Abstract
While improving patients’ abilities to manage medications is important for home health care patients for a successful recovery, we have a limited understanding on differences in the improvement by neighborhood disadvantage. Thus, we examined whether improvements in oral medication management vary by neighborhood socioeconomic status among Medicare home health care patients. Our findings show the odds of improvement in oral medication management for patients in the most disadvantaged neighborhoods were 1.13 times higher (95% CI, 1.11-1.14; p < .001) than those from less disadvantaged neighborhoods. At the same time, we also find that patients in the most disadvantaged (33.6%) were more likely to be excluded due to inpatient facility transitions, compared to patients in less disadvantaged neighborhoods (26.6%). The excluded sample due to inpatient facility transition was more associated with higher ADL scores, risks of hospitalizations, and prevalence in the majority of the chronic condition indicators compared to the main sample.
Introduction
Quality of care measures are critical tools in the current US health care system as they are used in several initiatives to monitor and track the quality of care, encourage health care providers to improve their care, and support patients and families in making informed decisions.1,2 The Centers for Medicare and Medicaid Services (CMS) also uses quality measures for home health care, which measures the ability of home health agencies (HHAs) to provide high-quality care. 3 To ensure receiving high-quality home health care is equally accessible to all patients and not limited to specific subpopulations, previous studies have examined possible disparities in care quality.4 -8 However, the majority of studies have focused on particular types of measures such as timely initiation of care and/or measures of potentially avoidable events, both which have been highlighted and targeted by several policy initiatives.4 -8 Considering the importance of each care quality measure which represents unique measures/characteristics of care quality, expanding our knowledge on health disparities in care quality measures is important to better understand potentially overlooked gaps in provided care and to identify opportunities to improve overall care quality and health outcomes.
Improvement in management of oral medication is one of the quality measures used by CMS to assess the quality of care provided by HHAs. It measures whether patients’ ability to reliably and safely take oral medications improved at discharge compared to the start or resumption of care. Correctly taking oral medications is a complex issue which is important for successful recovery from a short-term illness, control of a chronic illness, and prevention of adverse reactions and health outcomes such as serious complications and rehospitalization. 9 The importance of oral medication management is emphasized for older adults who have a greater number of chronic conditions and take multiple medications, increasing the likelihood of potential mistakes in medication management.9 -11 While improvements in the ability to correctly take medications are critical for Medicare home health care users to remain in their home and community, medication management has been relatively overlooked, especially in the context of disparities in Medicare home health care.
Neighborhood context could contribute to disparities in such care quality measures. People in more disadvantaged neighborhoods are known to have limited access to high care quality health care settings than those in less disadvantaged neighborhoods.12 -14 Moreover, additional external factors which HHAs may not be able to control for could lead to differences in care quality and health outcome measures. For example, people living in disadvantaged neighborhoods are known to be associated with less social and financial support and resources.15 -18 which affects health recovery and maintenance. Thus, people in disadvantaged neighborhoods could experience more challenges in receiving the desired level of care, which reduces the possibility for patients to achieve improvement in health and functional abilities, further hindering the likelihood of staying at their home and community. Examining the variation in oral medication management improvement by neighborhood disadvantage could guide the assessment of whether all Medicare home health care users are receiving high care quality and also identify opportunities for target interventions and programs to achieve health equity in Medicare home health care.
This study, therefore, examined the differences in oral medication management improvements by neighborhood disadvantage. We hypothesized that disadvantaged neighborhoods will be less associated with improvements in the ability to manage oral medications.
Methods
Data
We used the 2019 Outcome and Assessment Information Set (OASIS) for this study. The OASIS is an extensive assessment tool that collects information for home health care patients including functional, clinical, behavioral, service utilization domains from home care clinicians. 19 We also used and merged the 2019 Area Deprivation Index (ADI) with the OASIS data based on 9-digit zip codes. The ADI is a composite measure of 17 indicators based on Census Block Group-level income, education, employment, and housing conditions from the American Community Survey.20,21 Moreover, we used Medicare Beneficiary Summary Files (MBSF) for information including race/ethnicity, Medicare Advantage enrollment, and chronic conditions. We also used providers of services (POS) to include HHAs’ information.
