T0946 2019 BRFSS Data Collection Guide 110218.docx

BRFSS Data Collection Protocol with Disposition Table

Table of Contents

Introduction In 1984, the Centers for Disease Control and Prevention (CDC) initiated the state-based Behavioral Risk Factor Surveillance System (BRFSS)—a cross-sectional telephone survey that state health departments conduct monthly over landline telephones and, since 2011, cell phones; the states conduct the BRFSS survey with the use of a standardized questionnaire and the technical and methodologic assistance from CDC. BRFSS collects prevalence data among noninstitutionalized adult US residents regarding their risk behaviors and preventive health practices that can affect their health. Respondent data are forwarded to CDC to be aggregated for each state, returned with standard tabulations, and published at year's end by each state. In 2017, over 450,000 interviews were conducted in the states, the District of Columbia, and participating US territories and other geographic areas.

This document provides data collectors with a BRFSS overview and outlines the processes involved with calling, disposition-code assignment, and data submission. This document does not cover details of sampling and weighting, as they are not data-collectors’ responsibility. Specific information regarding data quality, response and/or cooperation rates, or calling outcome can be found in the yearly Summary Data Quality Report released with the annual data set.

Find yearly data and support documents here: http://www.cdc.gov/brfss/data_documentation/index.htm.

Details of the data collection process are discussed in regularly scheduled conference calls and at the annual BRFSS meetings/training workshops. BRFSS encourages data collectors to participate in these events, as updating BRFSS data-collection protocol is a collective process that is strengthened when organizations and day-to-day stakeholders provide their input.

The BRFSS Process The BRFSS questionnaire was developed in collaboration with CDC and public health departments in each of the states, the District of Columbia, and participating territories. Data derived from the questionnaire provide health departments, public health officials, and policy makers with behavioral and health status information that, when combined with mortality and morbidity statistics, guide the development of health-related policies and priorities as well as help decision makers address and assess strategies to promote good health. A finalized version of the questionnaire is sent to the states each year. CDC also provides computer-assisted telephone interviewing (CATI) programming to states, but they may opt to use their own CATI programming software using the final version of the BRFSS questionnaire as a guide. States which develop their own programming systems are required to test it accurately against the CDC provided programming. States may not change the skip patterns or wording of questions in the questionnaire but are free to create state-added questions that can be customized to states’ individual needs (see below). In some instances, states may insert state-added questions into the questionnaire--with permission from CDC--when such questions fit into the context of extant topics/sections and do not impede the flow of the interview. Requests should be submitted to the state project officer or the survey methodologist on the Survey Operations Team.

Annual questionnaire construction The BRFSS questionnaire is comprised of an annual standard core, which includes questions asked of respondents each year; a biannual rotating core, which includes questions asked only in even- or odd-numbered years; optional modules, which include standardized questions adopted verbatim by the states; and state-added questions, which states individually customized. Appendix A provides a draft copy of the 2019 BRFSS questionnaire, including modules with skip patterns. Data collectors will note that the 2019 questionnaire includes skip patterns for landline and cell phone interviews that administrators should follow when they are conducting interviews using a sample provided by CDC (see Sampling below).

Standard Core Questions: The portion of the questionnaire that is included each year and must be asked by all states. The core may include Emerging Core questions about “late-breaking” health issues. After 1 year, these questions are either discontinued, incorporated back into the standard core or become part of the rotating core or optional modules.

Rotating Core Questions: The portion of the questionnaire asked by all states on an every-other-year basis. These questions regularly appear in even- and odd-numbered years. A few questions (dealing with immunization) appear on the BRFSS in three year rotating cycles. States may use rotation core questions as optional modules in off-year questionnaires.

Optional Modules: Optional modules are sets of standardized questions on various topics that each state may select and include in its questionnaire. Once selected, a module must be used in its entirety and asked of all eligible respondents. If an optional module is modified in any way (e.g., if a question is omitted), then the questions will be treated as state-added questions (see below).

In order to achieve a wide range of data, states may choose to “split” samples in order to give different modules to each smaller group of participants. For example, if a state adopts a questionnaire that is too long to ensure respondent cooperation, the state may choose to provide a version of the questionnaire with the core and a subset of modules. In this way a greater number of modules may be used if the state uses different modules on different versions of the questionnaire. Some respondents, therefore, will answer the core questionnaire and one set of modules, while others answer the core questionnaire and a different set of modules. States are required to conduct at least 2,500 interviews for each of the versions of the questionnaire in order to have enough responses for weighting purposes. States may adopt up to three versions of the questionnaire, each including the core (with standard and rotating core questions) and a specified number of modules, which will differ by version. States must include modules on both landline and cell phone interviews. Versions must also be included in both samples.

