Appendix B. National and State Estimates

Estimates of Paid Community Health Workers

The Standard Occupational Classification (SOC) System used by occupational data collection entities, including the Census Bureau and the Bureau of Labor Statistics, does not contain a code that clearly identifies the occupation of community health worker (CHW). Until now, CHWs have been counted in official reports under existing occupation codes with job descriptions that are similar, but not equivalent to, the activities performed by CHWs. The method used in this study to estimate the number of paid CHWs is described in the following section.

Despite best efforts, the estimates of the number of paid CHWs are tentative since assumptions had to be made about which occupational codes had been used by individuals and human resource managers to report CHW activities. The assumptions employed were reviewed by scientists knowledgeable about the methods and designs of the Census and BLS surveys, as well as by researchers and specialists who had experience in studying and working with CHWs. All technical advisors expressed the opinion that the estimates, while not ideal, were useful indicators and the best effort possible within the budget constraints of the study.

Data Sources

Two sources of data were available to provide information about employed CHWs: (1) The 2000 Census data collected by the Census Bureau and released to the public in de-identified format as the Public Use Microdata Sample (PUMS) and (2) The Staffing Patterns data, used to calculate occupational estimates, collected by the Bureau of Labor Statistics (BLS). Both of these data sets contain information about workers by occupational title code (from the SOC System) and industry codes (from the North American Industry Classification System - NAICS). The codes used for the estimates were selected through a multistage process that began with a review of the available literature on CHWs.

Even though the coding system is the same, the Bureau of Labor Statistics collects more detailed occupation and industry data than the Census Bureau. This results in some discrepancy in the information available from each of the sources. For example, individuals responding to the Census "long form" may incorrectly identify either the industry in which they are employed or the occupational title used by their employer. Also, the PUMS industry and occupational categories are broader than the BLS categories and may overestimate the number of workers. Table B.1 shows a comparison of the two data sources.

Table B.1 Brief Comparison of Data Used for Computing Estimates of Paid CHWs

* PUMS data from the American Community Survey will be available on a yearly basis, as it will replace the "long form" of the decennial Census.
** Each employer is surveyed every 3 years.

Census Bureau's Decennial PUMS Data

The U.S. Census Bureau makes data collected from the "long form" questionnaires completed by individuals during the decennial Census available to researchers. Sampled persons are identified in PUMS areas of 100,000 or more to protect confidentiality. The PUMS files contain records representing 5 percent of the occupied and vacant housing units in the United States and the people in the occupied units. People living in group quarters are also included in the sample. The records include a large amount of data about persons and the housing units in which they live. The file contains individual weights for each person and housing unit, which, when applied to the individual records, expand the sample to the relevant total.

The person records provide a number of items useful for identifying CHWs, and codes for occupations and the industries in which persons work. A crosswalk between Census codes for industry/NAICS and occupation/SOC, where NAICS is based on the 2002 North American Industry Classification System and SOC is based upon the 2000 Standard Occupational Classification System, is available and described later in this appendix.

While there is no specific occupation code for CHWs, there were a limited number of occupations in which CHWs may be classified. By selecting these and the industries in which CHWs were most likely to be employed, it was reasonable to expect that this set of workers could be identified.

Bureau of Labor Statistics "Staffing Patterns" Data

The U.S. Department of Labor, Bureau of Labor Statistics (BLS), has rigorous guidelines in each State to survey firms in order to collect detailed SOC occupational content in each type of NAICS-based industry category. The resulting statistics are called the "staffing patterns" for each industry. Every year, each State surveys one-third of all of its industries under strict sampling guidelines set forth by the statistical sampling techniques based on employment concentrations. After 3 years, an entire round of all industries has been updated. Considerable effort is made by each State to deliver survey results that meet these specified criteria, and follow-up activity is utilized to overcome any shortfalls in sampling.

