Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC): a cluster randomised controlled trial study protocol

Abstract

Introduction

The annual prevalence of elder mistreatment (EM) in cognitively intact older adults is estimated to be 11%, yet the annual prevalence in older adults with Alzheimer’s disease and related dementias (AD/ADRD) is estimated to be as high as 75%. Associated with a decrease in quality of life and increase in risk of mortality, EM represents a significant public health burden. Home-based primary care (HBPC) providers are uniquely positioned to address the critical need for robust EM screening and reporting, especially among individuals with AD/ADRD. This protocol seeks to adapt the Detection of Elder mistreatment Through Emergency Care Technicians (DETECT) screening tool, previously used by emergency medical technicians, for use by HBPC providers.

Methods and analysis

The protocol consists of two main phases which include four substudies. Substudy 1 uses a qualitative approach to understand the current barriers to clinician identification and reporting of EM in HBPC, including what adaptations need to be made to DETECT for use in HBPC. Substudy 2 is a cluster randomised controlled trial evaluating the impact of Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC) on clinician identification of older adult patients with increased risk of EM and referring their concerns to the appropriate authorities and service providers. Substudies 3 and 4 apply a mixed-methods approach to postscreening interviews with clinicians and caregiver/care recipient dyads, respectively. These substudies aim to evaluate DETECT-RPC’s impact on barriers to EM identification and reporting as well as the harms and benefits of using the screening tool from the perspective of patients and their caregivers.

Ethics and dissemination

All components of this study are conducted with the approval of the Institutional Review Board of the University of Texas Health Science Center at Houston (HSC-SPH-22-0732, HSC-SPH-23-0105, HSC-SPH-23-0965, HSC-SPH-24-0123). The results of this study will be disseminated through a peer-reviewed journal as well as through presentations at professional conferences, invited talks and other standard channels.

Trial registration number

NCT05958654 (ClinicalTrials.gov).

Correspondence to Dr Nicholas Sevey; [email protected]

STRENGTHS AND LIMITATIONS OF THIS STUDY

• This study will sample 43 home-based primary care (HBPC) clinicians, 180 caregiving dyads and 6150 patients receiving HBPC services from seven academic HBPC clinics located in four geographically diverse states.

• Embedding the Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care screening tool within electronic medical record systems enhances integration into clinical workflows.

• The use of both quantitative and qualitative analyses allows for a thorough evaluation of the screening tool.

• Randomisation at the provider level could lead to contamination due to interactions between providers.

• The study includes measures to minimise selection and reporting biases through randomisation and defined intervention effects.

Introduction

Elder mistreatment (EM) is commonly defined as an intentional act, or failure to act, by a caregiver or another person in a relationship involving an expectation of trust that causes harm or creates a risk of harm to an older adult.^{1 2} EM takes many forms, including financial abuse/exploitation, neglect and physical, emotional/psychological and sexual abuse. Exposure is often chronic and involves polyvictimisation.^1–3 The prior two decades have seen a sharp uptick in clinical and empirical attention to the prevalence, public health impact, identification and modification of salient risk factors, as well as strategies for early detection, outreach and intervention efforts to address EM.^4–7 Epidemiological studies estimate that more than 1 in 10 cognitively intact older adults living in the community experience EM annually.⁸ Risk for mistreatment is higher among older adults with Alzheimer’s disease and related dementias (AD/ADRD), with estimates suggesting an annual prevalence as high as 30–75%.^9–11 The public health impact of EM is considerable, with EM increasing the risk for poor quality of life,¹² loss of property and security,¹² physical injury,¹³ depression,¹⁴ functional decline,^{15 16} emergency room visits,¹⁷ hospital admissions¹⁸ and all-cause mortality compared with non-mistreated older adults.^19–22 The costs associated with lost income, recovery from financial abuse and the medical, legal and social service intervention needed by mistreated older adults are estimated to be in the billions of dollars each year.² Nevertheless, EM is difficult to detect and often goes unrecognised. Effective and efficient EM screening tools are urgently needed to improve early detection efforts to preserve both family functioning and the health and wellness of older adults and their caregivers.