In many instances states may prefer to insert optional modules into the core questionnaire. This may be done to improve the flow of the questionnaire by grouping questions on similar topics. The following optional modules may be inserted into the questionnaire at the discretion of the BRFSS coordinators:

Diabetes/ Pre-diabetes After the diabetes question in the chronic disease section

All other insertions of optional modules into the core sections should be approved by CDC prior to implementation of the survey.

State-added Questions: BRFSS encourages states to add their own extra questions to their questionnaire, so they can gather data on additional topics related to their specific health priorities. All questions included in the BRFSS, with the exception of state-added questions, are cognitively tested prior to inclusion in the questionnaire. It is up to each state to decide whether or not it will cognitively test its state-added questions before use. State-added questions may not be inserted into the text of the core questionnaire or optional modules without approval from BRFSS. States should contact their project officers to request insertion of state-added questions into text that has been approved for use by all states.

The wording of the questions in any part of the BRFSS, with the exception of state-added questions, is determined at the annual BRFSS meeting (in spring or early summer), where BRFSS state coordinators vote to adopt questions submitted by CDC programs. A governing group including state BRFSS coordinators, CDC staff and others known as the BRFSS Working Group, may add questions on emerging issues (such as the H1N1 flu questions added in 2009 and e-cigarette use in 2014). A field test of new questions, modules and those sections of the questionnaire affected by new questions is conducted after the state voting process. CDC then designs core components and optional modules and produces data processing layouts, while considering state priorities, potential funding, and other practical aspects. Minor changes in question wording and format may be made after the field test. The new BRFSS materials for the next surveillance year are then sent to the states, which may add their own questions that they have designed or acquired. A target of October 1 is set for finalization of the questionnaire for the upcoming year.

Data collectors should have the capacity to make modifications, including addition of questions, during the course of the year. In addition, data collectors must be capable of adjusting screening questions that determine eligibility during the course of the year.

Data Collection Data collection follows a suggested BRFSS interviewing schedule; all calls for a given survey month should be completed in the same sample month if possible. In some cases samples begun in one month may be completed in the first 7-10 days of the next month. Up to 15 calling attempts may be made for each landline phone number and up to 8 for each cell phone number in the sample, depending on state regulations for calling and outcomes of previous calling attempts. Although states have some flexibility in distribution of calling times, in general, surveys are conducted using the following calling occasions:

· Conduct 20% of landline calling attempts on weekdays (before 5:00 PM)

· Conduct 80% of landline calling attempts on weeknights (after 5:00 PM) and weekends

· Conduct cell phone calling attempts during all three calling occasions (weekday, weeknight, and weekend), with approximately 30% on weekend calling occasions.

· Change schedules to accommodate holidays and special events.

· Make weeknight calls after 5:00 PM.

· Adhere to respondents’ requests for specific callback/appointment times whenever possible.

With larger portions of the samples allocated to cell phone numbers, states may modify the calling schedule for efficiency. Data collectors must develop and maintain procedures to ensure respondents’ confidentiality, assure and document the quality of the interviewing process, and supervise and monitor the interviewers. CDC does not authorize the taping of interviews. Data collectors should keep in mind that state laws on recording conversations may vary, and there is no certainty when dialing a cell phone number as to where respondents are currently residing and accepting calls. Should a data collector record a call for any reason (such as training), CDC must be notified and approve of the procedure. In all cases where quality is being monitored by recording calls, respondents should be notified before the first questions are posed.

Each telephone number in the CDC-provided sample must be assigned a final disposition code to describe the result of calling the number:

· A completed or partially completed interview (see definitions in Appendix B) or

· A determination that:

· A household was eligible to be included but an interview was not completed or

· A telephone number was ineligible or could not have its eligibility determined.

The final disposition codes are then used to calculate response rates, cooperation rates, and refusal rates. The distribution of individual disposition codes and the rates of cooperation, refusal, and response are published annually in the Summary Data Quality Reports. BRFSS uses standards set by the American Association of Public Opinion Research (AAPOR) to determine disposition codes and response rates. All BRFSS disposition codes and rules for assigning disposition codes are provided in Appendix B: Disposition Table with Callback Rules. Given the myriad outcomes for assigning specific codes associated with technological barriers, additional guidance is provided in Appendix C: Understanding Coding for Technological/ Telecommunication Barriers. Data collectors must follow the rules for assigning disposition codes and train and monitor interviewers in the use of specific dispositions.