The percentages from these patterns are used to calculate occupational estimates for State and sub-State areas. Since each State must cooperate in these efforts set forth by BLS, the staffing patterns represent a common methodologically collected series of statistical base ranges upon which to calculate estimates and - even more importantly - projections.

Selection of Data Sources for the Estimates

Prior to calculating any estimates of paid CHWs from existing data sources, the Institute for Demographic and Socioeconomic Research (IDSER) at the University of Texas at San Antonio, collaborating with the research team for this component of the study, reviewed possible sources of data to determine which would be used to produce the estimates. This list included the 2000 Census, 2000 Public Use Microdata Sample (5 percent sample), and the 2004 American Community Survey.

Given the NAICS codes identified in the CHW report for inclusion in the estimates, the data sources were reviewed to determine the industry coding or NAICS equivalencies as used by each of the data sources. This process helped identify which data set included the most detailed information for the NAICS categories of interest.

Prior to finalizing this decision, IDSER reviewed the unweighted numbers for the NAICS codes of interest using the 2000 PUMS data. The results were evaluated to ensure sufficient numbers of cases within each industry code to proceed with the CHW estimates. After review, it was determined that the numbers were sufficient to proceed as planned. The PUMS data were then weighted.

Based on the review of the data, it was determined that the 2000 PUMS (5 percent sample) from the decennial Census would be the most appropriate data source for the CHW estimates. Because earlier discussions had also suggested the use of estimates based on BLS data, it was decided that such estimates would be prepared and compared to the PUMS-based estimates.

Standard Occupation Classification (SOC)

According to the Bureau of Labor Statistics, the 2000 SOC System is used by Federal statistical agencies to classify workers into occupational categories for the purpose of collecting, calculating, or disseminating data.^[1] Workers are classified into one of more than 820 occupations according to their occupational definition. These occupations are then combined to form 23 major groups, 96 minor groups, and 449 broad occupations. The broader occupation includes detailed occupation(s) requiring similar job duties, skills, education, or experience. As per the American Community Survey,^[2] another survey conducted by the Census Bureau, Census occupation codes are classified into 23 major occupational groups based on the Standard Occupational Classification (SOC) Manual: 2000, published by the Executive Office of the President, Office of Management and Budget.

While the Census Bureau uses its own classification system for occupations, that is, Census Occupational Categories, the Census makes available a crosswalk^[3] to show how occupation codes used by the Census correspond to the SOC. The following is an example from a previous Census Bureau crosswalk:

In addition, the Census Bureau makes available descriptions for each SOC code^[4] when it is linked to occupation codes used by the Census. For instance, Social and Human Service Assistants (SOC 21-1093) corresponds to code 2020^[5] where the following SOC description is provided:

Assist professionals from a wide variety of fields, such as psychology, rehabilitation, or social work, to provide client services, as well as support for families. May assist clients in identifying available benefits and social and community services and help clients obtain them. May assist social workers with developing, organizing, and conducting programs to prevent and resolve problems relevant to substance abuse, human relationships, rehabilitation, or adult day care. Exclude "Rehabilitation Counselors" (21-1015), "Personal and Home Care Aides" (39-9021), "Eligibility Interviewers, Government Programs" (43-4061), and "Psychiatric Technicians" (29-2053).

A useful tool for viewing occupational descriptions was available from the Occupational Information Network (O-Net).^[6] O-Net, available online, allows users to select specific SOC codes containing descriptions and classifications of job requirements and worker competencies. In addition, O-Net makes available crosswalks, enabling users to "convert" several widely used occupation coding systems to current SOC codes.^[7]

The descriptions used by Census, BLS, and O-Net regarding occupations were the same.

Identification of Occupations for Estimates of Paid CHWs

Since the current SOC System did not have a job title or an occupational category specifically named, or designated for, community health workers, using descriptions of the work performed by CHWs and the tools provided by O-Net, Census Bureau, and BLS, it was possible to identify occupational titles in which CHWs were most likely to be classified in the current data collection systems. It was assumed that information collected by both the Census Bureau and BLS about CHWs was stored in existing SOC categories for workers with job duties that were similar to or overlapped with those of CHWs.