Sample
We focused on Medicare beneficiaries aged 65 years or older who received home health care. This study limited the sample to Medicare Fee-for-Service beneficiaries due to the differences in the use pattern of HHAs by Medicare program,22,23 and focused on the first care episode per patient. We excluded individuals without identifiable 9-digit zip codes and ADI values. According to the calculation of the quality measure: improvement in management of oral medications by CMS, 24 further exclusion criteria were applied. First, patients who (1) were able to take oral medications correctly without assistance or supervision, (2) has no prescribed oral medications, or (3) were nonresponsive, were excluded. Also, patients transferred to an inpatient care facility or who died were excluded. Our final study sample includes 1,334,600 patients. Supplemental Table S1 provides the sample size by each exclusion criterion.
Variables
Main outcome
The main outcome was improvements in management of oral medication. Management of oral medication indicates “Patient’s current ability to prepare and take all oral medications reliably and safely, including administration of the correct dosage at the appropriate times/intervals. 24 ” Improvements in management of oral medication is a dichotomous measure indicating whether the value reported by the home care provider on the discharge assessment indicates less impairment in the patient’s ability to correctly take oral medications compared to the assessment at the start (or resumption) of care (0 = no improvement in management of oral medication and 1 = improvement in management of oral medication).
Neighborhood Socioeconomic Status
Using the 2019 ADI values, we categorized the sample into 2 groups: “less disadvantaged neighborhoods” and “most disadvantaged neighborhoods.” The ADI is reported in percentiles, in which a higher ADI value indicates a higher level of deprivation. Based on the ADI percentile and previous studies,25,26 sample from an area with an ADI value above the 85th percentile was defined as the most disadvantaged neighborhood and the bottom 85th percentile was defined as less disadvantaged neighborhoods.
Other Covariates
We used the Aday-Andersen health behavior model to select covariates. Predisposing factors included age, gender, race/ethnicity, and living arrangements and enabling factors include Medicaid status. We also included several factors including indicators of chronic conditions, activities of daily living score, and indicators of risk for hospitalizations to capture beneficiaries’ health needs. In addition, we included HHAs’ characteristics including tax status, size, affiliation, and accreditation status. A list of the covariates and categories of covariates is provided in Supplemental Table S1.
Analysis
First, we conducted descriptive analyses to compare characteristics of home health care episodes by neighborhood disadvantage (LDN: less disadvantaged neighborhood vs MDN: most disadvantaged neighborhood). We used chi-square tests for categorical variables and t-tests for continuous variables to test for significant differences. Next, we used multivariate logistic regression models to examine the association between neighborhood SES and improvements in management of oral medication adjusting for individual- and HHA-level characteristics. In addition, we compared the rate of excluded patients due to transition to inpatient care setting by neighborhood disadvantage to explore potential differences by neighborhood related to the medication management improvement measure. Also, we explored the characteristics of patients by the exclusion criteria: transition to an inpatient care facility. We considered p-values of <.05 to be statistically significant in all analyses. We used R version 4.4.1 to perform all statistical analyses.
As a sensitivity analysis, we examined whether the study results vary by HHAs in states with and without demonstration of the Home Health Value-Based Purchasing (HHVBP) program. The HHVBP demonstration program was designed to encourage HHAs to improve their care quality through financial incentives and penalties which was conducted in 9 states from 2016 to 2020. Also, we tested different thresholds of ADI to define neighborhood SES. In addition, we limited our analyses to patients who were discharged from an inpatient care facility within the past 14 days before the first assessment.
Results
The sample included 1,334,600 patients. Table 1 shows the sample characteristics by neighborhood disadvantage. Patients living in the most disadvantaged neighborhoods were less likely to be White (less disadvantaged neighborhoods: 82%; most disadvantaged neighborhoods: 70.4%) but more likely to achieve improvements in oral medication management (less disadvantaged neighborhoods: 75.5%; most disadvantaged neighborhoods: 78.7%), live alone (less disadvantaged neighborhoods: 22.7%; most disadvantaged neighborhoods: 29.2%), and be Medicare-Medicaid duals (less disadvantaged neighborhoods: 18.1%; most disadvantaged neighborhoods: 28.5%) than others. While patients in the most disadvantaged neighborhoods were associated with higher prevalence of chronic conditions including chronic kidney, chronic obstructive pulmonary disease, diabetes, and heart failure than others, they were less likely to have specific chronic conditions including depression and osteoporosis. Also, patients in the most disadvantaged neighborhoods were less likely to receive care from non-profit (less disadvantaged neighborhoods: 34.5%; most disadvantaged neighborhoods: 29.6%) and accredited HHAs (less disadvantaged neighborhoods: 39.1%; most disadvantaged neighborhoods: 31.6%) than patients in less disadvantaged neighborhoods.