Survey Protocol BRFSS sets standard protocols for data collection, in order to maintain consistency across states that permits state-to-state data comparison. Data collectors should follow the assignment of disposition codes provided in Appendix B: Disposition Table with Callback Rules. Disposition codes follow the format of 1000-1999 completed/partially completed; 2000-2999 non-completed interviews with eligible respondents/households; 3000-3999 non-completed interviews with unknown eligible persons/households; 4000-4999 ineligible numbers; 5000-5999 interim dispositions. A 2000 level disposition should not be assigned unless the interviewer is certain that both the household and respondent are eligible for the survey. Assigning incorrect disposition codes can lower response rates and efficient use of the sample. The following items are included in the BRFSS survey protocol:

1. All states must include the core questions and introductory scripts without modification. States may choose to add any, all, or none of the optional modules and state-added questions after the core component. Interviewers may not offer information to respondents on the meaning of questions, words, or phrases beyond the interviewer instructions provided by CDC and/or the state BRFSS coordinators. States may not insert state-added questions into the core component or into optional modules without permission. State coordinators should contact their CDC project officers to request the placement of state-added questions into text that has been approved for use by all states.

2. Systematic, unobtrusive electronic monitoring is a routine and integral part of monthly survey procedures for all interviewers. States may also use callback verification procedures to ensure data quality. Unless supervisory monitoring of 10% of all interviews is being routinely conducted, a 5% random sample of each month’s interviews must be called back to verify selected responses for quality assurance. Recording calls as part of quality assurance is not part of the BRFSS methodology and recording interviews without respondent knowledge is not legal in all states. Data collectors should remember that cell phone numbers may reach respondents in any state or country, where laws on recording calls may be different than in the state where the call originated.

3. An eligible household is defined as a housing unit that has a separate entrance, where occupants eat separately from other persons on the property, and that is occupied by its members as their principal or secondary place of residence. The following are non-eligible households: vacation homes not occupied by household members for more than 30 consecutive days per year, group homes, institutions, and (in the landline telephone sample) households in states other than the one conducting the BRFSS questionnaire. Persons in a state’s cell phone sample who are residents of other states are eligible for interview. The state contacting the respondent should complete the core questionnaire and then provide the data to CDC for transfer to the appropriate state of the respondents’ residence. States should especially attempt to obtain the state of residence of respondents who indicate that they have moved and retained their cell phone number from another state. States should collect verbatim county information on persons who live in other states in order to permit the correct weight for the respondent after data are transferred. Since 2012, persons living in college housing have been included as eligible respondents. Although it is rare to contact a college housing resident in the landline sample, this person would also be included as a single adult household. The BRFSS is a self-reported survey. If respondents report that they live in private residences, it is not the role of interviewers to question them. The only instances under which there is discussion of information on whether households qualify as private residences is when respondents initiate the question.

4. Eligible household members include all related adults (aged 18 years or older), unrelated adults, boarders/roomers, live-in au pairs or students and domestic workers who consider the household their home, even though they may not be home at the time of the call. College housing residents are treated as single adult households. Household members do not include adult family members (including students) who are currently living elsewhere.

5. Questions should be read verbatim. In many cases introductory phrases are provided which should also be read as written. Interviewer instructions are optional and can be read if the respondent is confused or needs additional information. Items in parentheses in statements are also optional and may be read for clarification. Interviewers should not offer their own interpretation of questions or response options.

6. Proxy interviews are not conducted in the BRFSS. For people interviewed on landline telephones, individual respondents are randomly selected from all adults living in a household and are interviewed in accordance with BRFSS protocol. Household members include all family members, domestic servants, and au pair or live-in students who have resided at the residence for at least 3 months. Cell phone interviews are conducted with respondents who answer the number called and are treated as one-person households.

7. An interview is considered complete if data are collected for all questions which would have normally been asked for any selected respondent. Partially completed interviews are defined as those where the first sections of the interview are completed and the portions of the demographic section which are used for weighting are also asked of the selected respondent in regular order of the questionnaire. For the 2019 questionnaire this will include through question 8.12 for landline respondents and 8.09 for cell phone respondents. If the respondent does not provide substantive responses for weighting variables (that is, the respondent refuses to answer or responds that he/she does not know), imputed values will be generated and used only to assign weights. If an interviewer codes a number of responses as “don’t know” or “refused” just prior to cut off in order to have an interview count as a partial complete, this will be noted by the CDC staff as potential falsification of data.

