WHY DON’T ELDERS ADOPT TWO-FACTOR AUTHENTICATION? BECAUSE THEY ARE EXCLUDED BY DESIGN

Abstract Two-Factor Authentication (2FA) provides effective protection for online accounts by providing efficient and highly robust access control. Adoption and usability, however, remain challenges for such technologies. Most research on 2FA focuses on students or employees in the tech sector. For example, our research with student populations found that lack of adoption was primarily due to a lack user risk concern matched with confidence in their ‘strong’ password strategies. The situation for older adults (> 60 years) was quite different, as we discovered through detailed interviews and think-aloud protocols targeted at understanding the registration, after installation, and their (un)willingness to use 2FA. We focused our research on USB security hardware tokens; additionally, we asked about other 2FA strategies which the participants adopted (if any). Their lack of adoption of the devices stemmed from its shortfall of inclusive design. Most available security tokens which were compliant with tablets have very small form factors; nearly invisible in a purse, and easy to slip through a pocket. The larger security keys are device and browser (Google Chrome) dependent. The organizations which would be most invested in protecting older people -- retirement management funds, the Social Security Administration, Medicare, and banking institutions - reasonably do not adopt 2FA because of its lack of acceptability. Our negative result is that older adults are caught in a negative feedback loop where lack of adoption prevents availability, and vice versa. The positive result is that these concerns are straight-forward to overcome.

assessing positive resources, negative attributes and outcomes of pain. Depression, stress, sleep, resilience, social networks, pain severity and interference were measured. Opioid and other medications were determined from claims. The population was propensity weighted to adjust for survey non-response; weighted to be generalizable to members with diagnosed pain. Multinomial logistic regression was used to determine associations of positive/negative attributes on pain. Among respondents (N=4,161; 29%), prevalence of pain severity and interference for no/mild, moderate and severe categories was 61%, 21% and 18% for severity and 67%, 16% and 17% for interference. In bivariate models adjusted for demographics/health status, negative attributes of depression, stress and poor sleep had stronger associations with pain severity and interference than moderating effects. In full multivariate models, the strongest associations with moderate and severe severity and interference remained depression, stress and sleep. Based on results, multidimensional pain management strategies should include management of depression, stress and poor sleep along with enhancement of positive resources and analgesics as needed for pain management.

HOME CARE SOCIAL WORKERS CLAIM MEDICARE IGNORES PATIENT NEEDS William D. Cabin 1 , 1. Temple University, Philadelphia, Pennsylvania, United States
There is significant literature on the importance of addressing social determinants of health (SDOH) in order to improve health care outcomes. In response, the Centers for Medicare and Medicaid Services (CMS) has expanded Medicare Advantage plans ability to cover SDOH-related services. Medicare home health does not cover SDOHrelated services. A literature review indicates no studies on the nature, significance, or impacts of the lack of SDOH coverage in Medicare home health. This article summarizes an initial, exploratory study to address the literature gap, based on interviews of a convenience sample of 29 home care social workers between January 2013 and May 2014 in the New York City metropolitan area. Results indicate social workers believe the lack of SDOH coverage in Medicare home health results in exacerbation of existing patient conditions; creation of new, additional patient conditions; increased home care readmissions and re-hospitalizations; increased caregiver burden; and exacerbation of patients' mental health and substance abuse needs. Policymakers are urged to consider adding coverage of SDOH to Medicare home health primarily through expanded social work coverage.

TECHNOLOGY ELECTRONIC PERSONAL HEALTH RECORD USE AMONG ELDERLY CANCER SURVIVORS AND NON-CANCER SURVIVORS
Yan Luo 1 , 1. University of Alabama, Tuscaloosa, Alabama, United States Background Electronic personal health records (ePHRs) are potential tools to improve clinical outcomes through increasing patients' self-management. Although elderly people, especially elderly cancer survivors, is a growing population who can benefit from ePHRs, little is known about its utilization among the elderly, particularly among those diagnosed with cancer. Objective By applying Anderson's Behavioral Model of Health Services Use, this study aims to examine and compare the associated factors with ePHRs use among elderly cancer survivors and noncancer survivors. Methods The data collected from the 2018 Health Information National Trends Survey (HINTS) was analyzed. The level of access to ePHRs among the elderly were assessed. Predictors of ePHRs use among elderly cancer survivors and non-cancer survivors were compared by conducting multiple linear regression. According to Anderson's Model, predisposing factors, enabling factors, and need factors were included in the statistical model. Results The overall use of ePHRs remained low among 577 participants (mean = .87, SD = 1.72, range from 0 to 4). Non-Cancer survivors reported lower ePHRs use (mean = .83, SD = 1.77, range from 0 to 4). Race/ethnicity, education, regular health care providers, health insurance, social support, and medical conditions were associated with ePHRs use among non-cancer survivors, while age, gender, social support, and self-reported health status were related to ePHRs use among cancer survivors. Conclusion This study suggests additional efforts to increase ePHRs utilization among the elderly, especially the elderly cancer survivors. The predictive findings reported in this study will contribute valuable implications to enhance the ePHRs use. Two-Factor Authentication (2FA) provides effective protection for online accounts by providing efficient and highly robust access control. Adoption and usability, however, remain challenges for such technologies. Most research on 2FA focuses on students or employees in the tech sector. For example, our research with student populations found that lack of adoption was primarily due to a lack user risk concern matched with confidence in their 'strong' password strategies. The situation for older adults (> 60 years) was quite different, as we discovered through detailed interviews and think-aloud protocols targeted at understanding the registration, after installation, and their (un)willingness to use 2FA. We focused our research on USB security hardware tokens; additionally, we asked about other 2FA strategies which the participants adopted (if any). Their lack of adoption of the devices stemmed from its shortfall of inclusive design. Most available security tokens which were compliant with tablets have very small form factors; nearly invisible in a purse, and easy to slip through a pocket. The larger security keys are device and browser (Google Chrome) dependent. The organizations which would be most invested in protecting older people --retirement management funds, the Social Security Administration, Medicare, and banking institutions -reasonably do not adopt 2FA because of its lack of acceptability. Our negative result is that older adults are caught in a negative feedback loop where lack of adoption prevents availability, and vice versa. The positive result is that these concerns are straight-forward to overcome.

