Lessons learned from CMS’s National Partnership to Improve Dementia Care: a thematic synthesis of multiple stakeholder-engaged studies

Abstract

Background:

Antipsychotic prescribing in United States nursing homes (NHs) has decreased since CMS debuted the National Partnership to Improve Dementia Care in Nursing Homes (NP); however, reductions have stalled. To help explain persistent antipsychotic use despite the NP’s reduction efforts, the perspectives of diverse NP stakeholders were qualitatively assessed. This study aimed to re-evaluate these individual perspectives in combined thematic synthesis to discover NP improvement opportunities undetectable in single stakeholder assessments.

Methods:

Thematic synthesis. Through immersive crystallization, original source coding results were organized into related descriptive themes. Similarities and differences were identified, and descriptive themes were regrouped into new, increasingly abstract, analytical themes. This cycle continued until variances were resolved and analytic themes sufficiently described and explained all initial descriptive themes.

Results:

Three analytic themes emerged regarding NP improvement opportunities. The NP’s positive impacts would be augmented by 1) a deeper and expanded appreciation of stakeholder perspectives; 2) more urgent and rapid adaptation to unintended adverse outcomes; and 3) greater recognition of the contextual and environmental factors influencing decisions to prescribe or not prescribe antipsychotic medications. Stakeholder groups described: perspectives they perceived as inadequately considered by the NP; insufficient NP engagement with the stakeholders capable of creating evidenced, affordable, and available non-pharmacologic therapies for dementia symptoms; recognition that dementia interventions effective for a specific individual at a specific time in a specific community may not generalize; and diverse ongoing undesirable outcomes from NP policies that could be mitigated by NP modifications.

Conclusions:

The NP has done much to advance dementia care in NHs. Notwithstanding, these results suggest the NP would only be improved through increasingly comprehensive inclusion of stakeholder perspectives, enhanced incorporation of individual contextual factors, and a more decisive mechanism for ongoing and continual adaptation.

Introduction:

Risky antipsychotic prescribing in United States’ nursing homes (NHs) has decreased since the Center for Medicare & Medicaid Service’s (CMS’s) 2012 debut of the National Partnership to Improve Dementia Care in Nursing Homes (NP). In recent years, however, progress has slowed and even backslid with relapses in hard-earned antipsychotic prescribing reductions. More than one in six NH residents continues to be treated with antipsychotic medications despite known risks, weak evidence for efficacy, and contrary to the NP’s ongoing and purposeful reduction efforts.^1–3 The reasons for this variance remain imperfectly understood.

To help explain the gap between what is desired at a policy level and what is happening in practice, we qualitatively assessed the varied perspectives of a diverse group of NP stakeholders. Stakeholder perspectives, including those of NH residents treated with antipsychotics for dementia symptoms, their family and nursing caregivers, clinicians who prescribe antipsychotics in NHs, and key NH facility personnel, were evaluated through semi-structured qualitative interviews and open-ended surveys.^4–6 While each investigation has already added to what is known about dementia care in NHs, we believe that a re-evaluation of these individual perspectives, grouped in a thematic analysis and surveyed through a fresh lens and with a broader perspective, may reveal new NP improvement opportunities indiscernible by single stakeholder assessments alone.

We set out to explore the persistent mismatch between the NP’s antipsychotic prescribing reduction goals and ongoing antipsychotic prescribing in NHs by applying qualitative thematic synthesis on to our completed evaluations of different NP stakeholders. Thematic synthesis can systematically integrate findings of multiple qualitative studies to identify interpretative constructs and explanations beyond those described by the original primary studies.^{7, 8} We hope that a better understanding of the factors and processes underpinning the gap between antipsychotic policy and antipsychotic practice can offer insight into the barriers and obstacles to ongoing NP improvement efforts.

Methods:

Between 2014 and 2022, we qualitatively assessed the perspectives of multiple stakeholder groups invested in the NP’s initiative to reduce antipsychotic prescribing in NHs. While these original works have been summarized in the Table 1 of the supplement, additional important details are available in published form.^{4–6, 9} All qualitative evaluations iteratively evolved over the course of each investigation to incorporate new incoming data, and every evaluation of individual stakeholder groups also built upon earlier investigations of other stakeholders. Thus, while assessment instruments varied across and between different investigations, the overarching goal for each evaluation was the same: to understand why antipsychotics continue to be prescribed in NHs for dementia symptoms despite growing safety and efficacy concerns and in the face of the NP’s purposeful efforts to curb exactly such prescribing.^4–6

For this focused thematic synthesis, the coding team involved in all the original qualitative work returned to the source interview transcripts, codebooks, and line-by-line coding. Exemplar quotes are provided in Table 2 of the supplement and additional details on a priori codes and specific findings from this original work is publicly available.^{4–6, 9} In a highly iterative process of immersion crystallization, findings were organized into related descriptive themes using the methodology for thematic synthesis of qualitative research in systematic reviews described by Thomas and Harden.⁷ Similarities and differences between descriptive themes were identified, and related descriptive themes were grouped together. Grouped descriptive themes were organized and reorganized into new and increasingly abstract analytical themes. This process of developing, grouping, and organizing themes was first done independently and then together as a group. These thematic extractions were inspected recurrently over multiple meeting cycles where additional or disconfirming evidence was sought and differences resolved through ongoing negotiated consensus. This cyclical process continued until all variances were resolved and the resultant analytic themes sufficiently described and explained all the initial descriptive themes.