States should monitor data collection to ensure that the percentage of partial completes is not greater than 10% of the number of completes. A large percentage of partial completes could result in missing data for variables that follow the demographic sections including all optional modules. The number of partial completes which are not part of the RFP may be higher, if the state has contracted with the data collector ONLY on the number of 1100 dispositions.

8. With the exception of verbally abusive respondents, eligible people who initially refuse to be interviewed may be contacted at least one additional time and given the opportunity to be interviewed. Preferably, this second contact will be made by a supervisor or a different interviewer. Some states have regulations on whether refusals should be called again and the manner of the refusal conversion. For example, a period of two days between the initial refusal and second attempt is often standard protocol. Data collectors should contact the state BRFSS coordinator to determine the state’s policy on calling back refusals. Some states maintain an internal list of numbers which connect to households/persons who have been particularly adamant about being taken off calling lists. Numbers from new samples are matched against these list to prevent calling the numbers. States generally maintain a number on the list for up to two years. Numbers de-duped as a result of such lists should be coded with dispositions of 3700.

9. States are required to give a final disposition for every number in the sample, usually within the same month of the sample. States should complete all calling on each monthly sample within that month. A few states receive and account for all calling on a sample on a quarterly basis rather than a monthly basis. Data collectors should contact the state BRFSS coordinator to verify whether the state is receiving a monthly or quarterly sample from CDC.

10. The BRFSS OMB number and burden statement must appear on the header page of all interviewer forms. The CDC will provide the header with the questionnaire each year. Please note that the interviewers do not need to read any part of the OMB number or burden statement to the respondents unless asked. The entire burden statement does not need to be read if the respondent is simply asking how long the interview will take. If the respondent asks for any information at any time about the authority by which information is being collected, it is imperative that the OMB approval information be available to the interviewer. The interviewer may then cite the OMB control information, which would allow the respondent to review the project plan online.

11. CDC provides the states with a Spanish translation of the BRFSS questionnaire. Unlike the English version, states may change wording of the Spanish version in order to match local dialects.

12. Each data collector must assign a unique identifier to each interviewer, so that state health departments and the CDC may conduct interviewer-level analyses. This identifier should not be assigned to other interviewers working on the BRFSS for that state.

General callback and disposition coding rules are established by CDC (see Appendix B), and states are encouraged to adhere to them whenever possible. The calling rules are not universally applicable to each state. Data collectors contracted by the states should have the capacity to adhere to the calling rules listed below as well as those to in Appendix B.

1. All cell phone numbers must be hand dialed. Data collectors should seek legal advice if they are uncertain whether their practices are in any way contradictory to the regulations.

2. Interviewers should be trained specifically for the BRFSS and retrained each year.

3. If possible, calls made to non-English-speaking households and assigned the interim disposition code of 5330 (household language barrier) should be attempted again with an interviewer who is fluent in the household language (e.g., Spanish).

4. States should maximize calling attempts as outlined in Appendix B. The maximum number of attempts (15 for landline telephone and 8 for cell phone) may be exceeded if formal appointments are made with potential respondents. There are many instances in which the maximum number of callbacks is not required. States and data collectors should refer to the callback table provided in Appendix B with this document for the required number of callbacks for each calling outcome/ disposition.

5. Calling attempts should allow for a minimum of 6 rings and up to 10 rings if not answered or diverted to answering devices.

6. The maximum number of attempts may be set by the states. CDC recommendations for the maximum number of attempts are 15 for landlines and 8 for cell phones.

7. All numbers must be assigned a final disposition. Data should not be submitted with interim dispositions.

8. Messages left on answering devices/voice mail devices should be left by interviewers. Messages should never be left by any automated voice devices. States may have their own standard scripts for messages, describing the reasons for the call and when respondents might expect a return call. Messages can be left after any attempt. It is not recommended that respondents be burdened by repeated messages. States should adopt protocols to leave one or two messages during the calling attempts for a single number during the calling period.

9. If a respondent indicates that they will be available for a callback at a specific time/date, appointments may be made. Data collectors may send texts reminding cell phone respondents of the appointments after receiving verbal permission to text that information.