HEALTH LITERACY, MEDICAL NON-ADHERENCE, AND SELF-REPORTED HEALTH PROBLEMS AMONG OLDER INTERNET USERS Gul Seckin 1 , 1. University of North Texas, Denton, Texas, United States
The Internet presents new options for the elderly to gather information to support their health care. Health information gathering is among the major motivations for using the Internet among aging baby-boomers. However, insufficient e-health literacy presents challenges for the aging baby boomers. We examined the extent to which healthrelated internet use and e-health literacy are associated with non-adherence and self-reported negative health outcomes. Respondents were randomly sampled from the largest national online probability-based research panel (N = 710; M= 48.8, SD= 16.4). The age range in our research allowed us to examine the hypothesized associations across the full sample while focusing on older adults (age ≥ 60; N = 194). Older adults with greater e-health literacy reported higher averages for non-compliance because of information obtained from the Internet [(t (194) = 5.06, p ≤ .0001]. Ordinary least squares regression analyses showed that older adults who reported greater averages on health-related internet use reported higher averages on self-reported health problems (β = .292, p ≤ .01). However, women reported fewer health problems (β = -.217, p ≤ .01). Non-adherence with doctor recommendations is a significant positive predictor of selfreported health problem in the full sample (β = .244, p ≤ .0001) but not among older respondents (β = .032, p ≤ .061). Older individuals will make better utilization of the Internet if health professionals guide them to credible sources for health-related information. Empowerment of individuals to utilize the Internet in an informed manner requires addressing their needs for e-health literacy skills.

INTERNET USE AND NEGATIVE HEALTH PERCEPTIONS: THE MODERATING ROLES OF EDUCATION AND HEALTH LITERACY
Gul Seckin, 1 Valarie Bell, 1 and Susan Hughes 1 , 1.

University of North Texas, Denton, Texas, United States
There is a mixed support regarding the effect of Internet use on health and well-being. We estimated the extent to which e-health literacy predicted two domains of negative assessment of well-being: negative affect and self-reported experience of health problems. Respondents were randomly sampled from the largest national online probability-based research panel (N = 710). Hierarchical ordinary least squares regression analyses were employed for hypothesis testing. We computed interaction terms (e-health literacy x strain; e-health literacy x education; and education x strain) as determinants of negative subjective assessment of well-being. Older adults with higher levels of e-health literacy reported significantly more health information consumerism [(t (194) = 7.32, p ≤ .0001] but less strain in medical encounters [(t (194) = 2.92, p ≤ .004]. They reported less negative affect [(t (194) = 2.11, p ≤ .036] and more satisfaction with medical encounters [(t (194) = 4.70, p ≤ .0001]. The effect of perceived strain in medical encounters on negative affect was weaker among those with higher levels of education (β = -.314, p ≤ .01). Education had a significant moderating effect on the association between perceived strain in medical encounters and self-reported health problems, (β = -.550, p ≤ .05). Those who reported lower averages for e-health literacy but higher educational levels indicated lower averages on negative affect (β = -.597, p ≤ .05). Given that conventional methods of acquiring health-related information shift to the Internet, our study holds significant health and social implications for a rapidly growing Internet-connected older population.

DOES CHANGE IN INTERNET USE PREDICT PSYCHOLOGICAL WELL-BEING AMONG OLDER ADULTS?
Ronald W. Berkowsky 1 , 1. California State University Channel Islands, Camarillo, California, United States Previous work focusing on the relationship between Internet use and quality of life among older adults (aged 65+) has found evidence of various positive impacts. This project expands upon this work by examining the relationship between Internet use and measures of psychological well-being (PWB) including autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, and self-acceptance. The analytic sample is derived from two waves of data (Time 1 = 2004, Time 2 = 2011) taken from the Wisconsin Longitudinal Study and includes a sample of older adults aged~65 at Time 1 (N = 4943). Participants were separated into four categories: those who did not use the Internet at Time 1 or 2, those who used the Internet at Time 1 only, those who used the Internet at Time 2 only, and those who used the Internet at both Time 1 and 2. Regression analyses were performed with the Time 2 PWB measures as the outcomes and the Internet use categories as the primary predictors. Results indicate that while continuous Internet users typically reported higher PWB scores compared to non-users, those who stopped use between Time 1 and 2 also reported higher scores and those who started use between Time 1 and 2 reported lower scores. These results generally held when introducing Time 1 PWB measures as controls, suggesting changes in Internet use may affect PWB but not necessarily in the predicted directions. Additional control variables, potential explanations, and implications for future research are discussed.

A COMPREHENSIVE DIGITAL SELF-CARE SUPPORT SYSTEM FOR OLDER ADULTS: A MULTIDISCIPLINARY FRAMEWORK
Priya Nambisan, 1 Kalle Lyytinen, 2 Kurt Stange, 1 Eva Kahana, 2 and Gary L. Kreps 3 , 1. University of Milwaukee,Wisconsin,United States,2. Case Western Reserve University,Cleveland,Ohio,United States,3. George Mason University,Fairfax,Virginia,United States This paper presents an innovative conceptual framework for designing a Comprehensive Digital Self-care Support System (CDSSS) to meet the health needs -physical, mental and social health needs of older adults and their caregivers. Older adults deal with multiple co-morbidities, medications