Additionally, the development of early detection strategies that are appropriate for use with older adults who are living with AD/ADRD is critical as none currently exist.^{23 24} In the absence of such strategies, between one-third and three-fourths of older adults living with AD/ADRD will experience EM this year—3–5 times as many as their cognitively intact peers.^{9 10} Primary care clinicians represent a key cadre of first-line medical professionals who may be uniquely positioned to identify early risk factors for EM. Older adults are seen by their primary care physician in the outpatient office setting 2–3 times a year, on average, and these encounters may be the only interactions that some older adults have with someone other than a family member.²⁵ Accordingly, the American Medical Association and others suggest that physicians have an ethical and professional obligation to routinely inquire about possible EM. Yet, they provide little guidance on exactly how to conduct those inquiries.^26–28 The evidence base for appropriate EM screening strategies in primary care settings is currently insufficient as evidenced by the US Preventive Services Task Force’s (USPSTF) 2018 recommendation statement.²⁹ Therefore, there is need to rigorously evaluate the potential benefits and harms of EM screening in primary care. Among primary care providers, those providing home-based care are particularly well-positioned to identify and respond to EM. In contrast to office-based visits, home-based primary care (HBPC) providers may see their patients more frequently, often 4–6 times per week or more if needed.³⁰ During these visits, providers are able to examine their patient’s environment, observe interactions between the patient and caregivers, and recognise important changes over time. Patients who receive HBPC are typically patients with complex medical needs such as functional impairment, multiple chronic conditions and isolation, which are also known risk factors for EM.³¹

We previously partnered with MedStar Mobile Healthcare and Texas Adult Protective Services (APS) to develop and test an EM screening tool designed specifically for use by Emergency Medical Technicians (EMT) and paramedics in older adults’ homes. The project and screening tool is called Detection of Elder mistreatment Through Emergency Care Technicians (DETECT). The DETECT tool was designed to (1) be brief, (2) be based on the clinician’s direct observations of the older adult and the older adult’s physical and social environment with no direct questioning of the older adult about EM, (3) provide reporting guidance and (4) be integrated into their existing procedures and medical charting software.³² The results from a 4-year test of DETECT indicated that the number of reports to APS increased threefold (Relative Risk, 3.03; 95% CI 2.06 to 4.46) when the DETECT tool was used without any decrease in the proportion of those reports being validated by APS.³³ Further, a subset of the DETECT screening items were adapted for an emergency department EM screening protocol and implemented in five diverse emergency departments that screened over 15 000 patients,³⁴ demonstrating the feasibility of adapting DETECT to diverse practice settings.

Study objectives

Guided by the Abuse Intervention Model (AIM)³⁵ and the Integrated Behavioral Model (IBM),³⁶ the overarching objective of this study is to adapt the DETECT screening tool for use by HBPC clinicians, with special consideration given to screening older adults living with AD/ADRD. This study will consist of two phases (figure 1), each with its own set of aims. In phase I of the study, we will adapt the existing DETECT tool for a new environment—HBPC. The specific aims of phase I are:

• Evaluate the potential impact of state-specific laws and site-specific policies on the reporting guidance that will be provided by Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC) tool.

• Evaluate provider-level barriers to recognising and reporting EM in HBPC environments.

• Adapt the DETECT screening tool for HBPC, including accompanying training, procedures and reporting guidance.

Figure 1. Phase I of the study adapts the existing Detection of Elder mistreatment Through Emergency Care Technicians (DETECT) tool for home-based primary care clinicians (HBPC). Phase II of the study uses a combination of approaches, including a multisite randomised controlled trial to investigate the feasibility, acceptability, harms and benefits of using Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC) in HBPC.

In phase II of the study, we will use a combination of approaches, including a provider randomised controlled trial, to investigate the feasibility, acceptability, harms and benefits of using DETECT-RPC in HBPC. The specific aims of phase II are:

• Rigorously evaluate the impact of DETECT-RPC on clinician identification and referral of older adult patients with increased risk of EM to the appropriate authorities and services.

• Rigorously evaluate the effect of DETECT-RPC on targeted mechanisms of action (ie, current barriers to identification and reporting) elucidated in phase I of the study.

• Rigorously evaluate the benefits and harms of screening with DETECT-RPC.

Importantly, this study does not aim to position DETECT-RPC as a diagnostic tool for EM, nor will the tool be used in this study as a psychometric scale meant to measure a latent construct of EM. Instead, each item of DETECT-RPC represents a specific cue to action intended to increase the likelihood of a report in the context of each clinician’s existing knowledge of the patient as well as the DETECT-RPC educational training component.