10. States adopting the Adverse Childhood Experience (ACE) Module, should train interviewers on how to handle respondents who become upset or stressed as a result of the questions. Appendix G provides a crisis protocol to train interviewers who will be administering this module.

Using the BRFSS Sample In some instances, states design samples within boundaries of sub-state geographic regions. States may determine that they would like to sample by county, public health district, or other sub-state geography in order to make comparisons of geographic areas with their states. To conduct the BRFSS, states get samples of telephone numbers from CDC. States then review their sampling methodology with a state statistician and CDC to make sure data collection procedures are in place to follow the methodology. States must consult with CDC before making changes to methodology. States must maintain sample phone numbers in files that are separate from responses, in order to maintain standards of respondent confidentiality.

The BRFSS uses two samples: one for landline telephone respondents and one for cell phone respondents. State BRFSS coordinators work with CDC to produce all samples The CDC recommends that the range of completed cell phone interviews be a minimum of 50% to a maximum of 80% of total completed interviews. The 50-80% range of cell phone interviews is set to ensure the geographic distribution of the sample (since landlines samples can be geographically distributed across the state) and to ensure that the sample is demographically representative of the state. Data collectors are ultimately responsible to states for the distribution between landline and cell phones. Since landline telephones are often shared, household sampling is used in the landline telephone sample. Household sampling requires interviewers to collect information on the number of adults living in a residence and then select randomly from all eligible adults (see questionnaire). Cell phone respondents are treated as single adult households and therefore do not require household sampling. The samples are fully overlapping, so that any eligible person in the landline frame may also be eligible in the cell phone frame. States receive the sample monthly or quarterly, approximately by the 15th. Note that the BRFSS is a sample with replacement. It is possible, therefore, for a single household/respondent to be eligible and appear in a sample more than once within a year. Some states eliminate duplicate (“de-dup”) numbers that appear within the same quarter. A state with sub-state regions that represent small areas is more likely to encounter repeat numbers in the sample. States that wish to send advance letters should request addresses with their regular landline and cell phone samples. For states that send advance letters, mailing addresses are appended to telephone numbers. Data collections should release all replicates (of 30 numbers) in the sample in the first week of each month. Those who receive samples quarterly should release them in a manner that allows for sufficient calling prior to the end of the quarter. The table below provides the format for the landline and cell phone sample files received by the states.

Field Name

Size

Position

Format/Values/Explanation

Phone Number (AREACODS, PREFIXS, SUFFIXS)

1-22

9,1-NNN-NNN-NNNNv20191

Geographic Stratum (_GEOSTRS)

23-25

First position = 1 for Landline / 2= Cell phone Then States with no geographic strata=01 in each record. Others according to provided information.

Density Stratum (_DENST2S)

1=Listed number, 2=Not listed one-plus block, 3=Zero block, 9=Not applicable (GU, PR, VI).

Sequence Number (SEQNO)

27-36

A unique 10-digit number for a state for a year with year in the first four digits. For example: 2015000001.

Number of Records Selected From Stratum (NRECSELS)

37-42

Number of telephone numbers (eligible sampling units) selected from stratum.

Number of Records in Stratum (NRECSTRS)

43-51

Number of telephone numbers in the stratum from which sample was selected.

Precall [GENESYS-ID] Status (PRECALLS) Landline sample

1=To be called, 3=Non-working number, 5=Business phone. (Including GU, PR, VI) 4 = Cellular – PRO-T-S, 6 = Cellular - Interviewer

Precall [Cell-WINS Screening] Status (PRECALLS) Cell phone sample

1=Active, 3=Inactive, 7=Unknown Status

Replicate Number (SMONTH, REPNUM)

53-58

The first two digits, 01-12, represent months, the last four digits a sequential number starting with 0001 each month.

Replicate Depth (REPDEPTH)

59-60

A sequential number from 0130 in each replicate.

State FIPS Code (_STATE)

61-62

FIPS code of assigned state.

County FIPS Code (ASGCNTY)

63-65

FIPS code of assigned county.

Blank=GU, PR, VI.

County FIPS Code of Listed Number (LISTCNTY)

66-68

For listed numbers, FIPS code of the county in which number is located. For not listed numbers=999. Blank=GU, PR, VI, and cell phone sample

Number of Landline Assignments in 1K Blocks in Assigned County (NOHHCTY)

69-72

Number of Landline Assignments in 1K Blocks that are in assigned county.