Results:

Three analytic themes regarding NP improvement opportunities were conceptualized through our thematic synthesis. These themes organize our results and emphasize opportunities for the NP to advance its intended policy goals regarding the use of antipsychotics.

The potential exists for the National Partnership to further improve outcomes by:

Theme 1. Ever increasing consideration of diverse stakeholder perspectives.

Each stakeholder group, whose perspectives were previously evaluated by our research team, shared viewpoints they perceived as inadequately informing the NP. For example, all stakeholder groups believed the NP efforts to improve dementia care would benefit from greater recognition of the palliative and quality-of-life goals of residents and their families, goals which may define some palliative-oriented antipsychotic prescribing outside NP guidelines as appropriate. Conversely, some stakeholders including clinicians, nursing caregivers, and NH personnel described feeling unfairly singled out as the primary contributors to antipsychotic overutilization and consequently bearing a disproportionate expectation for facilitating antipsychotic reductions. All evaluated stakeholder groups described frustration with the absence of evidenced, affordable, and available alternatives to treat dementia symptoms beyond drugs, suggesting insufficient NP engagement with the stakeholders capable of resolving that deficiency. Payors were repeatedly singled out as an example of such a stakeholder, as was the pharmaceutical industry and dementia care researchers.^4–6

Theme 2. Adapting more rapidly to unintended adverse outcomes.

The engaged stakeholder groups described a diversity of ongoing undesirable outcomes from NP policies, adverse effects that could be reasonably moderated with modest modifications to the initiative. NH clinicians, for example, were frank in acknowledging purposefully altering their approach to documentation and diagnosing to skirt mandatory antipsychotic reporting while also increasing their prescribing of non-superior but unmonitored antipsychotic alternatives even as they prescribed antipsychotics less. These measures were done specifically to reduce their NH’s reported antipsychotic prescribing rate while reliance on risky drugs to treat dementia symptoms continued unchanged.⁵ Similarly, key NH personnel endorsed changing their NH admissions’ process in response to the NP, purposefully avoiding admitting residents likely to require antipsychotic medications.⁴ ‘Cherry-picking’ desirable, and avoiding undesirable, NH admissions leads to increasing disparities in NH quality and orphans sicker individuals most in need of quality NH care.⁴ Finally, some family and nursing caregivers described distress at the inappropriate underutilization of necessary antipsychotics perceived as accompanying laudable reductions in inappropriate antipsychotic use.⁶ These negative outcomes, attributed to the NP by stakeholders, were regarded by stakeholders as mitigatable by reasonable modifications to NP’s policies.^4–6

Theme 3. Greater attention to contextual and environmental factors.

Stakeholder groups recognized that what works for a specific individual at a specific time in a specific community may not work among other people, or at other times, or in other places. Resident attributes, facility characteristics, community factors, geographic variables, and a diversity of other environmental and contextual elements were recurrently highlighted as influencing the likelihood that each unique resident would be prescribed or not prescribed an antipsychotic. Stakeholders endorsed that the efficacy and impact of the NP would only be improved by greater recognition of the importance of these influences.^4–6

Discussion:

The NP was created as a public-private collaboration with the quality improvement principal of ‘broad design and scope’ where all are invited to participate. Early partnership stakeholders included NHs, Federal and State agencies, providers, advocacy groups, state coalitions, partnerships, and communities of practice; and the NP went to considerable lengths to be maximally inclusive of all stakeholders.^{2, 10, 11} Despite these painstaking efforts, the stakeholder groups, most particularly resident families and resident caregivers, described perspectives they perceived as insufficiently incorporated into the NP. All care, including the management of dementia symptoms in nursing homes, is a relational activity, not simply a transactional one. The importance of the individuals directly involved in that relationship, and the motivations they have for serving that relationship, are impossible to over-emphasize. That said, integrating every diverse stakeholder perspective into a domestic initiative is an ideal to be strived for, not a reasonable expectation. Patients and their families are also perhaps the most challenging stakeholders to meaningfully engage in a national effort and the science of engaging and integrating these most critical stakeholder groups into quality improvement is still emerging.^12–14 For example, patients and their families were only more recently added to the National Quality Forum’s expert panel tasked with the consensus building process of endorsing NP quality measures, including “The Percent of Residents Who Received an Antipsychotic Medication (Long Stay.)”¹⁵ It is possible that the NP’s approach to curbing inappropriate antipsychotic use would have reflected a greater consideration of the palliative perspective of some patients, their families, and their caregivers had these stakeholders been more comprehensively involved early in the partnership.