Methods and analysis

Study overview

The study has two phases that include four distinct sampling and analysis substudies spread across 60 months of data collection between July 2022 and June 2027 (figure 1). The phase/sample combinations, which we will refer to as substudies 1–4, include

Substudy 1: screening tool adaptation with n=12 HBPC clinicians (project phase I)

The primary objective of this substudy is to understand the current barriers to clinician identification and reporting of EM in HBPC, and what adaptations need to be made to DETECT to transport it from its current use in emergency 911 responses to use in HBPC.

Substudy 2: cluster randomised controlled trial with n=6150 HBPC patients clustered within 43 HBPC clinicians (project phase II)

The primary objective of this substudy is to evaluate the impact of DETECT-RPC on clinician identification of older adult patients with increased risk of EM and referring their concerns to the appropriate authorities and service providers.

Substudy 3: mixed-methods clinician follow-up interviews with n=43 HBPC clinicians (project phase II)

The primary objective of this substudy is to evaluate the impact of DETECT-RPC on the barriers to EM identification and reporting that were identified in the phase I—screening tool adaptation sub-study.

Substudy 4: mixed-methods caregiving dyad follow-up interviews with n=180 caregiver/care recipient dyads (project phase II)

The primary objective of this substudy is to explore the harms and benefits of using DETECT-RPC in HBPC from the perspective of patients and caregivers.

Study setting

We will partner with HBPC clinics providing care across five cities in the USA (Baltimore, MD; Birmingham, AL; Dallas, TX; Houston, TX; San Francisco, CA). We will sample clinicians (geriatricians, nurse practitioners and palliative care physicians) at each site for substudy 1 (phase I—screening tool adaptation), substudy 2 (phase II—cluster randomised controlled trial) and substudy 3 (phase II—mixed-methods clinician follow-up interviews) from our partner HBPC clinics. All patients and patient caregivers participating in substudy 4 (phase II—mixed-methods caregiver dyad follow-up interviews) will also be sampled from our partner HBPC. Because the remaining methods and study design elements (ie, recruitment, sample size, outcomes, etc) differ across substudies, we describe each substudy separately below.

Substudy 1: screening tool adaptation

Evaluate laws and policies

Currently, each state in the USA has its own unique laws defining EM and establishing who is responsible for reporting and investigating EM. In addition to differences in state reporting requirements, each of our study sites has unique organisation-level policies and procedures for identifying (eg, other screening tools currently being used), reporting and offering services to patients who are at risk of EM. Accordingly, the first task completed as part of phase I will be to rigorously document each of the various state and organisation-level EM screening and reporting guidelines. This information will be gathered by the study team via online legislative review and interviews with key informants (eg, clinicians, hospital administrators, APS administrators and state legal experts). After gathering all relevant information, we will develop an EM reporting guidance to be shared with our clinical partners during dedicated focus groups.

Focus groups

We will enrol approximately n=43 HBPC clinicians from across all of our study sites.

These focus groups consisting of HBPC clinicians will last 90 min and seek to achieve the following key objectives.

• Describe their previous experiences when EM was identified.

• Assess individual clinician motivations, the extent to which motivations are aligned and changes in motivation that may impact EM detection.

• Describe current barriers to identifying EM and making referrals for services.

• Describe how current structure, workflows, policies and referral processes facilitate or impede the adoption of EM screening tools and the identification and referral of EM in HBPC environments.

• Collect feedback on DETECT content and process to inform specific modifications for DETECT-RPC

• Describe the impact of ADRD on clinical judgement of the occurrence and reporting of EM.

Focus group participant recruitment

The clinical director and/or clinical champion at each site will have primary responsibility for distributing information about the study to clinical providers to participate in the focus groups. They will be told their participation is not mandatory and will not affect their employment, salary or promotion. Participants will receive $60 Amazon gift cards as incentives to participate.

Focus group participant inclusion criteria

Focus group participants must be a physician, nurse practitioner or physician assistant who actively provides HBPC to patients enrolled in one of our partner HBPC programmes at least part-time.

