Accountable Care Organizations (ACOs) Could Potentially Improve the Quality of Care in Those Afflicted With Dementia

To the Editor:

Introduction:

Accountable Care Organizations (ACOs) could potentially improve the quality of care of high-need, high-cost persons, including those afflicted with dementia. Little research has examined the relationship between patient characteristics, and ACO organizational characteristics on whether a person with Alzheimer’s Disease and Related Dementias (ADRD) remained attributed to an ACO.

Methods:

To study attribution of persons with dementia in an ACO, we assembled a retrospective cohort study of 2018 Medicare decedents with a nursing home (NH) stay between 91 and 180 days prior to death who were attributed to an ACO in the in 2017. Based on the Minimum Data Set (MDS) assessments obtained during a nursing home (NH) stay, decedents with the following characteristics were selected: at least two activities of daily living (ADL) impairments, a dementia diagnosis, and a Cognitive Function Scale (CFS)¹ score indicative of mild to severe dementia. A multivariable random effects model with clustering was used to examine the association between continued ACO attribution and various decedent and ACO-level characteristics. Decedent-level variables included age, gender, Medicare/Medicaid eligibility, race, functional impairment, CFS score, and whether the NH stay was a post-acute stay. ACO organizational variables included ACO age, size, nurse practitioner (NP) and physician assistant (PA) staffing measured as percent of total number of ACO providers.

Results:

In the year prior to death, 26,136 (19.2%) out of 134,937 decedents with dementia were attributed to an ACO. Overall, 52.3% of dementia patients remained attributed, which ranged across ACOs from 0% to 94.9%. Decedent sex, level of function, cognitive impairment, and Medicare/Medicaid eligibility were not associated with continued ACO attribution (see Fig. 1). A skilled NH stay was associated with a lower likelihood of remaining attributed to an ACO in the year of death (Adjusted Odds Ratio (AOR) 0.82, 95% CI 0.75-0.89). Dementia patients attributed to ACOs with more than 10.9% of all providers comprised of NPs were 1.47 times more likely to remain ACO attributed in the year of death (95% CI 1.1-2.0). PA staffing did not have a similar association.

Fig. 1.

Presents the results of key independent variables’ association with continued attribution to an ACO based on multivariate random effects model with clustering at the level of the ACO dentification number. ACO size was based on quartiles of the number of patients in the year prior to death. Prior hospitalizations indicate the Minimum Data Set assessment was based on skilled nursing home stay. CFS indicates the Cognitive Function Scale.

Comment:

Nursing home care costs are projected to increase in the coming decade, with the highest costs occurring in the last month of life due to terminal hospitalizations. In a prospective cohort study of long-stay NH residents conducted in 2013, ACO attribution was associated with fewer hospitalizations, but not lower total Medicare spending, with 14.6% of residents either switching in or out of an ACO.² We found that in 2018, only about one-half of ACO decedents with dementia remained attributed to an ACO in the year of death. Retrospective attribution rules and organizational practices could result in discontinuity of care, potentially reducing quality of care for this high-need, high-cost population.

After a 2016 rule change, non-physician practitioners have grown in ACOs from 25.5% to 38.7% with much of this growth being PAs.³ ACOs may use non-physician practitioners like NPs to manage complex needs of persons with either subacute or long-term NH stays through more frequent monitoring that improves the quality of care and prevents hospitalizations. In our study, we found that a higher NP staffing ratio was associated with higher continued ACO attribution in the year of death.

It should be noted these results are from 2018, before the 2019 ACO rules aimed to increase the number of ACOs taking downside risk as well as upside potential savings.⁴ This rule may either increase the use of NPs to manage NH stays or have unintended consequences of even more decedents with dementia diagnosis losing their attribution. Future research is needed to monitor the impact of ACO policies on continuity and quality of care for this vulnerable population.