Effect of Educational Level and Minority Status on Nursing Home Choice After Hospital Discharge

METHODS

RESULTS

Table 1 ▶ summarizes the characteristics of the 95 HSAs in the study and the patient-level characteristics of the sample. Our sampling strategy yielded HSAs with a high percentage of African Americans (24%). The average HSA contained an average of 6 acute general hospitals (SD = 7, range = 1, 44), 20 freestanding nursing homes (SD = 16, range = 4, 75), and 3 hospital-based nursing homes (SD = 4, range = 0, 18). Patient-level characteristics are presented separately by race/ethnicity. A high proportion of African Americans (52%) and other minority elders (54%) did not complete high school. African Americans in our sample were generally younger (28% older than 85 years) than non-Hispanic Whites (37% older than 85 years).

In an effort to validate our quality measure, we compared the worst quartile of nursing homes in each HSA (n = 465) to other nursing homes with fewer deficiencies (n = 1439) across various proxies for facility resources and quality. Compared with nursing homes in the better quartiles of deficiencies, nursing homes in the worst quartile of deficiencies had significantly higher percentages (P < .05) of Medicaid residents (72% vs 63%), significantly lower occupancy rates (85% vs 87%), and significantly higher prevalence of pressure ulcers (11% vs 9%). The worst-quality nursing homes were more likely to be for-profit (78% vs 69%) and to have higher percentages of African American residents (33% vs 22%).

Table 2 ▶ presents the independent effect of being an ethnic/racial minority on the relative risk of admission into the worst quartile of freestanding nursing homes in the 95 HSAs. When we controlled for other factors, African Americans were more likely (relative risk [RR] = 1.26; 95% confidence interval [CI] = 1.0, 8.45) than non-Hispanic Whites to be admitted to a low-quality nursing home. Individuals without a high-school education were more likely (RR = 1.22; 95% CI = 1.0, 1.46) to be discharged to the lowest quartile of nursing homes in a given HSA. Women and older patients were significantly less likely to be discharged to low-quality nursing homes. In a set of unreported robustness checks, we tested for interaction effects between ethnic/racial minority status and other demographic characteristics such as marital status, female gender, and less than high-school diploma and found no significant interactions. We also ran the model with dummy variables for the 5 major regions in the United States and observed no changes in the effects for ethnic/racial minority status and education.

DISCUSSION

Ecological analyses have shown residential racial segregation patterns to be independently associated with racially disparate mortality rates.29^–31 Segregation clearly plays a role in perpetuating racial and ethnic disparities in quality of care, but identifying the mechanisms through which segregation persists has been difficult. Our findings offer evidence that existing ethnic/racial disparities in nursing home care are attributable in part to hospital discharge practices that refer minorities to lower-quality nursing homes. Focusing on the HSA level allowed us to narrow the range in which nursing home “choice” typically occurred, thereby characterizing quality differences in a highly localized manner. Moreover, by focusing on postacute Medicare patients, we were able to control for the role of insurance because all these individuals had Medicare-covered stays.

Not having completed high school was independently related to the likelihood of being discharged to poor-quality nursing homes. Because current policy assumes that elders and their advocates should be able to participate in making quality choices and avail themselves of the relevant information, we assumed that those with less education would have more difficulty finding, interpreting, and acting on this information.32

The Centers for Medicare and Medicaid Services rulings requiring providers to offer patients referred to postacute settings a choice of facilities may be counterproductive in guiding patients to those facilities with the highest quality (the goal of public reporting of quality information). If those discharged from hospitals are making choices solely on the basis of geography (proximity to their families), it may well be that African American and other minority patients “choose” facilities that are systematically of lower quality precisely because of their location in the community.33 Perhaps a more proactive approach will be required to give minority patients a choice based on quality information that could be known to the discharge planner. Indeed, discharge planners may be in the best position to translate publicly available information about nursing home quality to prospective nursing home admissions and their families, particularly because such decisions tend to be made in a crisis situation. This would involve expanding the role for discharge planners to include more than simply finding an empty bed as quickly as possible.34 One mechanism to expand the role for discharged planners could involve collaborative efforts initiated by Programs of All Inclusive Care for the Elderly, whereby the care of frail elders is proactively managed to reduce lengths of stay while facilitating high-quality transitions between settings.

Inspection data documenting nursing home health deficiencies were publicly available for several years before the Centers for Medicare and Medicaid Services online publication of new quality indicators in early 2003.35^,36 The inspection results and new quality measures are now available to all hospital discharge planners to use in guiding nursing home placement decisions. Hospital discharge planners have been included as an audience in the national quality improvement organization efforts to promote higher-quality nursing home care.37 However, the extent to which hospital discharge planners are sharing the new information with patients and their families remains unknown.

Most interventions targeting discharge planners for education and awareness have focused on increasing discharges to home,38 reducing rates of readmissions,39^,40 lowering total days rehospitalized,41 and improving patient satisfaction with and involvement in the discharge process.42 New initiatives to sensitize hospital discharge planners to the quality of nursing home care should build on these education and awareness efforts while recognizing that the choice of a nursing home can be influenced by the ethnic/racial makeup of nursing homes: racial/ethnic minorities may choose to reside with other ethnic/racial minorities.

Our study was limited by the lack of information about individuals’ preferences for specific nursing homes or the relative importance of nursing home geographic proximity to family. Also, we did not include information on the distribution of all hospital discharges receiving other forms of postacute care such as rehabilitation, hospital, or home care. Our findings may be biased if minorities are significantly more or less likely to use these other forms of postacute care. However, because recent research has shown that greater use of Medicare home health services among minorities is largely explained by differences in health needs, it is likely the same would be true in the case of nursing home care.43 We also lacked data on supplemental Medicare insurance for skilled nursing home care, which may influence placement into certain facilities. A final potential limitation is the omission of hospital quality information within our analyses. For example, minority patients may disproportionately be admitted to public hospitals with lower operating margins. Thus, policies that address broader disparities in hospital resources may ultimately lessen disparities in nursing home placements.

In spite of numerous commentaries on postacute care and the need for continuity of care for the frail Medicare population and the increasing recognition of the importance of chronic care systems of care, little attention has been devoted to the relations that hospitals have with the nursing homes to which they discharge their patients.44 This lack of attention has allowed racial disparities to persist in the quality of nursing homes to which hospitals’ minority patients are discharged. Hospital discharge planners’ referring of minority hospital patients to lower-quality nursing homes may contribute to racial and ethnic disparities in quality of care that exist within a geographical area. Educating hospital discharge planners about nursing home quality is a critical first step in leveling the playing field for vulnerable minority patients.