistics Home
N.C. Public Health Home
 
 

Technical Notes 2010

Behavioral Risk Factor Surveillance System (BRFSS)

Detailed data tables for NC BRFSS survey items are posted by the SCHS on an annual basis. The exact wording for each question is used as a title for the majority of the web tables, however due to space limitations the titles may not include some introductory remarks and explanations for some questions. Although not all survey questions are presented in the web tables, the full questionnaire may be viewed online.

For the majority of survey items, the results are displayed by sex, race, age, education, and household income. For the statewide and regional tables, disability status, veteran status and Hispanic origin are also displayed as demographic groups.

Results are also posted separately by gender, by race (white and African American), and by selected risk factors (defined below).  In addition, the results are posted for 23 of the state's largest counties. For 2010, if the total number of respondents was less than 100 for a particular survey item, the results were limited to the total sample and further subgroups are not presented. This occurred most often for county-level tables regarding topics, such as diabetes, that pertained to a small sample of respondents. As was the case in 2009, the results for gender by race, e.g., white females, have not been posted for 2010.  These results may be obtained by request by contacting the BRFSS staff.

Interpreting Results
Following guidance from the CDC, the reader should give consideration to both the sample size and the width of the confidence intervals when interpreting results. Readers should use caution in interpreting cell sizes less than 50. The cell size is either the number of respondents found in the denominator, i.e., the table column labeled, "Total Respond.," or the numerator. The cell size for the numerator is defined by the number of respondents associated with the response categories (e.g., "Yes-No").

When interpreting the results, the reader should also give consideration to the width of the confidence intervals. The CDC suppresses the results if the half-width of the confidence interval is greater than 10 percent. Although the SCHS does not suppress these results, the reader should be aware when this occurs and should view these results as NOT reliable. For example, for diabetes screening, 89 of the 165 Spanish-speaking respondents reported having a diabetes screening test in the past three years. Here, the number in the denominator is 165 and the number in the numerator is 89, yielding a weighted percentage of 48.8. Although the cell sizes are greater than 50, note that the confidence interval ranges from 37.5 to 60.3; the half-width for this confidence interval is 11.14. These results, therefore, have a large degree of statistical error and are considered unreliable.

Although it is the policy of the SCHS to post results for all survey questions, including results with fewer than 50 respondents, it is up to the reader to exercise caution when small numbers appear in the results.

Some topics are age-dependent topics, such as prostate cancer screening, which was asked only of men 40 years or older. Please note age ranges for particular survey items in the web table’s footnotes or as cited in the table title.

Weighted BRFSS data are used in all calculations; therefore percentages shown in web tables cannot be derived exactly from the numbers presented. BRFSS data are weighted for the probability of selection of a telephone number, the number of adults in a household, and the number of phones in a household and adjusted to reflect the demographic distribution of North Carolina's adult population (ages 18 and older).

Respondents who refused to answer or did not know the answer were excluded from most calculations in these tables. However, there are instances when "Don't know/Not sure" responses may provide valuable information and are included in the response categories.

For more technical information about BRFSS, please visit national BRFSS web site.

Risk Factors and Calculated Variables Presented in the 2010 BRFSS Results

Body Mass Index Grouping

Body mass index (BMI) is computed as weight in kilograms divided by height in meters squared:(kg/ m2). BMI is an intermediate variable used in calculating these measures:
Underweight: BMI less than 18.5, Recommended Range: BMI 18.5 to 24.9, Overweight: BMI 25.0 to 29.9, Obese: BMI greater than 29.9. and Overweight or Obese: BMI greater than 24.9.

Smoking Status

Current Smoker (every day): Respondents who have smoked at least 100 cigarettes in their lifetime and now smoke every day.
Current Smoker (some days): Respondents who have smoked at least 100 cigarettes in their lifetime and now smoke some days.
Former Smoker: Respondents who have smoked at least 100 cigarettes in their lifetime and currently do not smoke.
Never Smoked: Respondents who have not smoked at least 100 cigarettes in their lifetime.

Current Smoker

Yes: Current Smoker (every day or some days)
No: Former Smoker or Never Smoked.

Binge Drinking

No : Respondents who report they did not drink in the past 30 days, or who drank in the past 30 days but did not have five or more drinks for males or four or more drinks for females on an occasion.
Yes : Respondents who report they did drink in the past 30 days and had five or more drinks for males or four or more drinks for females on one or more occasions in the past month.

Heavy Drinking

Yes: Respondents reported having MORE than 2 drinks/day for MALES and MORE than 1 drink/day for FEMALES.
No: Respondents reported having LESS than or equal to 2 drinks/day for MALES and LESS than or equal to 1 drink/day for FEMALES.

Have had a mammogram within the past 2 years (Age 40+)

Yes: Females age 40 and older who reported ever having a mammogram and reported having their last mammogram within the past 2 years (less than 2 years ago).
No: Females age 40 and older who reported ever having a mammogram and reported having their last mammogram 2 or more years ago.

