|Home | About | Journals | Submit | Contact Us | Français|
Medication use in nursing homes (NH) occurs under some of the most complex circumstances in all of medicine. Most NH residents are frail with multiple medical conditions typically related to cardiovascular disease, arthritis, stroke and diabetes.1, 2 Nearly half are over the age of 85 and have disabling dementia that impairs their ability to communicate or perform daily activities.3 NH residents also receive, on average, seven to eight medications each month, putting them at high risk of medication-related problems including use of inappropriate therapies.1, 2, 4, 5
NHs may also be one of the most difficult settings in which to improve medication use. The challenge of poor NH prescribing quality is well documented and includes the underuse of indicated medications, overuse of harmful or ineffective medications, and failure to adequately and appropriately monitor narrow therapeutic window medications such as warfarin.1 Prior efforts to improve NH prescribing include academic detailing,6 audit and feedback to physicians,7 Food and Drug Administration (FDA) labeling with ‘black box’ warnings for high risk drugs,8, 9 and federal policy proscribing drug appropriateness for NH residents.10 Despite these efforts, the problem of suboptimal NH prescribing persists. To illustrate this point, we focus on the example of antipsychotic medications.
Antipsychotics are widely used in the NH setting and have become the dominant therapeutic modality in NHs for treatment of behavioral symptoms of dementia despite clinical trial evidence showing little benefit for behavioral management11, 12 and growing evidence of excessive morbidity and mortality.8, 12, 13 In fact, the growth of atypical antipsychotic use between 1999 and 2006,14 despite federal regulations10 and FDA15 warnings calling for greater restraint in antipsychotic use among older adults suggests that the overuse of antipsychotics in NHs represents one of the great failures of evidence–based medicine to date.
This commentary describes a framework for improving prescribing in NHs by focusing on the whole facility as a system that has created a “prescribing culture.” We offer this paradigm as an alternative to targeted interventions that focus on educating and reforming “bad” prescribers, using the example of the atypical antipsychotics to illustrate the approach. We highlight elements of the NH culture change movement that are germane to medication prescribing, and illustrate which elements of NH culture have been shown to be associated with suboptimal quality of care. We conclude by describing current models including our study funded by the Agency for Healthcare Research and Quality (AHRQ) to identify the best methods of disseminating evidence-based medication use guides in NHs.
Approximately one-quarter to one-third of NH residents in the United States currently receive antipsychotic drugs.14, 16 This is the highest level of use reported in more than a decade. This affects the health and welfare of approximately 390,000 frail, institutionalized elders, with wide geographic variation in the United States ranging from approximately 20% to 45%. (Figure 1) In 2006, antipsychotics had become the most costly drug class for the Medicaid program, a main payer of NH medications,17 including $176 million in Medicaid reimbursements for dual eligibles and an additional $2.6 billion for nondual eligibles. By 2007, antipsychotic drugs were the third-largest therapeutic class in U.S. sales for all payers combined, accounting for over $13 billon in sales.18 Overall sales for antipsychotics followed only lipid-lowering agents and proton pump inhibitors, and experienced a growth in prescription sales of 12.1% between 2006 and 2007.
Up to 80% of antipsychotics prescribed in NHs are for off-label uses, mainly for the management of the behavioral symptoms associated with advanced dementia.11,14,16,19 In 2006, a significant proportion of NH residents with dementia were prescribed an antipsychotic, including 22.6% without behavioral symptoms, 29.5% with non-aggressive behavioral symptoms, and 51.2% with aggressive behavioral symptoms.14 Essentially, most use was for residents without an FDA-approved indication, with as many as 21% of all NH residents in the US receiving antipsychotics without a psychosis-related diagnosis.14,16
The evidence base for using antipsychotics in the NH setting is limited and mostly extrapolated from what is known about treatment effects in community-dwelling adults. The landmark National Institutes of Health sponsored randomized clinical trial on atypical antipsychotics in older adults with Alzheimer’s dementia---the CATIE-AD trial---was conducted in the community-setting.12 The CATIE-AD study found that the overall effectiveness of antipsychotics was limited, and outweighed by the risks of adverse effects. In 2005, the FDA issued black-box warnings on the atypical antipsychotics for excess stroke and overall mortality risk in older patients with dementia.15 Lastly, a recent Cochrane review of high quality evidence on using atypical antipsychotics to treat patients with dementia (including NH residents) concluded: 1) only risperidone and olanzapine had sufficient data demonstrating efficacy in treating aggression; and 2) no antipsychotic has sufficient data on improving cognitive function.20 Furthermore, “the atypicals should not be used routinely to treat dementia patients unless there is severe distress or risk of physical harm to those living and working with the patient.”
