Health information disparities? determining the relationship between age, poverty, and rate of calls to a consumer and patient health information service * {

Growth and demand for consumer health information services has been well documented [1–4]. In the past twenty years, medical librarians have transitioned from cautious providers of consumer and patient health information to leaders of outreach programs in which they coordinate with medical professionals and public librarians to serve their communities' health information needs. So prevalent are these services that the Medical Library Association provides guidance on its Consumer and Patient Health Information Section (CAPHIS) website [5], and the American Library Association has issued a policy to guide the growing number of services in providing consumer health information [6]. 
 
Beginning with a grant from the Tennessee State Library and Archives, Preston Medical Library, located in Knoxville, Tennessee, has provided a free telephone-based consumer health information service for area citizens and University of Tennessee Medical Center patients and their families since 1989. Despite the wide availability of the Internet, a recent survey of consumers in Tennessee funded by the National Network of Libraries of Medicine (NN/LM), Southeastern/Atlantic Region, to which 233 individuals responded, showed that the inability to identify reliable sources was the single most important stumbling block to finding health information. In the same survey, 37% of respondents ranked libraries as one of their top 3 sources of health information along with doctor's office (61%); Internet and magazines (54%); and newspapers, magazines, and books (25%) [7]. A recent Pew report on the demographics of Internet users shows that while 71% of all adult Americans go online, only 32% of those over 65 and just 55% of those living in households with less than $30,000 in annual income have Internet access [8]. Because a library-based telephone reference service is a low technology way to obtain health information and libraries are a trusted source of health information, populations who are not using the Internet are an important target for the Preston Medical Library's consumer health service. 
 
Preston markets its service in various ways. Brochures in physicians' offices and flyers sent to public health departments promote the service to outpatients. In the hospital, information is provided in new patient packets. The library website includes a page describing the Preston Consumer and Patient Health Information Service and suggests other reliable consumer health sites. Preston partners with Knox County Public Library to distribute brochures and provide train-the-trainer classes to staff. Librarians exhibit at community events and speak to consumer groups. 
 
This project was undertaken to analyze the demographics of those who called the service in order to assess outreach methods, and change them if needed, and was designed to reach two target groups without Internet access: the elderly and those living below the poverty level. Demographic information could also be helpful in obtaining additional funding to reach these target groups.


INTRODUCTION
Growth and demand for consumer health information services has been well documented [1][2][3][4]. In the past twenty years, medical librarians have transitioned from cautious providers of consumer and patient health information to leaders of outreach programs in which they coordinate with medical professionals and public librarians to serve their communities' health information needs. So prevalent are these services that the Medical Library Association provides guidance on its Consumer and Patient Health Information Section (CAPHIS) website [5], and the American Library Association has issued a policy to guide the growing number of services in providing consumer health information [6].
Beginning with a grant from the Tennessee State Library and Archives, Preston Medical Library, located in Knoxville, Tennessee, has provided a free telephone-based consumer health information service for area citizens and University of Tennessee Medical Center patients and their families since 1989. Despite the wide availability of the Internet, a recent survey of consumers in Tennessee funded by the National Network of Libraries of Medicine (NN/LM), Southeastern/Atlantic Region, to which 233 individuals responded, showed that the inability to identify reliable sources was the single most important stumbling block to finding health information. In the same survey, 37% of respondents ranked libraries as one of their top 3 sources of health information along with doctor's office (61%); Internet and magazines (54%); and newspapers, magazines, and books (25%) [7]. A recent Pew report on the demographics of Internet users shows that while 71% of all adult Americans go online, only 32% of those over 65 and just 55% of those living in households with less than $30,000 in annual income have Internet access [8]. Because a library-based telephone reference service is a low technology way to obtain health information and libraries are a trusted source of health information, populations who are not using the Internet are an important target for the Preston Medical Library's consumer health service.
Preston markets its service in various ways. Brochures in physicians' offices and flyers sent to public health departments promote the service to outpatients. In the hospital, information is provided in new patient packets. The library website includes a page describing the Preston Consumer and Patient Health Information Service and suggests other reliable consumer health sites. Preston partners with Knox County Public Library to distribute brochures and provide train-thetrainer classes to staff. Librarians exhibit at community events and speak to consumer groups.
This project was undertaken to analyze the demographics of those who called the service in order to assess outreach methods, and change them if needed, and was designed to reach two target groups without Internet access: the elderly and those living below the poverty level. Demographic information could also be helpful in obtaining additional funding to reach these target groups.