Using the descriptions provided in the literature of the roles and functions fulfilled by working CHWs, as well as the titles used to identify CHWs, the research team identified an initial set of occupational classifications (SOC codes). In addition, skills outlined in the National Community Health Advisor Study^[8] and other studies provided guidance in identification of CHWs within SOC codes. These skills included:

• Advocacy skills - Ability to "speak up" for patients and communities to overcome barriers; ability to act as an intermediary with bureaucracy
• Bilingual skills - Fluency in the preferred language of clients and ability to translate technical terms
• Capacity-building skills - Empowerment skills; leadership skills; ability to influence communities and individuals to change behavior and take more control of their own health
• Communication skills - Ability to listen and use oral and written language confidently
• Computer skills - Performing data entry and using the Internet to locate health information
• Confidentiality skills - Ability to keep matters private, comply with HIPAA laws
• Interpersonal skills - Friendliness, sociability, counseling and relationship building skills; ability to provide support and set appropriate boundaries
• Organizational skills - Ability to set goals and develop an action plan, manage time, keep records
• Service coordination skills - Ability to identify and access resources; ability to network and build coalitions; ability to make and follow up on referrals
• Teaching skills - Ability to share information, respond to questions, and reinforce ideas; ability to adapt methods to various audiences

From all available occupational titles, those that best fit the CHW tasks described in the literature were those of persons employed in health and social services occupations (SOC codes 21-, 29-, 31-). These categories included:

21-0000 Community and Social Services Occupations
21-1011 Substance Abuse and Behavioral Disorder Counselors
21-1012 Educational, Vocational, and School Counselors
21-1013 Marriage and Family Therapists
21-1014 Mental Health Counselors
21-1015 Rehabilitation Counselors
21-1019 Counselors, All Other
21-1090 Miscellaneous Community and Social Service Specialists
21-1091 Health Educators
21-1092 Probation Officers and Correctional Treatment Specialists
21-1093 Social and Human Service Assistants
21-1099 Community and Social Service Specialists, All Other

29-0000 Healthcare Practitioners and Technical Occupations
29-1000 Health Diagnosing and Treating Practitioners
29-1010 Chiropractors
29-1020 Dentists
29-1030 Dietitians and Nutritionists
29-1040 Optometrists
29-1050 Pharmacists
29-1060 Physicians and Surgeons
29-1070 Physician Assistants
29-1080 Podiatrists
29-1110 Registered Nurses
29-1120 Therapists
29-1130 Veterinarians
29-1190 Miscellaneous Health Diagnosing and Treating Practitioners

29-2000 Health Technologists and Technicians
29-2010 Clinical Laboratory Technologists and Technicians
29-2020 Dental Hygienists
29-2030 Diagnostic-Related Technologists and Technicians
29-2040 Emergency Medical Technicians and Paramedics
29-2050 Health Diagnosing and Treating Practitioner Support Technicians
29-2060 Licensed Practical and Licensed Vocational Nurses
29-2070 Medical Records and Health Information Technicians
29-2080 Opticians, Dispensing
29-2090 Miscellaneous Health Technologists and Technicians

29-9000 Other Healthcare Practitioners and Technical Occupations
29-9010 Occupational Health and Safety Specialists and Technicians
29-9090 Miscellaneous Health Practitioners and Technical Workers

31-0000 Healthcare Support Occupations
31-1010 Nursing, Psychiatric, and Home Health Aides
31-2000 Occupational and Physical Therapist Assistants and Aides
31-2010 Occupational Therapist Assistants and Aides
31-2020 Physical Therapist Assistants and Aides

31-9000 Other Healthcare Support Occupations
31-9090 Miscellaneous Healthcare Support Occupations

The review found that the occupations associated with health care required that the persons have specific professional training or technical skills or be associated with providing direct personal services. Therefore, two occupations categories that initially appeared to hold potential as CHW occupations were eliminated when job descriptions from O-Net were examined. These occupations were: Other Healthcare Practitioners and Technical Occupations (29-9000) and Healthcare Support Occupations (31-0000).