Antipsychotics are also among the most highly regulated medications in NHs and have been so since the 1980s when the Institute of Medicine reported widespread misuse of these agents to sedate and chemically restrain NH residents.21 Today, NHs must maintain careful documentation on the necessity of using antipsychotics for each resident. This includes the reason for the treatment from a list of approved indications, plans for monitoring medication efficacy and tolerability, and future consideration for dose reduction or treatment discontinuation. Failure to do so may result in fines and sanctions on the NH for deficiencies in care. Unfortunately, despite initial reductions in antipsychotic use immediately after the implementation of this federal-level regulatory intervention,10 antipsychotic prescribing remains essentially unchanged in 2007.22
NHs are complex institutions with a wide variety of organizational models and health care professionals. Medication decisions often represent concerted efforts between the on-site NH staff of nurse practitioners, registered and licensed practical nurses, certified nursing aides, consultant pharmacists and physicians. These physicians are typically off-site and communicate orders by telephone or faxes.23, 24 Social workers often convey family or patient preferences in this process, and consultant pharmacists review the medication order for safety and compliance with regulations. The quality of the medication decision reflects the quality of communication within the NH.25, 26
Organizational culture offers a cohesive way for understanding the many influences on prescribing decisions in NHs.27, 28 (Figure 2) Organizational culture is a broad concept that encompasses the shared values, beliefs and assumptions of a group or members working together as a group.29 In this context, each NH may be understood as a micro-society of internal stakeholders who are subject to external pressures and whose inter-related decisions result in observable patterns of medical care.27, 30 For some NHs, high staff turnover31 and many off-site physicians with limited training in NH care may distinguish the culture, while in other NHs a litigious environment may influence medical decisions.
Some of the best examples of how organizational culture affects health care processes and patient outcomes come from the hospital setting. For example, Shortell et al.32 have shown from data collected from 61 US hospitals that quality improvement implementation in the healthcare setting varied by four types of cultures: group culture, based on norms and values of teamwork; developmental culture, based on risk-taking and change; hierarchical culture, based on values associated with bureaucracy; and rational culture, emphasizing efficiency and achievement. Organizational culture type has also shown associations with differences in approaches to quality improvement implementation [e.g. defensive (focusing primarily on meeting external accreditation requirements) or integrative (called ‘prospector’ in their framework, reflecting integration into an overall plan of implementation)]. These differences have predicted variation in patient outcomes. Hospitals with a patient-centered focus have also demonstrated greater investments in change to improve chronic illness care.33
Current research on NHs organizational culture most closely reflects studies regarding the role of patient-centeredness on healthcare outcomes. Also termed “resident-centered care” and “resident-directed care” in the NH literature,34–36 these studies have found that values such as compassion, dignity and respect to be associated with less use of feeding tubes in NH residents with advanced dementia37 and better performance on quality measures such as the number of NH residents having high-risk pressure ulcers, low-risk pressure ulcers, and being bedfast.38
Specific to medications, evidence from a single-site, multi-pronged intervention suggests that antipsychotic use may be reduced by providing combinations of resident-centered activities, prescribing guidelines, and educational rounds to improve NH dementia care.39–43 Furthermore, medication reviews and/or educational interventions alone44–46 do not significantly improve the prevalence of antipsychotic prescribing in a pooled analysis.47 Other studies suggest that differences in organizational culture may also explain the wide variation in the use of antipsychotics in NHs that are unexplained by patient case-mix,6, 48 organizational (e.g. profit status) or market characteristics.49 Although the specific drivers of off-label antipsychotic prescribing remain unclear, facility-level factors such as staffing levels,10, 50 nurse beliefs,25 interdisciplinary communication,25, 39, 51 and availability of resident activities40 appear to be associated with variations in antipsychotic use.