Adaptation of the DETECT screening tool for HBPC

Following the development of an EM reporting algorithm, a team of experts in EM, primary care and ADRD will meet to tailor the DETECT screening tool for use in HBPC, thus adapting the new DETECT-RPC screening tool. Adaptations may include the decision to retain, discard or modify existing screen items and modifications to screening and reporting protocols. The first complete draft will then be shared to a sample of HBPC clinicians not included in the study for comment as well as a sample of caregiver/recipient dyads. Data from these comments will be used to revise the tool prior to finalising the DETECT-RPC version for use in phase II.

Data collection and management

All focus groups will be audio-recorded and professionally transcribed in preparation for a four-step approach to analysis. Basic demographic information (eg, gender, age, race, marital status, education, income, employment status and caregiving status) about the focus group participants will also be collected. All non-digital data (paper forms, notes, etc) will be destroyed immediately after completing the transcription process. The digital recordings, transcripts and data sets will be stored in an encrypted, password-protected, cloud-based storage solution.

Planned analyses

First, the team will collectively read the transcripts collected from each focus group to develop a deeper understanding of the general themes. Through this process, a deductive codebook will be created to allow labelling of the text from the focus groups. These codes will be generated in group analysis sessions until the team has reached an agreement on all codes to be used. Once the team codes the text, an immersion-crystallisation approach³⁷ will be deployed to inductively identify themes from emerging categories. Transcripts will then be read by a second coder, and coding inconsistencies will be discussed and resolved by consensus between the two coders.

Intervention mapping

After completing the focus groups, we will complete an intervention mapping process (figure 2).³⁸ The result of that process will be a model that maps (1) HBPC clinician barriers to identifying EM and referring patients and caregivers for services to (2) theory- and evidence-based determinants of behaviour change and finally to (3) the specific DETECT-RPC intervention components described below. Taken together, these components may be thought of as the mechanisms of action through which the DETECT-RPC intervention modifies HBPC clinician screening and referral behaviour. The intervention map will also contain a set of measures that will be used to evaluate the effectiveness of the DETECT-RPC intervention in reducing or eliminating these barriers in phase II of the study.

Figure 2. Inputs into the intervention mapping process include the study team’s previous experience with designing, implementing and evaluating the Detection of Elder mistreatment Through Emergency Care Technicians (DETECT) screening tool, the focus groups and key informant interviews described above, a review of the literature, and law and policy evaluation. Successive steps in the intervention map are guided by the Integrated Behavioral Model (IBM) and Abuse Intervention Model (AIM). Continuous feedback collection from study site champions underlies every step of the process.

Figure 2 provides an overview of the DETECT-RPC intervention mapping process. The study team’s previous experience with designing, implementing and evaluating the DETECT screening tool, the focus groups and key informant interviews described above, a review of the literature and the law and policy evaluation described above will serve as inputs into the intervention mapping process. The continuous collection of feedback from study site champions will underlie every step of the process.

In step 2 of the intervention mapping process, we will create a logic model of change from the process inputs and the logic model of the problem created in step 1 (figure 2). The goal of step 2 will be to identify expected behavioural outcomes, performance objectives and programme outcomes. Importantly, the logic model of change will be guided by the IBM, which is an extension of the Theory of Reasoned Action/Theory of Planned Behavior.³⁶

In step 3 of the intervention mapping process, we will design/redesign DETECT-RPC’s core features, sequence and the choice of specific theory- and evidence-based programme components used to promote behaviour change. Inputs into step 3 will include the logic models and matrices developed in steps 1 and 2 and the AIM.³⁵

Substudy 2: cluster randomised controlled trial

While phase I will focus on developing the DETECT-RPC screening tool, phase II will focus on investigating the feasibility, acceptability, harms and benefits of using DETECT-RPC in HBPC (figure 3). This second substudy of phase II will be a provider randomised controlled trial designed to address the first study aim of this phase—rigorously evaluate the impact of DETECT-RPC on clinician identification and referral of older adult patients with increased risk of EM to the appropriate authorities and services. The DETECT-RPC tool will be incorporated into the electronic charting software (electronic medical record (EMR)) that our partner clinicians are already using. Sites without DETECT-RPC EMR integration will capture the data via REDCap.

Figure 3. A provider randomised controlled trial will be used to rigorously evaluate the impact of Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC) on clinician identification and referral of older adult patients with increased risk of elder mistreatment (EM) to the appropriate authorities and services.