The occupation category that held the most potential to identify community health workers was Community and Social Services Occupations (21-0000). The specific occupations category that seemed likely to be useful in identifying CHWs in PUMS data was Miscellaneous Community and Social Service Specialists (21-1090). Three of the four occupations in this category seemed best suited as occupations most closely related to work CHWs do:

21-1091 Health Educators
21-1093 Social and Human Service Assistants
21-1099 Community and Social Service Specialists, All Other

The selected set was reviewed again by members of the research team and some of their advisors for further assessment of occupational titles. A second set of codes was identified as possible occupational categories in which CHWs might be classified:

21-1010 Counselors
21-1090 Miscellaneous Community and Social Service Specialists
29-1129 Therapists, All Other
29-2090 Miscellaneous Health Technologists and Technicians
29-9000 Other Health Care Practitioners and Technical Occupations
31-909X Medical Assistants and Other Health Care Support Occupations

The list of classified positions was then reviewed by a task force of individuals identified by the office of the State of Texas Regional and Local Services Division. This office oversees CHW accreditations in Texas.^[9] The individuals of the task force were chosen because of their knowledge about CHWs and professional interest in this emerging workforce. This review task force included representatives from:

• Regional and Local Services Division, Texas Department of State Health Services
• Forecasting and Research, Texas Health and Human Services Commission
• Center for Health Statistics, Texas Department of State Health Services

In comparing the descriptions of specific subgroups for specific occupational categories to the work that CHWs perform, members of the Texas task force identified the following categories as those most likely to include CHWs:

21-1011 Substance Abuse and Behavioral Disorder Counselors
21-1012 Educational, Vocational, and School Counselors
21-1014 Mental Health Counselors
21-1019 Counselors, All Other
21-1091 Health Educators
21-1093 Social and Human Service Assistants
21-1099 Community and Social Service Specialists, All Other

The review by the Texas task force was followed by an informal review by the research team of the Center for Health Workforce Studies (CHWS) at the State University of New York at Albany, which had been engaged as a subcontractor for another component of the study. The Albany research team had past experience using data from the BLS and agreed with the SOC codes identified by the Texas task force.

In conclusion, the SOC codes in Table B.2 were those found most likely to include employed CHWs.

Table B.2 Standard Occupational Classification (SOC) Code Included in Methodology for Estimates of Paid CHWs

*Broad categories reported in PUMS data.

North American Industry Classification System (NAICS)

According to the Census Bureau,^[10] "Federal statistical data published for reference years beginning on or after January 1, 2002, should be published using the 2002 NAICS United States codes. Agencies may adopt the 2002 NAICS earlier at their discretion. Publication of a 2002 NAICS United States Manual is planned for January 2002." Some of the new features of NAICS 2002 include: (1) Relevance: new, emerging, and advanced industries are included; (2) International comparability: Canada and Mexico both cooperated in development of the latest industry classification; and (3) Consistency: businesses that use similar production processes are grouped together.^[11]

Similar to a crosswalk provided by O-Net for occupational categories, a crosswalk was available from the Census Bureau to assist in converting current NAICS to previous versions of NAICS and to Standard Industrial Classification (SIC).^[12]

Identification of Industries for Estimates of Paid CHWs

Since the SOC code 21-1090, Miscellaneous Community and Social Service Specialists, was found to be the one most promising for identifying CHWs, after selecting only individuals with SOC code 21-1090, a list of industries was produced using PUMS data:

