Participation1 apha 05

Participation in a Nursing Home Admission Decision Among Working Age Individuals: 

Participation in a Nursing Home Admission Decision Among Working Age Individuals Nancy A. Miller, Ph.D. Charlene Harrington. Ph.D. Ab Brody, R.N. Yu Kang. M.P.A. American Public Health Association Annual Meeting, December 14, 2005 Funded by the National Institute of Disability and Rehabilitation Research

Introduction: 

Introduction The lifetime risk for a nursing home admission is substantial Working age individuals (ages 18-64) are a growing segment of the nursing home population Limited information is available on the circumstances associated with admission and their involvement in this health care decision

Working Age Individuals in Nursing Homes: 

Working Age Individuals in Nursing Homes Approximately 10 percent of the nursing home population is working age Relative to older individuals, younger individuals are: Less often married (44.6% vs. 11.8%) More often minority (27.3% vs. 12.9%) More often male (50.4% vs. 25.7%) More often insured by Medicaid (75.8% vs. 56.8%) Less often in need of assistance with ADLs (13.4% vs. 4.1%) (DHHS, 2002)

Working Age Individuals in Nursing Homes: 

Working Age Individuals in Nursing Homes The clinical profile of working age individuals varies by age Younger adults typically experience conditions arising from mental retardation and developmental disabilities or acute trauma Older adults experience chronic conditions such as diabetes and cardiac disorders (Fries et al., 2004)

Patient Involvement in Health Care Decision Making: 

Patient Involvement in Health Care Decision Making Increasing emphasis has been placed on involving individuals in health care decisions Ethical reasons – individuals have a right to understand their condition, its prognosis, and treatment options (Sullivan, 2003) Practical reasons -- informed, involved individuals are more engaged in their care, more compliant with treatment recommendations, experience improved outcomes, report greater satisfaction (Epstein, Alper & Quill, 2004; Greenfield, Kaplan & Ware, 1985; Roter, et al., 1997; Sherbourne et al., 1992)

Patient Involvement in Health Care Decision Making: 

Patient Involvement in Health Care Decision Making Physicians do not routinely engage individuals in decision making (Braddock et al., 1999; Epstein, Alper & Quill, 2004) Studies generally suggest that individuals who are younger, white, and of higher SES are more often engaged Individuals in poorer health or with more severe illness are less often engaged

Patient Involvement in Health Care Decision Making: 

Patient Involvement in Health Care Decision Making Limited research has focused on individual involvement in long term care decisions Miller and Weinstein (2005) found individuals’ race, insurance source and knowledge of the health care decision maker were predictors of involvement among newly admitted working age individuals in Maryland

Study Purpose: 

Study Purpose This research extends Miller and Weinstein (2005) to explore if there are geographic differences in perceived participation in the nursing home admission decision Different Medicaid programs Different “safety net” health care systems

Methods: 

Methods Study Sample Nursing homes in seven Maryland counties (n=17), San Francisco (n=2), and Washington, D.C. (n=2) were recruited Participants met the following inclusion criteria: Age 18-64 Cognitively capable of understanding the purpose of the study

Methods: 

Methods Data Collection Semi-structured in-person interviews Sociodemographic characteristics History of medical conditions and functional status Whether the individual could identify the primary medical provider in the decision Self-reported level of involvement in the admission decision

Methods: 

Methods Conceptual framework of Ong, deHaes and Lammes (1995) used to explore the relationships among patient and physician characteristics, nature of the physician/patient relationship (e.g., participatory, parternalistic), process of interaction (e.g., instrumental relative to affective) and patient participation and outcomes

Methods: 

Methods Variables Dependent variable Self reported level of involvement in the nursing home admission decision (Wetle, et al., 1988). Individuals were asked to recall the admission decision process and to rate their actual level of involvement, using a four point Likert scale of being involved “a lot”, “a moderate amount”, “a little”, or “not at all”.

Methods: 

Methods Variables Independent variables Patient Age Gender Race/ethnicity Marital status Education Health insurance

Methods: 

Methods Variables Independent variables Provider Health care provider name. Individuals were asked if they could identify the primary medical provider involved in the nursing home admission decision. Categorized as “yes, could identify by name”, “yes, could identify by role”, or “no, could not identify”. Proxies for the patient/provider relationship.

Methods: 

Methods Variables Independent variables Disease Self-reported whether a physician or nurse had ever told them they had one of 36 conditions, following a self-report list used in other surveys (e.g., Medicare Current Beneficiary Survey). Created five clinical conditions identified by Fries, et al. (2004) – diabetes, cardiac conditions, mental retardation/developmental disabilities/seizures, paralysis, and cardiovascular accidents (CVA) with paralysis. Added HIV/AIDS as a condition.

Methods: 

Methods Variables Independent variables Site – Maryland, San Francisco, Washington, D.C.