Eligibility criteria

At the beginning of phase II, we will randomise approximately n=43 HBPC clinicians to either use the DETECT-RPC screening tool at every qualified patient encounter (experimental condition) or continue to provide standard of care (control condition). Criteria for a qualified HBPC patient encounter include (1) the patient is enrolled in the site-specific HBPC programme and (2) and the encounter is taking place at the patient’s primary residence.

Blinding

A graduate research assistant working on the study team will generate the randomised list of HBPC clinician study group assignments. The study principal investigator, data analysts, and patients will be blinded to the study group assignments. It is not possible to blind the HBPC clinicians in the experimental group because they have to use the DETECT-RPC screening tool. However, HBPC clinicians in the control group are blinded to the existence of the study.

Intervention

The complete DETECT-RPC intervention will be composed of three high-level components: (1) an HBPC clinician education and training component, (2) an indicators of EM screening module built into the EMR and (3) a reporting guidance module built into the EMR.

The education and training component of DETECT-RPC will be delivered to all HBPC clinicians who are randomised to the experimental condition at the beginning of phase II. The training and education session will be delivered online. Topics covered in the training and education session will include the purpose of the study, instructions for using the DETECT-RPC screening tool and other theory-based determinants of behavioural change identified during the intervention mapping process described above. HBPC clinicians will receive semi-annual booster sessions in addition to the initial training and education session.

At each qualified encounter, HBPC clinicians randomised to the experimental group will complete the indicators of EM module (see figure 4). The indicators of EM screening module will be a checklist of screening items adapted from the DETECT screening tool. This module is a tool that will help HBPC clinicians determine if the patient is likely living with EM. The DETECT-RPC indicators of EM module do not require direct input from the patient; rather, it is a purely observation-based (eg, ‘absence of necessities (eg, inadequate food in cabinets and/or fridge, utilities not functioning)’) and typically takes less than 2 min to complete.

Figure 4. HBPC clinicians randomised to the experimental group will complete the indicators of EM module, a checklist of screening items adapted from the Detection of Elder mistreatment Through Emergency Care Technicians (DETECT) screening tool. This module will help HBPC clinicians determine if the patient is likely living with EM.

After completing the DETECT-RPC indicators of abuse module, clinicians who have determined that the patient is living with potential EM will be given access to a reporting guidance module in the EMR. This module will give the clinician applicable laws, policies and procedures for reporting concerns about potential EM to the appropriate state-mandated authorities (eg, contact APS) and other referral resources for patients and caregivers.

Outcomes

The primary outcome of interest for substudy 2 is the number of reports of potential EM made to the appropriate authorities. To observe preintervention baseline EM reporting patterns for each partner site and to establish an ongoing process to collect these data, site champions at each of our partner HBPC sites will record some basic information about any reports of potential EM made by any of the clinic’s clinicians into a REDCap database for 24 months preceding DETECT-RPC implementation. Specifically, they will be asked to record the name of the clinician who made the report, the organisation the clinician is affiliated with, the date of the report, the method used to submit the report and the report number (if provided by the authorities). These data will serve as our primary baseline measure of reports of potential EM made to the authorities. When a report is made to the authorities in phase II, clinicians will continue to inform the site champion using the same protocol used throughout phase I. At the end of each study year, we will extract the number of reports made from the EMR at each study site and compare the percentage of medical encounters that resulted in an EM report by study condition using the methods described below. Secondarily, we will also evaluate differences in mortality by study condition.

Planned analyses

The proposed multisite provider-randomised controlled trial consists of clustered data with patients nested within providers. Providers are further nested within sites. Any statistical analysis that ignores the lack of independence due to nesting will be subject to inflated type 1 errors. We briefly describe our general strategy for specifying the planned models while accounting for non-independence for the proposed analyses below.

Generalised linear mixed models will be our primary statistical approach for analysis. With this cluster randomised design, it can be assumed that observations are similar in terms of their properties representing exchangeable observations. Therefore, it is appropriate to use a compound symmetric covariance approach for the modelling of patients nested within providers as it assumes that any two observations within the same cluster have the same covariance and all observations have the same variance. Provider differences by site will be accounted for by including site as a fixed effect in all models. Implementing our planned covariance model can be done for both continuous and binary screening outcomes in PROC GLIMMIX in SAS by specifying a random intercept for provider and a fixed effect for site. In addition to efficiently accounting for non-independence, a generalised linear mixed model approach provides the flexibility to specify different link and variance functions for our mix of continuous and binary outcomes variables.