Blank=GU, PR, VI, and cell phone sample

NXX Type (NXXTYPE)

73-74

Blank

Number of Landline Assignments in 100 series bank (BLCKSIZE)

75-77

Number of landline assignments in hundred block=000-100. Blank=GU, PR, VI, and cell phone sample

Number of Landline Assignments in 1K Block (LSTHHPRE)

78-82

Number of Landline Assignments in 1K Block Blank=GU, PR, VI, and cell phone sample

Estimated Total Households in Prefix (TOTHHPRE)

83-87

BLANK

Core Based Statistical Area

(CBSACODE)

88-92

99999=Not in an MSA.

Blank=GU, PR, VI, and cell phone sample

Metropolitan Status Code (MSCODE)

1=In the center city of an MSA, 2=Outside the center city of an MSA but inside the county containing the center city, 3=Inside a suburban county of the MSA, 4=In an MSA that has no center city, 5=Not in an MSA. Blank=GU, PR, VI.

Rate Center Name (RCNAME) Cell phone only

94-123

Rate Center Name (RCNAME)

V&H Coordinate (VNHCOORD)

124-133

BLANK

Date Sample Generated (DATESMP)

134-143 mm/dd/yyyy

Pre-screening Process Used

(PRESCREN)

0= Not screened 1=ID 2=ID Plus 3 = CSS, 4 = Cell

Date Sample Pre-screened

(DATESCRN)

145-154

Release Date of Active Prefix Database (PHNRLDAT)

155-164

Blank= cell phone sample

Release Date of Listed Phone Number Database (LSTRLDAT)

165-174

Blank=GU, PR, VI, and cell phone sample.

CLEC Number (CLEC)

1=Yes, 2=No. Blank=GU, PR, VI, and cell phone sample.

Replicate designated for inclusion in Multi-Mode Mail Survey and address match status (MSREPMCH) Landline only

Blank = All States

Time Zone

Eastern = 7, Central = 6, Mountain = 5, Pacific = 4, Alaska = 3, Hawaii =2

Listed in one of the following Databases: InfoUSA, Experian, (DIRLST)

1=Yes, 2=No.

Blank=GU, PR, VI, and cell phone sample.

Secondary Screening Flags

0/Undetermined/Residential

1/ No Answer

2/ Busy

3/ Fax/Modem

4/ Language Barrier

5/ Privacy Manager

6/ Residential Voice Mail

7/ Residence/Phone Answered 8/Cellular Number 9/Business/Non-Working

Indication of Address Matching Landline only

2=Not matched Blank=GU, PR, VI, and cell phone sample

Path variable (PATH) used to help identify which questionnaire is used when there are dual questionnaires.

(States may update this variable to use for multiple paths in split samples)

20 = Default Genesys value Cell Phone Survey Sample

Note: Monthly files will be sorted by stratum, replicate, and depth. The order of numbers within a replicate will be randomized before assignment of depth numbers. All numeric fields are right aligned and padded with leading zeros. All character fields are left aligned with trailing blanks.

Each phone number is assigned a precall status to indicate whether the number should be called. States may opt to call landline telephone numbers with precall status >1 but are not required to do so. States may also choose not to call landline numbers from the unlisted portion of the sample with precall status = 1 which also have secondary screening status as fax/modem lines or are listed as “busy” by the precall screener. States are not required to call cell phone sample numbers with an “inactive” precall status, but may choose to do so. States should call all cell phone numbers with active and unknown precall status in the cell phone samples. Given that the precall status indicates the potential for reaching an eligible respondent, calling landline numbers with precall >1 or cell phone numbers with inactive precall status may reduce response rates. States may also use the secondary precall status to assign bilingual interviewers to numbers with language barrier precall assignments, or make extra efforts to reach numbers which have precall status indicating residence/household status. In 2019, landline phone numbers will be differentiated in the sample as precall > 1 if a technical review indicates that the number is never answered or a business. This should reduce the calling efforts for landline no answers, which typically represent more than half of the final dispositions for landline sample.

The landline sample is taken from listed and unlisted numbers at a ratio of 1:2. The ratio has changed from previous years, when it was 1:1.5, due to the changes in stability of the landline samples. NOTE: The number of callbacks required on the landline sample is different based on the density strata. Numbers from the low density strata have eight required callbacks, while those on the high density strata should be called 15 times. Please see the callback rules table provided in Appendix B for information on the number of required callbacks for each calling outcome/disposition.