5241 Insurance Carriers
6112 Junior Colleges
6113 Colleges, Universities, and Professional Schools
6213ZM Offices of Other Health Practitioners
621M Other Health Care Services
622 Hospitals
623M Residential Care Facilities, Without Nursing
6242 Community Food and Housing, and Emergency Service
6244 Child Day Care Services
6243 Vocational Rehabilitation Services
711 Independent Artists, Performing Arts, Spectator Sports
712 Museums, Art Galleries, Historical Sites, and Similar Institutions
713Z Other Amusement, Gambling, and Recreation Industries
721M Recreational Vehicle Parks and Camps, and Rooming and Boarding Houses
8121M Nail Salons and Other Personal Care Services
8129 Other Personal Services
8132 Grant Making and Giving Services
8133 Social Advocacy Organizations
81393 Labor Unions
813M Civic, Social, Advocacy Organizations, and Grant-Making Institutions
814. Private Households
9292 State Government, Exclusive Education and Health
9393 Local Government, Exclusive Education and Health

While CHWs may work in these industries, it was found to be unlikely that hospitals, offices of health practitioners, other health care services, etc., were actually employing many persons who work in the community. The review reduced the list of industries likely to employ community health workers to the following: Community Food and Housing, and Emergency and Other Services (6242) and Civic, Social, Advocacy Organizations, and Grant-Making Institutions (Census NAICS 813M). The BLS provided sub-categories for NAICS 813M: Grant-Making and Giving Services (8132), Social Advocacy Organizations (8133), and Labor Unions (81393).

After the SOC codes were reviewed and the additional SOC of 21-1010 was identified, the research team repeated the process described earlier to identify another set of industries (NAICS codes) most likely to employ CHWs.

As with the SOC codes, the list of NAICS codes identified were reviewed by members from the Texas task force, the research team in New York, and members from the Center for Sustainable Health Outreach^[13] (CSHO) at the University of Southern Mississippi.

Matching of Verified Employers to State Employment Data Sets

Each State collects employer and their wage and salary employee records in order to collect payments on unemployment compensation. This information is collected for almost 97 percent of all persons working in the civilian labor force. Under special confidentiality agreement for this project, it was possible to use a selected number of records to match the names of organizations verified as employers of CHWs to the BLS list of employers.

The record of verified employers in 10 States^[14] (N=1,327) was used to locate the employer record in the American Labor Market Information System. The matching process started with searching one verified record at a time by telephone number, followed by address, city, and last by organization name.

Of the verified employers in these selected States, 57 percent (759 of 1,327) were successfully matched against the employer records database. For the successfully matched records, 92 percent (701 of 759) corresponded to the industries identified for inclusion in the estimates of paid CHWs. The industries with the most overlap included 6214 or Outpatient Care Centers, 6241 or Individual and Family Services, 8133 or Social Advocacy Organizations, and 9231 or Administration of Education Programs.

The processes just described resulted in the identification of NAICS codes most likely to include employed CHWs. They are listed in Table B.3.

Table B.3 North American Industry Classification System (NAICS) Codes Included in Methodology for Estimates of Paid CHWs

BLS data contain two BLS-designated NAICS codes, 9992 - State Government sector and 9993 - Local Government sector, which are not found in the 2000 PUMS data. It was determined that these two BLS-designated NAICS categories are comparable to the NAICS categories 9211 - Executive, Legislative, and General Government and 9231 - Administration of Education Programs and will be used as counterparts in the State and national comparisons.

Data from the BLS were used to determine a proportion of individuals employed in each NAICS industry in cases where two or more NAICS codes were combined in the 2000 PUMS data. This assumed that the proportions of the combined NAICS industry codes represented in the BLS data were appropriate to use with the 2000 PUMS data.

Adjustment factors

The list of SOC and NAICS codes provided the occupational/industry categories within which employed CHWs were likely to be classified.

Next, it was determined that an adjustment factor should be employed to reflect the proportion of persons working as community health workers within the identified occupation and industry combinations.

