Methodology
research design
The Northwest Arkansas Hometown Health Improvement Project in coordination with the Social Work Research Center, the Madison County Health Coalition, the Benton County Community Coalition, and the Carroll County Resource Council had representatives who comprised the Health Data Team.
With the guidance of the Social Work Research Center, the Health Data Team worked to develop an effective tool to assess the health perceptions of the people in the Northwest Arkansas four county area. A survey instrument was developed. The Health Data Team and the coalition members were asked to choose sites that would be favorable to administer the assessment. A plan to administer surveys in Benton , Carroll, Madison , and Washington Counties was put in place. In addition, the survey was translated into Spanish and four targeted focus groups were completed. Therefore, both objective and subjective information was collected.
survey instrument
The Health Data Committee members and the Social Work Research Center (SWRC) staff collaboratively developed the survey instrument. SWRC conducted a comprehensive review of the relevant literature in order to assist in the design of the survey instrument. The Committee also reviewed the existing Hometown Health Guide, "Assessing Your Community's Health" (3) available within the public domain. Based upon a review of the relevant literature, the health data committee identified preliminary health care items to be considered for inclusion in the survey. The Hometown Health Guide also included a sample of community-based surveys, which provided the framework for the final survey tool.
The survey underwent four major revisions and was piloted at the Madison County Health Fair in May 2003 before the formal survey instrument was adopted. The finalized instrument included seventy-two items divided into six sections, along with fifteen demographic questions. The principal sections of the Four-County Community Heath Needs Assessment include the following categories of questions:
Medical Care Costs
Access to Health Care Services
Physical Health Care Issues
Mental Health Care Issues
Multicultural Health Care Issues
General Questions and Demographic Data
Answer sets to the survey questions included both open-ended and closed responses, depending on what was being asked. Health care questions contained within each category required responses from 1 (not an issue) to 4 (a major issue) on a four-point Likert scale. The instrument also included 0 for "don't know" responses. Open-ended questions were also included at the end of each category, providing the participant with an opportunity to elaborate upon their responses in narrative form and to prioritize the importance of their health care concerns. Demographic and general questions required open ended and discriminant responses.
The Four-County Community Health Needs Assessment was administered in both English and Spanish. Several train the trainers sessions were held prior to the administration of the instrument. These sessions were aimed to provide standardized instructions to those administering the survey. These train the trainer sessions occurred in September of 2003. Surveys were administered throughout Northwest Arkansas through the end of January 2004. Public service announcements, advertisements in newspapers, flyers, announcements in meetings and informal communication were utilized to inform the public of the Health Needs Assessment.
Survey sites were initially selected by members of the four county Hometown Health Improvement Coalitions, based upon the unique knowledge of each county's coalition members and recommendations provided by the Social Work Research Center.
Since the collection of the survey data utilized a sample of convenience, volunteers were trained to survey every third participant at a survey site. This strategy was an attempt to ensure that a representative sample would be obtained.
Volunteers were provided with an Interviewer Tip Sheet, a site monitor form, an introduction to the survey, an Informed Consent form and the actual survey instrument. English and Spanish versions of the surveys were provided. Each survey site was staffed by a site monitor and survey interviewers. The site monitor was responsible for securing the survey location, overseeing the actual administration of the survey and providing the necessary information, decisions and materials to ensure smooth data collection. Interviewers were requested to read the survey aloud to respondents. This administration technique was designed to increase the accuracy and completion rates of the surveys. For a complete listing of survey locations, monitors and survey interviewees see Appendices.
A total of 1,445 surveys were administered between September 10, 2003 and January 31, 2004. Eight surveys administered to survey respondents were disallowed due to incompleteness or gross inaccuracy in completing the assessment tool response sets.
Surveys Completed By Each County
County |
English Surveys |
Spanish Surveys |
|---|---|---|
Benton |
332 |
42 |
Carroll |
108 |
6 |
Madison |
146 |
8 |
Washington |
718 |
77 |
Total (by Language) |
1,304 |
133 |
Total Surveys |
1,437 |
|
One hundred twenty-one of the English surveys were verbally translated into Marshallese and administered to members of the Marshall Islander community in Northwest Arkansas.
sampling
After 600 samples had been collected, demographic data obtained from the Health Needs Assessment was analyzed in order to determine the representativeness of the sample participants. Demographic data obtained from the subset of surveys collected, was compared with the 2000 United States Census Data, for Benton , Carroll, Madison and Washington Counties. Results of this analysis revealed significant differences in the gender of survey participants (70% - 75% female vs. 25% to 30% males). This overrepresentation of female respondents in the sample data was contrasted with the equal distribution of the gender contained within the 2000 US Census Data (50% females to 50% males). Based upon the disparity in the gender data obtained from sub sample of the Health Needs Assessment, changes in the type of sites selected for administering the surveys were recommended, by the Social Work Research Center.