Characteristics of Patients by Neighborhood Disadvantage.
Note. LDN = less disadvantaged neighborhood; MDN = most disadvantaged neighborhood; ED = emergency department; BPH = benign prostatic hyperplasia;
We only displayed risks for hospitalization with >20%, discharge location and chronic conditions with >10%.
Table 2 reports the odds ratio (OR) of improvement in management of oral medication using multivariate logistic regression models. In the unadjusted model, we find that the odds of improvement in management of oral medication in the most disadvantaged neighborhood sample were 1.20 times higher than those in the less disadvantaged neighborhood sample (OR = 1.20; 95% Confidence Interval (CI), 1.18-1.21; p < .001). Adjusting for individual-level characteristics (model 1) and adding adjustments for HHA-level characteristics (model 2) narrowed down the differences in the odds of improvement in management of oral medication between the most disadvantaged neighborhood sample and less disadvantaged neighborhood sample, but we find that ORs were 1.16 (95% CI, 1.14-1.17; p < .001) and 1.13 (95% CI, 1.12-1.15; p < 0.001) from model 1 and model 2, respectively. Our sensitivity analyses (Supplemental Appendix Tables 2–5), focusing on different subsamples (patients in HHVBP states and not in HHVBP states, and patients who were discharged from inpatient care facilities within the past 14 days) and using different ADI cutoffs (70th and 80th percentile), show the same pattern of the main results.
Association Between Neighborhood Disadvantage and Improvement in Oral Medication Management using Multivariate Logistic Regression Models.
Note. Reference group: patients in less disadvantaged neighborhoods. CI = Confidence Interval.
Adjusted for individual-level factors including age, gender, race, living status, Medicaid enrollment, ADL score, inpatient facilities discharged within the past 14 days, risks for hospitalizations, cognitive impairment, and chronic condition indicators.
Adjusted for individual- and home health agency-level factors including ownership, affiliation, CMS program enrollment, size, hospice care program, and accreditation status.
Figure 1 displays the percentage of patients excluded due to being transferred to an inpatient facility by neighborhood disadvantage. The figure shows that patients in the most disadvantaged neighborhoods (27.8%) were more likely to be excluded because they were transferred to an inpatient facility compared to patients in less disadvantaged neighborhoods (26.6%).
Percentage of excluded patients due to inpatient facility transition.
Table 3 compares the characteristics of the main sample and the “excluded sample,” indicating patients who were excluded due to being transferred to an inpatient facility during their home health care. The excluded sample was more likely to be Medicare-Medicaid duals (Main sample: 19.6%; Excluded sample: 22.2%) and report higher ADL score (Main sample: 4.09; Excluded sample: 3.78), but less likely to live alone (Main sample: 23.6%; Excluded sample: 21.9%). Also, we find that the excluded sample was associated with a higher risk of hospitalization based on previous history of multiple hospitalizations (Main sample: 20.6%; Excluded sample: 36.5%) and multiple emergency department visits (Main sample: 18.8%; Excluded sample: 31.9%) in the past 6 months. In addition, the excluded sample shows higher prevalence in the majority of the chronic condition indicators, except cataract (Main sample: 21%; Excluded sample: 16.4%) and glaucoma (Main sample: 17.1%; Excluded sample: 16%).
Characteristics of Episodes: Main Sample Versus Excluded Sample due to Inpatient Facility Transition or Death.
Note. ED = emergency department; BPH = benign prostatic hyperplasia;
We only displayed risks for hospitalization with >20%, discharge location and chronic conditions with >10%.
Discussion
We examined the association between neighborhood disadvantage and improvements in oral medication management in Medicare home health care. This study shows that the patients in the most disadvantaged neighborhoods were more likely to be racial/ethnic minorities, living alone, enrolled in Medicaid, and less likely to receive care from non-profit and accredited HHAs. Our main study results based on multivariate logistic regression models are inconsistent with previous studies and our prediction as we found patients in disadvantaged neighborhoods to be more likely to achieve improvements in management of oral medications. However, our additional analyses show that the excluded sample due to transitions to inpatient care facilities was associated with poorer health status and higher rate of transition to inpatient facilities in the most disadvantaged neighborhood sample than others. While it is possible that HHAs serving patients in disadvantaged neighborhoods are more likely to focus on improving patients’ medication management ability, these study results raise the possibility that the differences in quality measures by neighborhood disadvantage may not appropriately capture the gaps in care quality, and the quality measure may not adequately inform patients and families.