To estimate the change in screening, mortality and summary scales measures due to the adapted DETECT-RPC screening intervention, we will analyse models of the following form: g(y _ips) = β₀ + β₁Trt_p + X _i + Z _s, where g() represents the appropriate link function for outcome y, Trt is a fixed effect for treatment assignment, X _i is a set of individual-level covariates chosen to improve model precision and Z _s is a series of site fixed effects. For binary outcomes, we will estimate the risk ratio using a log-binomial model. For continuous outcomes, we will estimate standardised mean differences from the standard linear model. The parameter β₁ estimates the average effect of DETECT-RPC based on within-site variability. While our primary analysis assumes no heterogeneity of the treatment effect by site, we explore site-specific effects by including treatment by site interactions.

We estimate power for each of our aims by first converting the expected sample size into an effective sample size (ESS) accounting for within-provider clustering. The expected number of providers available for randomisation is fixed at 43, and based on initial data we expect approximately 143 patients per provider. We assume a conservative within-provider intraclass correlation of 0.05, resulting in an ESS of approximately 1984. We then estimate a detectable effect size based on a χ² statistic assuming a power of 0.80 and a type 1 error rate of 0.05. This power analysis is likely conservative as planned analyses will incorporate prognostic covariates which increase precision. Initial data from study sites indicate that approximately 5% of older adults are currently being reported to APS. We are powered to detect a risk ratio of 1.62 or an increase in reporting to 8%. Given that we are powered to detect an increase that falls below published rates of EM, we are well powered to achieve study aims.⁸

Substudy 3: mixed-methods clinician follow-up interviews

In this third substudy of phase II, we will address the second study aim of phase II—rigorously evaluate the effect of DETECT-RPC on targeted mechanisms of action (ie, current barriers to identification and reporting) elucidated in phase I of the study—using a concurrent mixed-methods triangulation approach. We described our plan for collecting information about HBPC clinician barriers and facilitators to identifying EM and referring for services above in the section titled ‘Substudy 1: phase I—screening tool adaptation’. Substudy 3 builds directly on that substudy by following up with HBPC clinicians to investigate if the DETECT-RPC intervention is adequately reducing or eliminating the barriers to HBPC clinician screening and referrals for services that were previously identified.

Evaluating determinants of clinician behaviour change

As part of the intervention mapping process described above, we will map (1) HBPC clinician barriers to identifying EM and referring patients and caregivers for services to (2) theory- and evidence-based determinants of behaviour change and finally to (3) the specific DETECT-RPC intervention components. In addition, we will create 5- or 7-point scale measures to evaluate changes in the behavioural determinants after clinicians receive each of the DETECT-RPC intervention components. Clinicians will complete a brief survey containing these scale measures in conjunction with the training sessions described above.

Follow-up interviews with HBPC clinicians

We will conduct brief quarterly interviews with HBPC clinicians inquiring about their subjective experience with the DETECT-RPC intervention. We will additionally conduct more formal investigations of these mechanisms through the administration of the theory- and evidence-based measures generated from the intervention mapping process and through clinician focus groups (n=43) virtually identical to those conducted during phase I.

Planned analyses

We will use a concurrent mixed-methods triangulation approach to evaluate the effect of DETECT-RPC on determinants of clinician behaviour change. The qualitative data will be analysed concurrently with quantitative data obtained from the scale measures of the determinants of clinician behaviour change. We will then triangulate data to evaluate the impacts of DETECT-RPC on determinants of clinician behaviour change as outlined by the intervention map.

Substudy 4: mixed-methods caregiving dyad follow-up interviews

In this fourth substudy of phase II, we will address the third study aim of phase II—rigorously evaluate the benefits and harms of screening with DETECT-RPC—using a mixed-methods approach. We will recruit a purposive sample of n=180 caregiving dyads (figure 5) consisting of family caregivers and their care recipients, half of which will be living with ADRD. We will recruit dyads because we are interested in caregiver behaviours and their relationship to care recipient outcomes. The caregiving dyads will be recruited from among patients who are actively enrolled in one of our site-specific HBPC programmes. Dyads will be recruited within approximately 30 days of the older adult patient’s primary care visit by a specially trained interviewer (eg, licensed clinical social workers) employed by each of our partner clinical sites.