Volunteers who administered the Health Care Assessment were encouraged by members of the Social Work Research Center to select survey sites which would be disproportionately frequented by males (e.g. , hardware stores, feed stores, etc. ) in order to obtain a more representative sample. In addition, surveyors were also encouraged to "over sample" populations of interest (e.g. , multicultural populations, elderly, rural population, low socioeconomic groups) who were at risk of not receiving adequate health care services. These recommended adjustments in the choice of survey selection sites, remained in effect through January 2004, when the data collection phase of the Northwest Arkansas Health Needs Assessment was completed.
comparisons to existing health data
The data garnered from this project was compared to existing health care information in order to assess the accuracy of respondents' perceptions regarding health care issues in the region. Responses to items within the Health Care Concerns section were averaged and then compared to available health data information.
focus groups
The Four-County Community Health Needs Assessment utilized four focus groups in order to augment health information obtained from the analysis of the survey data. These focus groups were selected to bolster the survey findings relative to specific subgroups in Northwest Arkansas and to glean further information on those groups. Members of the health data committee and the Social Work Research Center selected subgroups who appeared to experience difficulty in successfully accessing health care services or who offered a unique perspective of the health care issues that affect Northwest Arkansas.
The four focus groups selected for participation in the health needs assessment were:
- First Responders
- First Generation Immigrant Hispanic Men
- Marshall Island Women (Madame Pastors)
- Senior Citizens
The first focus group was composed of first call critical care responders. NWHHIP and SWRC members believed that these healthcare providers could provide a unique perspective on the healthcare concerns currently existing in Northwest Arkansas , as critical care responders are often utilized as the initial access point for health care services.
First generation immigrant Hispanic males were selected because of the rapidly growing Hispanic population in Northwest Arkansas and due to the salient influence that Hispanic men exert within their families and communities.
The Marshall Islander community within Northwest Arkansas is of particular interest to healthcare providers and researchers. Not only does the Marshall Islander community in Northwest Arkansas represent the largest concentration of Marshall Islanders outside their own homeland (Republic of the Marshall Islands) but their residence in the United States provides a unique opportunity to assess the influence of cultural and language issues as they impact their ability to successfully access healthcare. Their presence also provides an opportunity to better understand the influence that Western culture has on their health.
The final focus group was comprised of elderly citizens residing in Northwest Arkansas This focus group was conducted at the board meeting for the Northwest Arkansas Area Agency on Aging meeting in Harrison , Arkansas. Senior citizens were selected to participate in a focus group due to increases in their life expectancies and the advances in medical care resulting in greater numbers of people who are active and healthy into their 80s and beyond. This selection was also influenced by the unique challenges that seniors experience in successfully accessing healthcare services.
Members of the Social Work Research Center facilitated each of the focus group sessions. The format of the focus groups required participants to respond to questions related to their own or their family's experiences with selected health care issues. A member of Northwest Arkansas ' Hispanic community conducted the First Generation Immigrant Hispanic Men focus group in Spanish, while a recognized leading member of the Marshall Islander community conducted the focus group in Marshallese. Each of the four focus group sessions were recorded and translated into English, as needed. A transcription of each of the focus group sessions may be obtained through the Social Work Research Center at the University of Arkansas.
data analysis
Results of the data obtained from the Four-County Community Health Needs Assessment were analyzed utilizing several different techniques. These techniques were designed to compare the survey's sample statistics with known demographic data for the region. They were also designed to help assess the relative importance of health care concerns for the region, to examine the impact of health need issues on select subgroups in Northwest Arkansas and to conduct a qualitative analysis of health care responses.
The statistical data from the survey were compared to existing population demographics in order to ascertain whether or not the sample could be considered representative. Gender, race, socioeconomic status (SES), educational attainment and employment status were compiled for each of the four counties from 2002 US Census data. This analysis provided important information regarding the similarities and differences between the survey's sample data created and known demographic characteristics for Benton , Carroll, Madison and Washington Counties.
Items within each content area were averaged and rank ordered according to their perceived importance, as indicated by the survey scores. The top three items most frequently selected by respondents as a major issue are highlighted in the county profiles of this document.
A Wilkes Lambda analysis was completed for each item to determine the strength of the relationship between respondent subgroups and survey responses. A Tukey post hoc analysis was conducted to indicate the direction and level of significance between the participant subgroups. The Tukey post hoc analysis allows statements to be made about which subgroup is more likely to see a given health care issue as a major issue when compared to the range of responses taken from the other groups. This analysis is summarized within the secondary analysis section of each county profile.
A qualitative analysis of narrative responses contained within the survey was also conducted. Survey respondents were asked to identify the most important perceived health care issue within each county. Additional qualitative analyses of physical health care issues and health care data, as well as a summary of the findings obtained from the focus groups are also provided.
The statistical significance of differences observed between survey responses and participant subgroups was determined at the (P<. 05) level.