Using the framework of organizational culture, we are conducting a cluster-randomized trial to improve the antipsychotic use in NHs using a stepwise approach that begins with an assessment of the prescribing culture of each NH. We have recruited 72 NHs in Connecticut with baseline use of atypical antipsychotics ranging from 2% to 65% of all residents in the facility.
The first step in this study characterizes each NH’s prescribing culture. Specifically, we will conduct a mixed-method study with NH leadership to assess important cultural domains known to be associated with variations in psychotropic medication use. A mixed-method study uses both qualitative and quantitative approaches to characterize observed phenomenon.52 Qualitative components include direct observation of facility characteristics using the Observable Indicators of Nursing Home Care Quality Instrument53 and semi-structured interviews with NH leadership. Questionnaires for leadership, direct care staff, nurses, prescribers, and associated consultant pharmacists based on validated NH culture change tools.34 Summarized in a 2006 CMS report Measuring Culture Change, six constructs were identified in the definition of culture change: resident-directed care and activities; home environment; relationships with staff, family, resident and community; staff empowerment; collaborative and decentralized management; and measurement-based continuous quality improvement processes.34 Sample questions based on these domains, and others described in related resident-centered assessments,25, 54 are specified in Table 1, and include assessments of NH structure (e.g. presence of a special care dementia unit and home-like atmosphere), decision-making process (e.g. protocols for the management of residents with behavioral issues), staffing (e.g. consistent assignment and staff turnover), interdisciplinary collaboration (e.g. inclusion of direct care staff, geriatric psychiatry in behavior management plans), and resident-centered communication (e.g. resident activities and family involvement).
The goal for the assessments is twofold. First, these assessments will guide the development of tailored educational and interventional approaches to improve antipsychotic prescribing by incorporating the perspectives of multiple stakeholders, and will depend on tailoring the intervention to the NH’s culture. Potential products for targeted stakeholders are described in Table 2. Second, these assessments will contribute to the development of instruments to characterize a NH’s medication prescribing culture as 'resident-centered’ or more ‘traditional’, validated with actual facility-level antipsychotic use, change in antipsychotic use after the intervention, and direct observation of each NH’s facilities’ structure and process of medication decision-making. We expect to find that some NHs will have hierarchical, traditional cultures with defensive approaches to quality improvement implementation, while others will have group, resident-centered cultures with integration of team-level interventions with physicians educated in partnership with nurses. Other facilities may have strong medical director models where the most effective interventions will require a centralized coordination of efforts through the medical director. Improving prescribing in NHs must incorporate the perspectives of multiple stakeholders and their particular NH’s culture.
Approximately 56 percent of NHs have either adopted culture change or are committed to culture change adoption.38 Emerging research on the role of organizational culture in NHs and its role in quality of care show promise. While different types of NH culture and organizational approaches to care currently exist--such as the Eden Alternative, Evergreen, and Green Houses,55 --further work can elucidate which elements of these models map to important cultural elements that contribute to improved pharmaceutical care. The Veterans Affairs Geriatric and Extended Care Program that oversee the 133 nursing homes (now called community living centers) nationwide have been leaders in implementing culture transformation. In fact they have each of the NH facilities self-report on a website twice yearly their scores in the six categories that make up the Artifact of Culture Change Tool, and are currently working on projects in influence change and determine their effectiveness.56 It remains unclear whether marrying elements of prescribing intervention trials46 with the broader strategy of the NH culture change movement will lead to meaningful reductions in inappropriate antipsychotic use, increases in the implementation of federally-mandated tapers of currently prescribed antipsychotics, and increased resident-centered care. We hypothesize that a combined interventional approach that educates NH leadership and prescribers, addresses organizational behavior, and improves interdisciplinary communication and direct care staff involvement can achieve sustained improvements in prescribing. We need a better understanding of the various ways patient-level prescribing decisions, and facility-level quality improvement decisions, are made in NHs so we can effectively intervene to finally align the use of medications such as antipsychotics with the evidence-base.
Funding Sources: This project was supported by R18 HS019351from the Agency for Healthcare Quality and Research. Dr. Tjia was supported by R21 HS19579 from the Agency for Healthcare Quality and Research. Dr. Briesacher was supported by a Mentored Research Scientist Award (K01AG031836) from the National Institute on Aging.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.