Clearly, the over-reliance on risky drugs to manage dementia symptoms in nursing homes is not a problem that can be easily attributed to a single stakeholder group. Rather, it is a problem all stakeholders share, and in which all need to engage collaboratively to resolve. Potentially, the NP missed opportunities to improve care by focusing on certain stakeholder groups over others. Multiple barriers exist to the optimal application of non-pharmacologic therapies to treat dementia symptoms in NHs which do not exist for drugs. Best care for dementia will never be accomplished until non-drug treatments for dementia symptoms are as affordable, available, and accessible as medications. Increased engagement with multiple stakeholder groups (and payors, the pharmaceutical industry, and dementia researchers were particularly singled out) will be required for this to occur in a more meaningful fashion. The perspectives of these stakeholders could offer invaluable insight into factors influencing the decision to prescribe or not prescribe medications for dementia symptoms in nursing homes, and perhaps even hint at new avenues for advancing NP policies.

The NP’s development incorporated tests of change to detect and remediate harm.¹⁰ However, our results are not unique in still suggesting a diversity of ongoing undesirable policy effects.^{10, 16–20} We view this as unavoidable. In addition, both the NP and the Centers for Medicare & Medicaid Services (CMS) have mechanisms to evaluate for initiative outcomes, including those that are adverse or unintended, and respond to them. As examples, through the NP, CMS recognized that providers were inappropriately using the diagnosis of schizophrenia to artificially improve their antipsychotic quality-measure score and facility rating and debuted a focused NH survey to better understand the problem and guide future actions. CMS identified staffing adequacy as a critical factor influencing antipsychotic prescribing and developed new staffing-based quality measures. CMS detected that NHs were not accurately self-reporting data to the NP and added claims-based quality-measures. CMS screened for NHs with persistent high rates of antipsychotic use and subjected them to more severe enforcement actions and penalties.^{10, 15, 20–22} Notwithstanding the above, our results suggest that stakeholders perceive a capacity for more rapid and decisive adaptation to adverse initiative outcomes as an improvement opportunity. In fact, we believe these results argue that rapid adaptive modification may be the most important quality of all effective and efficient quality improvement.

Finally, the NP would be well served by greater and more direct consideration to context. Since the NP engages stakeholders in a deeply relational environment with inherent biases advantaging and disadvantaging certain social groups, sensitivity to individual idiosyncratic impacts could improve NP outcomes, perhaps most notably, equality of care. In fairness, incorporating contextual factors comes with many challenges and its own potential for unintended adverse consequences. Still, there should be a solution that does not conceal disparities, normalize poor performance, or tolerate inadequate care, but rather recognizes that NHs serving vulnerable populations in communities with fewer resources may need additional assistance rather than penalties to achieve excellence in meeting NP goals. CMS is taking a Rewarding Excellence for Underserved Populations (REUP) approach to promoting equity in other incentive-based programs. This method is intended to award a high level of care to individuals with limited access to excellent care without lowering standards, thereby avoiding the unintended penalties that tend to be associated with delivering services to underserved populations.²³ Applying this approach to the NP may at least be a partial solution to the deficiency identified by NP stakeholders. Of course, as described above, our results would suggest that REUP will need to be highly adaptable in responding to the inevitable unintended adverse consequences resulting from its application in order to reach its full potential.

Limitations:

This was a focused study. Our sample was limited, and only inclusive of evaluations of Virginia stakeholders. It was, however, purposive and had the benefit of a consistent research team and a consistent research question. The exhaustive sampling called for by meta-analysis was not required as the purpose of this study was ‘interpretive explanation and not prediction.’²⁴ While this thematic analysis of multiple earlier assessments did not allow for results to be validated with original stakeholders, the broader approach, seeking consistent patterns across multiple groups and multiple studies, enhances external validity.^{7, 8}

Conclusion:

The NP remains an important initiative that has done much to advance dementia care and outcomes in NHs.^{25, 26} Notwithstanding these undeniable positive impacts, critical stakeholders endorse that there are still important NP stakeholders insufficiently engaged by the NP, that there is the potential for greater consideration of contextual and environmental elements, and that adverse outcomes from the NP continue to exist that could be moderated by a more decisively adaptive process. Thus, we believe the NP would only be improved through increasingly comprehensive inclusion of all stakeholder perspectives, enhanced recognition and incorporation of individual and community factors, and an explicit mechanism for swift and continual ongoing adaptation.