Have had a mammogram within the past 2 years and a clinical breast exam within a year (Age 40+)

Yes: Females age 40 and older who reported having their last mammogram within the past 2 years (less than 2 years ago) and their last clinical breast exam with the past year (less than 12 months ago).
No: Females age 40 and older who reported never having a mammogram or never having a clinical breast exam, or reported having their last mammogram 2 or more years ago or their last clinical breast exam more than 1 year ago.

Women age 18+ who have had a Pap Smear in the last 3 years

Yes: Females age 18 and older who reported ever having a Pap test and reported having their last Pap test within the past 3 years (less than 3 years ago).
No: Females age 18 and older who reported never having a Pap test, or reported having their last Pap test 3 or more years ago.
No: Women who reported having a hysterectomy were not included in these calculations.

Males aged 40+ that have had a PSA test in the past 2 years

Yes: Males age 40 and older who reported having a PSA (prostate-specific antigen) test within the past two years.
No: Males age 40 and older that had a PSA test greater than two years ago, or never had the test.

Disability Status

Yes: Respondents who reported being limited due to physical, mental or emotional problems or using special equipment or having learning problems or considering himself or herself as having disability.
No: Respondents who reported NOT being limited due to physical, mental or emotional problems or using special equipment or having learning problems or considering himself or herself as having disability.

Strengths and Limitations of the BRFSS Survey Data

One limitation of a telephone survey is the lack of coverage of persons who live in households without a telephone. Households without a telephone are, on average, of lower income. Therefore, for many of the health risks measured, the results are likely to understate the true level of risk in the total population of adults in North Carolina. A second limitation is due to the fact that the data are self-reported by the respondents. We expect that respondents tend to underreport health risk behaviors, especially those that are illegal or socially unacceptable. A third limitation is due to the fact that cell-phone-only households are not included in the estimates for 2010 – see below for discussion of potential bias inherent in omitting such households from telephone surveys.

There are some significant advantages of the telephone survey methodology, including better quality control over data collection made possible by a computer-assisted-telephone-interviewing system, relatively low cost, and speed of data collection. The BRFSS methodology has been used and evaluated by the CDC and participating states since 1984. The content of the survey questions, questionnaire design, data collection procedures, interviewing techniques, and editing procedures have been carefully developed to improve data quality and lessen the potential for bias. The data collection is ongoing, and each year new annual results become available.

Surveying Cell-Phone-Only Households

The number of wireless or cell-phone-only households in the U.S. is growing rapidly. The latest estimates from the July–December 2010 National Health Interview Survey (NHIS) indicate that about three in 10 American households had only wireless telephones; for the same time period in 2009, the estimate was one in four households.  Results from the NHIS tell us that wireless-only household members are more likely to be male, Hispanic, under age 34, renting their homes, and living in poverty, or near poverty.  From a health behavioral perspective, studies have also shown that adult members of wireless households are more likely to be binge drinkers, more likely to smoke cigarettes, and more likely to report cost as a factor prohibiting access to medical care.

In 2009, the CDC (Centers for Disease Control and Prevention) produced a special report for each state, entitled, the “2009 Weighting Comparison Report.”  The report highlighted two significant changes in the weighting process that will become standard procedures with the release of the 2011 state data. First, the CDC is moving from the traditional post-stratification method of weighting the data, where categories of gender, age, race/ethnicity, and regions within states are used to adjust the final results, to a method known as “raking,” or iterative proportional fitting.  This new raking methodology permits the inclusion of education level, marital status, and renter/owner status, along with gender, age, and race/ethnicity, in the final weights. These new raked weights will improve the representativeness of the sample population to the state population, particularly for low education groups. Secondly, the CDC will include cell-phone-only households in the final weights for 2011.

In the 2009 Weighting Comparison Report for North Carolina we find that for some indicators there is a significant difference between the post-stratified (traditional) weight and the new raked weight, which includes both landline and cell-phone-only respondents.  For example, based on the post-stratified weight, 17.10 percent of respondents (landline only) reported not being a able to see a doctor because of cost in the past twelve months. Whereas, when the 363 cell phone respondents (interviewed between March and October of 2009) were included in the new raked weight, the final estimate regarding the effect of doctor-cost was 21.01 percent, resulting in a difference of 3.91 percentage points. This difference is instructive of the degree of bias associated with both the old weighting methodology and the failure to include cell-phone-only respondents in the survey sample. However, for some indicators, such as cholesterol screening, there’s virtually no difference between the post-stratified and raked weights. For example, based on the post-stratified weight, 74.44 percent of respondents reported having their blood cholesterol checked in the past 12 months; the corresponding percentage was 74.46 percent using the raked weight with cell phone respondents included.

The takeaway message is this: the 2010 results, shown in the adjoining tables, are subject to some degree of bias based on the fact that the post-stratification weights do not adjust for cell-phone respondents, nor do they adjust for the number of North Carolina adults with low education, or non-married, or renter/owner status.

Return to 2010 BRFSS Annual Results Table of Contents