As for the selection of the occupational codes, the adjustment factors were selected by the research team in consultation with their technical advisors and State task force. The agreed-upon proportions were used to adjust the estimate of persons working in each NAICS/SOC category obtained from PUMS or BLS.

The adjustment factors used in the final set of estimates are shown in Table B.4.

Table B.4 Adjustment Factors Applied to NAICS/SOC Categories for Final Estimates

Table B.4 reflects the fact that, as with SOC categories, the Census Bureau collapses some detailed NAICS codes into a single broad industry category. For the final estimates, the adjustment factor for the broad category of 813M as used by PUMS was assigned to all the specific NAICS categories provided in the BLS data.

Reviews of the Methodology

Several groups reviewed the entire methodology in different stages of completion before the final estimates were computed by the IDSER/UTSA group. Those who participated in the reviews included the Texas State task force, research team members from the CHWS in New York, and specialists at the CSHO from The University of Southern Mississippi.

The methodology was shared with the entire State task force at a meeting in January 2005. Since Texas was the only State in the Nation requiring the certification of CHWs, availability of that group was significant as it included very knowledgeable individuals about CHWs, researchers, statisticians, and CHWs. The State task force members are listed in Table B.5.

Table B.5 List of Texas State Task Force Members

The study advisory group reviewed the entire methodology for the first time on May 25, 2005. During the meeting, five individuals were selected to participate in a special task force. Members included: Ms. Theresa Cosca, Bureau of Labor Statistics; Dr. Susan Chapman, University of California at San Francisco; Dr. Frederick Chen, University of Washington at Seattle; Dr. Steve Murdock, University of Texas at San Antonio; and Dr. Robert Giffin, Insititute of Medicine of the National Academies. Members of the Estimates Task Force were briefed twice in early 2006 on the progress made to date on the estimates. The final estimates were reviewed during a conference call held on August 29, 2006.

The Estimates

Table B.6 shows a comparison of the estimates of the total number of community health workers (CHWs) nationwide for each of the NAICS codes of interest for the 2000 PUMS and 2000 BLS data. Table B.7 shows the combined totals of CHWs for all NAICS categories by State. Summing the State estimates produced a national estimate of the number of paid CHWs.

Table B.6 CHW Estimates, National Comparison by NAICS Code, PUMS-Based and Staffing Patterns-Based

Table B.7 Estimates of Paid CHWs, PUMS-Based and Staffing Patterns-Based

The difference between the estimates from the Census PUMS and the BLS Staffing Patterns Survey shown in Tables B.6 and B.7 provides an estimated range for the number of CHWs working within a State. The results indicate that the aggregate estimates of CHWs based on BLS and PUMS Census data are similar at the national level (only a 6.1 percent difference between the two estimates), but they differ substantially for some States.

The estimates shown in Chapter 3 of the report are averages of the BLS- and PUMS-based estimates.

Estimates of Volunteer Community Health Workers

There were no existing databases containing information on the number of CHWs who were serving the community in a volunteer capacity. The only information available was an estimate of all volunteer workers by State.^[15]

Two sources of data were used to calculate the estimates of volunteer CHWs: the percent of paid workers from the CHW National Employer Inventory (CHW/NEI) and the estimates of paid CHWs calculated using PUMS and BLS data (discussed above).

The number of CHWs by paid and volunteer status was extracted for every State from the CHW National Employer Inventory. The States were then clustered based on geographic location into four groups designated as "Census Regions": Northeast, Midwest, South, and West.

Using the number of paid community health workers reported in the CHW/NEI, the proportion of paid workers was calculated for each Census Region and State.

The standard deviation for the four Census Regions was calculated, followed by a standardized score for every State. This process was carried out in order to identify those States with an extreme proportion of paid CHWs (either too large or too small) as compared to the regional average. According to the CHW/NEI Inventory, 67 percent of CHWs across the United States received compensation by an employer.