Figure 5. Half of the caregiver dyads will be living with Alzheimer’s disease or related dementias (ADRD) . The fourth substudy of phase II will rigorously evaluate the benefits and harms of screening with Detection of Elder Abuse Through Emergency Care Technicians Screening Tool Revision for Home-Based Primary Care (DETECT-RPC) using a concurrent mixed-methods approach highlighting differences in elder mistreatment (EM) detection among dyads living with ADRD.

Data collection methods and capacity assessment

Prior to data collection during these interviews, the interviewer will assess the patient’s capacity for consent. During this initial discussion, the interviewer will ask explicit and targeted questions designed to determine the capacity to consent to research such as rephrasing initial questions, stating the risks and benefits of participation and non-participation, and eliciting orientation to location, time, etc.³⁹ Once the capacity to consent to research has been determined, the interviewer will immediately administer the Montreal Cognitive Assessment (MoCA) adapted for phone-based administration. Should the MoCA suggest significant cognitive impairment or dementia, the patient’s legal guardian will be requested to provide informed consent. If proxy consent is warranted, assent will be obtained from the care recipient. Further, the interviewer will inform participants of all applicable mandated reporting laws prior to obtaining informed consent (see online supplemental materials). Once informed consent is obtained, the interviewer will spend approximately 1 hour collecting information across nine key domains (table 1). Interviews will be conducted using an interview guide developed by study team members with input from experts in trauma-informed interviewing and care. All interviewers will be trained in how to respond to trauma disclosure, and clear protocols will be put in place for connecting older adults and caregivers with additional services as needed.

Table 1

Caregiver dyad follow-up interview domains

Domain	Measure(s)
Sociodemographic information (CR/CG)	Household size, marital status, age, sex as a biological variable, gender identity, ethnicity, race, sexual orientation, educational attainment, household income and military service history
Social isolation/support (CR/CG)	PROMIS SF Social Isolation Scale,45 PROMIS SF Social Support Scale45
General health	Health history/medical records extracted from EMR (CR), medication use extracted from EMR (CR), PROMIS global physical health short form (CR/CG),45 geriatric depression scale46 or PhQ-947 as appropriate (CR/CG), Katz Activities of Daily Living index and the Lawton Instrumental Activities of Daily Living index (CR/CG)48 49
Alcohol use/misuse (CR/CG)	Alcohol Use Disorders Identification Test50
Self-reported measures of current and past EM (CR/CG)	Conflict Tactics Scale-Revised,51–53 National Elder Mistreatment Study 54
Adverse childhood experiences (CR/CG)	Revised Inventory of Adverse Childhood Experiences55
Caregiver desire to institutionalise	Desire to institutionalise scale56 57
Care-related financial strain	Financial subscale of the caregiver reaction assessment58
Caregiver coping strategies	Coping strategy indicator59

CG, caregiver; CR, care recipient; EM, elder mistreatment; EMR, electronic medical record.

Planned analyses

We will use a concurrent mixed-methods triangulation design⁴⁰ (survey instruments and semistructured interviews) to investigate the potential benefits and harms (to CR and CG) of using DETECT-RPC in HBPC. We hypothesise that screening, and subsequent intervention by the appropriate authorities, will act as a secondary prevention (ie, identify and treat EM before it results in significant symptoms) and subsequently reduce morbidity and mortality. Further, simply screening for EM (ie, raising the clinicians’ awareness of the patient’s EM risk) may alter the clinicians’ behaviour and/or treatment plan in a beneficial way. However, it is also possible that screening, and subsequent intervention by the appropriate authorities, will have unintended harmful impacts. For example, patients may forgo future HBPC encounters to avoid an APS report if one was made after a previous encounter. To assess potential harms and benefits from the screening, and the sequelae following the screening, we will use quantitative methods to objectively evaluate hypothesised outcomes and qualitative methods to understand the magnitude of the outcomes and identify unforeseen outcomes. Using the USPSTF ‘chain of evidence’ framework as a guide,⁴¹ we will triangulate the quantitative and qualitative data to provide an assessment of the estimated net benefits of the DETECT-RPC screening intervention.

Patient and public involvement

In the current protocol, feedback from healthcare providers, patients and caregivers is obtained throughout the study design through focus groups and interviews. This feedback will directly be used to guide the revision of the DETECT EM screening tool for HBPC.