Methods: 

Methods Analysis Dichotomized 4-point Likert scale Estimated a probit model for dichotomous dependent variable (Griffith, Hill & Judge, 1990) Estimated one model with all independent variables Estimated a second model with variables significant at p<.20

Table 1 Descriptive Characteristics of Participants by Site: 

Table 1 Descriptive Characteristics of Participants by Site Maryland San Francisco Washington, D.C. Patient *Age (Mean, Std) 45.77/9.60 48.76/8.02 51.12/9.60 Male (%) 57.56 68.00 68.00 African American (%) 63.41 39.58 88.00 Other race (%) 4.88 14.58 0.00 *Education (Mean, Std) 11.20/ 2.16 12.41/2.79 11.41/1.85 Never married (%) 52.68 48.00 44.00 Private Insurance (%) 8.87 0.00 4.00 Medicare (%) 18.72 22.00 24.00 Medicaid (%) 44.83 52.00 68.00 Uninsured (%) 27.59 26.00 4.00

Table 1 Descriptive Characteristics of Participants by Site-Continued: 

Table 1 Descriptive Characteristics of Participants by Site-Continued Maryland San Francisco Washington, D.C. Patient disease *Diabetes(%) 27.23 18.00 48.00 Paralysis(%) 25.85 22.00 28.00 Cardiac conditions(%) 49.76 40.00 44.00 Developmental disability/ 25.37 38.00 36.00 epilepsy(%) CVA (%) 8.29 12.00 8.00 *HIV/AIDS (%) 27.23 28.57 8.00 * p≤.05

Factors Associated with Self Reported Participation in the Nursing Home Admission Decision: 

Factors Associated with Self Reported Participation in the Nursing Home Admission Decision Model 1 Model 2 Coef/(SE) Coef/(SE) Patient Age -.03/.01* -.02/.01*** Male -.15/.18 African American .45/.20* .52/.19*** Other race -.05/.38 .13/.37 Education -.03/.04 Separated/divorced -.21/.33 Widowed -.15/.32 Never married .24/.36 Medicare -.99/.42* -.75/.38* Medicaid -.79/.39* -.65/.36* Uninsured -.96/.41* -.82/.37* Provider Provider role .18/.23 .12/.21 Provider name .39/.22* .42/.21** * p≤.10; **p≤.05; ***p≤.01

Factors Associated with Self Reported Participation in the Nursing Home Admission Decision-Continued: 

Factors Associated with Self Reported Participation in the Nursing Home Admission Decision-Continued Model 1 Model 2 Coef/(SE) Coef/(SE) Patient disease Diabetes .29/.21 Paralysis -.10/.22 Cardiac conditions -.04/.18 Developmental disability/ .30/.19 .30/.19 epilepsy CVA .17/.22 Site San Francisco .23/.25 .12/.22 Washington, D. C. -.17/.32 -.26/.29 Pseudo R² 0.09 0.08 Prob>chi² 0.06 0.004 * p≤.10; **p≤.05; ***p≤.01

Discussion: 

Discussion Profile similar to Fries et al. (2004) and Spector et al. (2000) Less well educated, lower income, more often male, more often minority Clinical conditions similar, for the most part Higher rates of HIV/AIDS Some site variation HIV/AIDS Diabetes

Discussion: 

Discussion Self reported involvement somewhat low 44% reported little or no involvement Self reported involvement lower among Washington D.C. participants, although not statistically significant. Those reporting no involvement: Maryland 27.9% San Francisco 34.5% Washington D.C. 40.0%

Discussion: 

Discussion Older individuals reported less involvement, consistent with the literature African America participants reported greater involvement Differs from broader literature on patient participation Type of decision differs Nature of physician/patient relationship differs Only 6 studies of medical decision making systematically assessed the effect of race/ethnicity (Rosu & Miller, 2004)

Discussion: 

Discussion No SES effect Differs from broader literature Participants were relatively low SES, measured by both education and income

Discussion: 

Discussion Insurance source matters Those who were uninsured reported less involvement Those who are uninsured report lower service use (LaPlante, Rice & Wenger, 1995), are admitted to the hospital in poorer health, while in the hospital receive fewer services (Hadley, Steinberg & Feder, 1991) and are at greater risk for substandard care (Burstin, Lipsitz & Brennan, 1992) Lack of insurance limits choices regarding care Organizational and payer policies may create structural constraints to physician/patient interactions (Stewart et al., 1999)

Discussion: 

Discussion Insurance source matters Medicaid effect may be similar to the uninsured effect, regarding limited choices and structural constraints Medicaid effect and Medicare effect may proxy for disease status Physicians may interact differently with their healthy, relative to less healthy patients Communication patterns may vary by disease progression (Ong, deHaes, & Lammes, 1995)

Discussion: 

Discussion Knowledge of the provider matters Individual who could identify the provider by name self reported greater involvement Suggests that, even absent an established relationship, the level of communication associated with knowing the provider increases perceived involvement Problems with communication and relationships with providers are the most frequently reported problems in hospital settings. Problems greatest for individuals of low SES and in poorer health (Cleary et al., 1990)

Discussion: 

Discussion Does participation matter? Working age individuals who perceived involvement in the decision were more satisfied with their site of long term care (Miller & Weinstein, 2005) Among older individuals in nursing homes, perceived involvement has been related to physical status (Reinardy, 1992), satisfaction, and morale (Harel & Noekler, 1982) Mirror previously discussed broader findings, in which involvement is related to compliance, health outcomes and satisfaction

Discussion: 

Discussion Despite potential constraints, previous work has demonstrated the importance of patient involvement There is an ethical obligation to involve patients in their health care decisions Clinical and ethical concerns are buttressed by the legal requirements of Olmstead