METHODOLOGY
In 2003, an SQL database was created of all calls made to the service since 1998. The database includes the caller's name and address and the health topic on which the caller requested information. For privacy reasons, the topic and contact information are kept in separate, linked tables. The database has been kept current and could therefore be used to answer the research question: Is there a relationship between the rate of calls to the service and the poverty level or proportion of population over sixty-five in a specific geographic area?
A retrospective analysis of the database was undertaken for calls from consumers in Tennessee zip codes who contacted Preston Medical Library between January 1, 1999, and June 1, 2007. The database included 1,938 complete records for calls received during the 8 years studied, representing 107 Tennessee zip codes. Out-of-state zip codes were not included. The database was queried to determine the number of calls from each of the zip codes.
Because population varied in size between zip codes, direct comparison of number of calls could be misleading. To correct for the variation, the rate of calls was determined by dividing the number of calls by the population of the zip code (number of calls/ population5rate of calls). US Census Bureau data from 2000 were used to establish the population for each zip code. Because age was not included in the database, rather than restricting to certain age groups, the total population of each zip code was used.
Once determined, the rate of calls by zip code was compiled into tabular format and compared to US Census data from those same zip codes for the following socioeconomic elements: percent of population below poverty level and percent of population above sixty-five years (Table 1 online). US census data is available online through Census 2000 U.S. Gazetteer File [9]. Data were analyzed using SPSS to determine if percent of population either below poverty level or over sixty-five were related to the call rate.
Data were examined for normality using both the Shapiro-Wilk test and visual inspection using histograms. Neither call rate nor poverty level was found to be distributed normally, but proportion of population over 65 was. Outliers were found for both call rate (0.04) and proportion of families living in poverty (0.47). Using scatterplots, neither variable related to call rate in a way that could be easily modeled. As a result, all variables were divided into low and high categories by splitting the distribution for each variable at the median (percentage living below poverty level median510.2%, percentage over 65 median513.5% , call rate median50.0009176).

RESULTS
Poverty was significantly related to call rate (Pearson chi-square(1)512.803, P50.0003), with only 32.1% of calls originating from zip codes where the percentage of individuals living below the poverty level was greater than 10.2%, while 66.7% of the calls came from zip codes where the percentage living in poverty was lower than 10.2%. There was no significant relationship between the proportion of population over 65 and call rate.

DISCUSSION
According to the State of Tennessee's Comptroller's Office, Tennesseans rank below the national average in literacy based on the National Adult Literacy Survey standards [10]. In fact, 53% of Tennesseans are classified in the bottom 2 of the 5 levels of adult literacy [11]. Because health literacy involves the ''degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions,'' lower literacy populations have challenges reading medical terminology, prescription instructions, and other patient education material [12]. Tennesseans are higher than the national average in terms of the prevalence of high-risk health behaviors and chronic diseases [13]. Librarians at the Preston Medical Library Consumer and Patient Health Information Service are aware of both the literacy problems and high prevalence of chronic diseases in their community and could provide health information at an appropriate literacy level to callers.
Although all demographic groups contain some individuals with low literacy, demographic groups likely to have low literacy rates are those who are greater than 65 years of age and those who live on limited incomes [14][15][16]. In addition, as people age, increasing health problems may increase their need for health information. In fact, the Centers for Disease Control and Prevention reported in 2008 that over the past 36 years, the percent of hospital inpatients who were 65 years of age and older grew from 20% to 38%, demonstrating an increase in health problems and probable subsequent need for information [17]. Although the authors expected calls from those who were over age 65 and those who live in poverty would occur at a higher rate, this was not the case. Several factors are considered below and offer further research opportunities.
A previous survey of health information needs of Tennesseans funded by the NN/LM, Southeastern/ Atlantic Region, in which 78% of the respondents were women, found that most women (82.4%) frequently sought information to address the health concerns of friends or family. Women requesting information for friends or family members may live in different zip codes than those they are assisting [7]. More adult children of aging parents are becoming caregivers [18], and this factor may have contributed to the absence of a relationship between the rate of calls to the service and the percentage of population of over 65 years of age. Calls may have come instead from younger caregivers in other zip codes. Declining mobility also may have limited the elderly population's exposure to the library's outreach efforts. In addition, some low health literacy patients use surrogate readers, such as family members, to request and understand health information, and these surrogates and family members may live in other zip codes [19]. Further research must be undertaken to explore the impact of these factors on the results.
Changes in the library's outreach practices have already been implemented with the specific aim of increasing calls from these target populations. Examples of current targeted outreach include exhibiting at health fairs during ''senior days,'' publishing related stories in community newspapers, and working with the rural public library systems to implement trainthe-trainer programs for frontline personnel. Additional planned changes in practice include distributing newly designed, easy-to-read brochures at locations more likely to reach those who are living in poverty and are above the age of sixty-five.

LIMITATIONS
This study looked at 1,938 complete records from 107 zip codes in the Knoxville area collected over a period of 8 years, and the small sample size might limit the validity of the results. Additionally, although calls could be identified as coming from zip codes with high or low percentages of the population below poverty level or above 65 years, there was no way to know if individual callers were actually impoverished or elderly.

CONCLUSION
Results show that zip codes with populations below the median poverty level had a call rate below the median. Several factors might contribute to this low call rate, including lack of knowledge of the consumer health service and a reluctance to request written material that would prove challenging to read for low literacy populations such as those living in poverty. There was no relationship between rate of calls and the percentage of population over sixty-five years old. This might be due to others requesting information on the behalf of the elderly and/or the inability to reach them through current promotion methods due to cognitive and mobility declines.
Because quality health information has been shown to contribute to all aspects of disease prevention and to better health care outcomes [20][21][22], outreach and promotion of the Preston Medical Library's Consumer and Patient Health Information Service will be altered to increase awareness of the service to assist in decreasing the identified health information disparity.