States that request addresses may send advance letters to those households to alert them to the fact that they will be receiving calls and the nature of the survey. States may include a toll free number for potential respondents to inquire about the BRFSS. Studies have shown that the use of advance letters does improve response rates. However, the proportion of the landline sample that is accurately matched to addresses is declining. Currently about 20-30% of the landline sample is accurately matched to an address. In 2019, for the first time, the BRFSS will also support sending advance letters to cell phone respondents. The ability to match cell phone numbers to addresses is improving, but states should only expect about 20-23% of cell phone numbers to match correctly to addresses. Since the cell phone sample is at the individual rather than household level, letters should specify which phone number connected to that address has been chosen. In order to maintain confidentiality, it is required that only portions of the phone number (such as the last four digits) be noted in the advance letters for cell phone respondents. The CDC will make every effort to provide addresses for advance letter as early as possible during each month. Data collectors should note that the cell phone sample is appended with landline numbers which have been ported to cell phone in previous months. Therefore the landline sample files may arrive earlier in the month than cell phone samples. Data collectors should speak to their BRFSS state coordinators about advance letters.

Samples for US territories differ from those from the states. BRFSS coordinators in US territories may deviate from the calling and sampling guidelines to fit the data needs of their jurisdictions. Data collectors should work closely with state BRFSS coordinators to ensure that the sample is properly managed. CDC will provide quarterly sample productivity tables on the upload/download site to alert the state coordinators of any problems with sample management.

Data collectors can track samples and productivity using the YTD Data Quality Reports (DQR) available with assigned logins on the upload site. Appendix F provides information on how to interpret information provided in the YTD Reports. The following table of contents lists the information available in the YTD Data Quality Reports. Changes in the information provided in the DQRs may change according to the needs of the data collectors and state coordinators. A short tutorial for using and understanding the YTD Data Quality Reports is found in Appendix F. BRFSS Coordinators and Data Collectors may attend training on using the YTD reports and Uploading Data at the annual BRFSS Questionnaire Meeting.

Year-to-Date Data Quality Reports Table of Contents

Definition of Variables

Final Disposition Codes

Table 1A. Interview Month By File Month (Landline only)

Table 1B. Interview Month By File Month (Cell phone only)

Table 2A. Discrepancy in Sex Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Landline only)

Table 2B. Discrepancy in Sex Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Cell phone only)

Table 3A. Discrepancy in Age Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Landline only)

Table 3B. Discrepancy in Age Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Cell phone only)

Table 4A. Discrepancy in Race/Ethnicity Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Landline only)

Table 4B. Discrepancy in Race/Ethnicity Between Population Estimates and Unweighted BRFSS Data, Year-to-Date (Cell phone only)

Table 5A. Geo-Stratum by File Month and Year-to-Date (Landline only)

Table 5B. Geo-Stratum by File Month and Year-to-Date (Cell Phone only)

Table 6A. Date, Day of Week, Number of Interviewers and Final Disposition Code, by File Month (records with one or more attempts; Landline only)

Table 6B. Date, Day of Week, Number of Interviewers and Final Disposition Code, by File Month (records with one or more attempts; Cell phone only)

Table 7A. Number and Percent of Completes in the First 5, 10, 15, 20, and 25 Days by File Month (Landline only)

Table 7B. Number and Percent of Completes in the First 5, 10, 15, 20, and 25 Days by File Month (Cell phone only)

Table 8A. Eligibility and Status by Categories of Disposition Code, Year-to-Date (Landline only)

Table 8B. Eligibility and Status by Categories of Disposition Code, Year-to-Date (Cell phone only)

Table 9A. Eligibility and Status by Categories of Disposition Code by File Month (Landline only)

Table 9B. Eligibility and Status by Categories of Disposition Code by File Month (Cell phone only)

Table 10A. Detailed Disposition Code, Year-to-Date (Landline only)

Table 10B. Detailed Disposition Code, Year-to-Date (Cell phone only)

Table 11A. Detailed Disposition Code by File Month (Landline only)

Table 11B. Detailed Disposition Code by File Month (Cell phone only)

Table 12A. Number of Attempts by File Month (Landline only)

Table 12B. Number of Attempts by File Month (Cell phone only)

Table 13A. Outcome Rates, Year-to-Date by File Month (Landline only)

Table 13B. Outcome Rates, Year-to-Date by File Month (Cell phone only)

Table 14A. Income (77 and 99 collapsed), Year-to-Date (Completes only; Landline only)

Table 14B. Income (77 and 99 collapsed), Year-to-Date (Completes only; Cell phone only)

Summary Data Quality Reports are also available on the BRFSS website for previous years. States may compare their data productivity to that of other states in the summary reports, but will not have access to the YTD reports from other states.