The proportion of paid community health workers using results from the CHW/NEI and estimates of paid CHWs were then used to calculate a total number of CHWs by Census Region and State using the following formula:

(Number of paid CHWs from estimates x 100) /

Proportion of paid CHWs from the CHW/NEI

Adjustments for the proportion of paid workers were made for States that were at least 1.0 unit from the standard deviation; reported a proportion of 100 percent paid from the Inventory; or had no responses to the Inventory (only one State). Adjustments were made as follows:

• One standard deviation was added to the proportion of paid CHWs from the Inventory for a negative standard deviation of 1.0 or greater.
• One standard deviation was subtracted from the proportion of paid CHWs from the Inventory if a positive standard deviation of 1.0 or greater was reported.
• One standard deviation was subtracted from the proportion of paid CHWs for States reporting a workforce that was 100 percent paid.

The number of volunteer CHWs was then calculated by subtracting the number of paid CHWs from the total number of CHWs. Table B.8 shows the number of volunteer CHWs per State.

Table B.8 Estimates of Volunteer CHWs

[1] Standard Occupational Classification (SOC) System [Internet]. Washington (DC): Bureau of Labor Statistics, U.S. Department of Labor; 2000 [updated 2006 Jun 13/cited 2005 Dec 14]. Available from http://www.bls.gov/soc/.

[2] The American Community Survey is expected to replace the decennial Census "long form" questionnaire beginning in 2010.

[3] Census 2000 Occupational Categories, With Standard Occupational Classification (SOC) Equivalents, Census 2000 Code Order [Internet]. Washington (DC): Bureau of Labor Statistics, U.S. Department of Labor; 2001 [updated 2001 Jan 01/cited 2005 Dec 14]. Available from http://www.census.gov/hhes/www/ioindex/occ2000t.pdf.

[4] Industry and Occupation 2002 [Internet]. Washington (DC): U.S. Census Bureau, Housing and Household Economic Statistics Division; 2005 [updated 2005 Mar 08/cited 2005 Dec 14]. Available from http://www.census.gov/hhes/www/ioindex/ioindex02/txtnew02.html#21-1011.

[5] In 2002, industry and occupation codes used by the Census Bureau underwent a major renovation from three-digit to four-digit codes in order to accommodate the possible additions of new industries and occupations.

[6] O-Net, administered and sponsored by the U.S. Department of Labor's Employment and Training Administration, is a comprehensive database system that replaced the Dictionary of Occupational Titles as the primary source of occupational information.

[7] Occupational Information Network O-Net Online: O-Net Online Help Crosswalk [Internet]. Washington (DC): National Center for O-Net Development; [updated 2005 Dec 14]. Available from http://online.onetcenter.org/help/online/crosswalk.

[8] Rosenthal EL, Wiggins N, Brownstein JN et al. The Final Report of the National Community Health Advisor Study. Tucson (AZ): University of Arizona, 1998.

[9] At the time, Texas was the only State in the country that required certification of CHWs when CHWs were compensated for their work.

[10] Office of Management and Budget. North American Industry Classification System-Revision for 2002; Notice. Fed Regist 2001; 66 (10).

[11] Ibid.

[12] 2002 NAICS United States Structure, Including Relationships to 1997 NAICS United States and 1987 Standard Industrial Classification [Internet]. Washington (DC): U.S. Census Bureau; 2002 [updated 2004 Mar 23/cited 2005 Dec 14]. Available from http://www.census.gov/epcd/naics02/naicod02.htm.

[13] CSHO was the partner in the study.

[14] States included: California, Colorado, Connecticut, Florida, Georgia, Hawaii, Kansas, New Jersey, New York, and Texas.

[15] Points of Light Foundation Announces State Volunteering Rates, Research Highlights Impact of Volunteer Center National Network [Internet]. Washington (DC): The Points of Light Foundation; 2004 [updated 2004 Sep 13/cited 2005 Dec 14]. Available from http://www.pointsoflight.org/about/mediacenter/releases/2004/09-13.cfm.