The majority of the previous studies have found that individuals in disadvantaged neighborhoods were associated with lower care quality and worse health outcomes compared to others.25,27 -30 Also, in the context of home health care, Medicare patients living in disadvantaged neighborhoods were more likely to use lower care quality HHAs 31 and experience adverse health outcomes.6,32 Low-care quality health care institutions are associated with lower levels of staffing and resources,33,34 and residing in disadvantaged neighborhoods is associated with scarce availabilities of social support and resources,17,18 which are all important factors for better care quality and achieving better health outcomes. Yet, our study results show that patients in the most disadvantaged neighborhoods were more likely to achieve improvements in oral medication management compared to others in less disadvantaged neighborhoods.
Several possible explanations exist for the observed differences in medication management improvements by neighborhood disadvantage. First, HHAs mainly serving patients in disadvantaged neighborhoods might prioritize enhancing patients’ ability to self-administer oral medications. Patients in disadvantaged neighborhoods are more likely to encounter barriers in accessing needed care services and resources.15,29 Also, while older adults receiving home health care services heavily rely on social support including caregivers and community resources, residents of disadvantaged neighborhoods often face greater challenges in finding social support and resources from the community.17,18,35 Under these circumstances, the ability to self-manage medication intake could be particularly important for patients in disadvantaged neighborhoods, enabling them to effectively manage their health conditions and remain at their home and community. Thus, prioritizing the ability to take medications for patients living in disadvantaged neighborhoods could have led to their improvement in oral medication intake relative to patients in less disadvantaged neighborhoods.
Another possibility is that sicker patients in disadvantaged neighborhoods are more likely to be transferred to an institutional care setting during their care. It is hard to consider being transferred to inpatient facilities or experiencing death a better health outcome than not achieving improvements in the ability of oral medication management, especially for home health care patients who prefer to stay at their home and community. However, patients in disadvantaged neighborhoods are more likely to be transferred to an institutional care setting during their home health care. 6 As such, patients who may not achieve improvements in oral medication management through home health care were being transferred to inpatient facilities. In other words, patients with worsened health status and lack of resources and social support were more likely to be transferred to inpatient care facilities whose possibility of achieving improvement in oral medications management through high-quality of care is low. It is also known that better medication management is associated with lower hospitalization risk and healthcare costs.36,37 Our findings also show that a higher number of patients living in the most disadvantaged neighborhoods were excluded from analysis since the calculation of improvements in oral medication management excludes patients who were transferred to inpatient facilities and the excluded sample was sicker with higher risks of hospitalizations and more chronic conditions. In such cases, our findings may imply a different kind of issue in neighborhood disparities in care quality that is not reflected in improvement in management in oral medications. Further research to clarify the effect of the exclusion criteria on differences in medication management improvements is critical to support efforts to ensure high care quality in Medicare home health care is not limited to specific subgroups. It could also deliver evidence on whether further modifications on current care quality monitoring system are needed.
Our study has several limitations. This study focused on Medicare-certified HHAs, thus study findings may have limited generalizability to HHAs overall. Also, our study results may not be generalizable to people enrolled in other types of insurance as we focus on Medicare fee-for-service beneficiaries. In addition, the recent decision to expand the HHVBP nationwide in 2022 could affect the study results. Further studies to track and examine how the expansion of the HHVBP affects quality measures and health disparities are needed.
While our study indicates that Medicare home health care patients in disadvantaged neighborhoods are more likely to experience improvement in oral medication management, our findings also raise the possibility that this finding does not capture the full phenomenon due to the characteristics of the excluded patients. Additional research should be conducted to understand the underlying mechanism of the differences in medication management improvement by neighborhood disadvantage. Such research can deliver helpful information to studies examining health disparities in care quality measures, support efforts to achieve health equity in home health care, as well as contribute to ensuring care quality measures represent the intended aspects of care quality.
Supplemental Material
sj-docx-1-hhc-10.1177_10848223241282338 – Supplemental material for Differences in Oral Medication Management Improvements by Neighborhood Socioeconomic Status in Home Health Care
Supplemental material, sj-docx-1-hhc-10.1177_10848223241282338 for Differences in Oral Medication Management Improvements by Neighborhood Socioeconomic Status in Home Health Care by Daniel Jung and Nam Hyo Kim in Home Health Care Management & Practice
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Owens Institute for Behavioral Research, University of Georgia.
IRB
University of Georgia, Protocol Number: PROJECT00004674.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