Data Submission CDC will provide a data layout file for monthly data submission. The BRFSS provides a data submission website to be used for uploading states’ data and monitoring the progress of processing. Access to this site is limited and requires a login accepted by CDC. Details on data submission are included in Appendix D: Uploading BRFSS Data Using OneEdits. Note that 2016 was the first year that OneEdits software will be used for data submission, so procedures have changed recently. Data collectors should download and run edit fix programs from the upload site prior to submitting data. Errors in submitted data will delay processing and may result in data sets being returned to states for corrections. Monthly data submission is preferable (and required during flu season, September-June). Data for each should be submitted by the 25th of April, July, October and January, respectively. Following the quarterly submissions, states will receive a data file for checking. Appendix E outlines the steps in submission of data files that have been cleaned through OneEdits.

Data Sharing, Security and Rights to the Datasets

State and territorial health departments are the owners of the datasets. Data Collectors which are contracted by health departments have NO data rights and should not share or publish from state data sets without written permission. States should consider carefully before permitting use of data which are not part of the public use dataset. Persons seeking access to county code information, industry and occupation information or other variables which are not part of the public use dataset may apply for access through the CDC Research Data Center (RDC) at https://www.cdc.gov/rdc/index.htm.

While the Privacy Act is not applicable, the appropriate security controls and Rules of Behavior should be incorporated to protect the confidentiality of information, proprietary, sensitive, and Personally Identifiable Information (PII) the data collector may come in contact with during the BRFSS data collection process.

As is noted elsewhere in this document, sample files and response files should not be merged or linked. Data collections should transmit data only through the upload sites provided by CDC and not by email attachments or other means. Data collectors should maintain the most recent virus protections, operating systems patch levels and other security measures to all computers used to collect BRFSS data. While computers used for CATI data collection are not in all cases accessible to the internet, data collectors should ensure that access to data using flash drives or other devices is restricted to authorized individuals for authorized purposes.

Sample files should not be retained past the data of final use for BRFSS purposes. BRFSS sample file use is restricted to completion of BRFSS data collection and data collection for BRFSS call-back surveys. BRFSS sample files are not to be used for any other or subsequent purpose. Data collectors may not obtain permission for such use from respondents and are restricted from soliciting BRFSS respondent to take part in other surveys, internet panels or any other related use.

Data collectors should not retain sample or data files beyond the time that is necessary to finalize data. During the time that datasets are retained by data collectors, they should meet all security requirements for data storage and firewalls that are included in the most current BRFSS OMB approval. Data collectors may choose to configure computers that contain BRFSS data with the applicable United States Government Configuration Baseline (USGCB) (see http://usgcb.nist.gov/). Note: USGCB is applicable to all computing systems, including desktops and laptops—regardless of function—but not including servers.

State Pilot Projects Using Protocol Adjustments

At any time during the data collection process, states may make greater efforts to reach respondents than the protocols listed here. These efforts may include increasing the number of attempts, increasing the ring times, calling all numbers in the sample regardless of the precall status or increased interviewer monitoring or training. On occasion states may wish to make adjustments to the data collection protocol in order to test the efficiency of a new procedure. For example, in 2017, one state determined that response rates might improve if the interviewer took a more conversational tone during the introduction of the survey. The protocol adjustment was approved and the change was made and tested by the state. States, and data collectors who wish to make protocol adjustments must have written approval from CDC in order to make adjustments. State coordinators should contact their project officers and the survey operations team at the Public Health Surveillance Branch of the Division of Population

Appendix A 2019 BRFSS Questionnaire (to be inserted)

Friday, November 2, 2018

Appendix B: Disposition Table with Callback Rules

Definitions of terms

weekday (before 5:00 pm on a weekday);

weeknight (after 5:00 pm on a weekday), and;

weekend (any time on Saturday or Sunday).

Disposition Code

Description

1-8 attempts (cell phone)

1-8 attempts (cell phone) Selected respondent may be called back to fully complete the interview. Give final disposition on 15th or subsequent call attempt even if there is only one occurrence of a refusal or termination.

(landline telephone only)