Discussion

The significance of EM is profound, impacting the health, well-being and finances of more than 1 in 10 community-dwelling, cognitively intact older adults annually, with even higher rates among those with ADRD.^8–11 As such, effective and efficient EM screening tools are urgently needed to improve early detection efforts, particularly for those in unique positions to observe EM indicators, like healthcare professionals in home-based settings. The DETECT-RPC study is a significant step toward addressing EM detection challenges. By integrating the evidence-based DETECT tool into routine healthcare assessments in HBPC, the current study aims to improve the early detection and reporting of EM, leveraging the healthcare encounters that many older adults, particularly those who are isolated, have as their sole regular interaction outside familial circles.

The DETECT-RPC study stands out as the first of its kind to adapt and rigorously assess an EM screening tool specifically for HBPC clinicians, focusing on systematic observations of older adults and their living environments. This approach not only aligns with prior research but also capitalises on a novel setting for EM screening. Our methodology is expansive, targeting over 2000 community-dwelling older adults annually across five diverse metropolitan areas in the USA, enhancing the study’s power and potential for significant impact. Additionally, the DETECT-RPC trial is pioneering in its use of multiple data sources, including healthcare and social service records alongside direct feedback from older adults and their caregivers, employing a mix of quantitative and qualitative data to comprehensively assess the effectiveness of EM screening. Such a robust methodological framework holds the potential of significantly advancing our understanding and potentially informing future US Preventative Task Force Guidelines.

If the study is successful, the next steps will include a stage IV multisite cluster randomised (at the clinic level) effectiveness trial. Such a trial would allow us to refine DETECT-RPC further, overcome any lingering doubts about the effects of contamination and investigate methods for widespread implementation of DETECT-RPC. In 2016, 5.4 million house calls were made, and the popularity of HBPC appears to be growing, with the predominant group consisting of elderly patients with complex, chronic conditions and functional impairment who are significantly home-limited.⁴² Fortunately, many HBPC programmes are increasingly funded by various payment structures such as fee-for-service, monthly fee-per-member/per-month and value-based or at-risk contracts housed within Medicare.^{42 43} Effective and efficient EM screening tools that are easy for HBPC clinicians to use could dramatically increase sentinel surveillance of EM in a very short period of time. Therefore, the successful completion of this project has the potential to make a significant, immediate public health impact.

Ethics and dissemination

The DETECT-RPC study will collect, analyse and interpret essential data from screening tools and interviews with older adults and their caregivers. We will also collect and analyse critical data from primary care clinicians, state health agencies and even from the very homes (environmental cue data) of older adults. This novel set of approaches will obtain data on and about human subjects, and every precaution will be taken to ensure all participants’ dignity, autonomy and confidentiality.

When conducting our dyad interviews, informed consent will be obtained prior to administering the MoCA. Therefore, there is always the possibility that an individual may consent to participating in the interview, only for the interviewer to immediately discover significant cognitive impairment or dementia as suggested by the MoCA. However, cognitive impairment does not preclude an individual from participating in low-risk research.⁴⁴ While interviewers will be trained to seek informed consent from the legal caregiver in these situations, ethical safeguards are further enacted through building capacity assessment into the entire informed consent process, including prior to the administration of the MoCA.

Additional ethical considerations include the consequences of identifying EM such as potential embarrassment and consequences related to mandatory EM reporting duties. We address these concerns by ensuring all participants are well-informed of our duties to disclose with added assurances that participants can skip questions and/or withdraw from participation at any time and for any reason. We also take careful steps to ensure participants are aware of the potential for an APS report should EM be suspected during any part of the study.

The trial conduct will be audited periodically by a data safety monitoring board convened by the National Institute on Aging to ensure adherence to the protocol, regulatory requirements and good clinical practice guidelines. These audits will occur every 12 months and will be coordinated independently from the study team to maintain objectivity.

Any important modifications to the study protocol, such as changes to eligibility criteria, outcome measures or analysis plans, will be promptly communicated to all relevant stakeholders. All site champions will be informed through official study meetings and email communications, with updated protocol documents provided. Any changes will also be submitted for approval to the Institutional Review Board (IRB) overseeing the study, and approval documentation will be obtained before implementing the amendments. If modifications directly impact participants, such as changes to the consent form or study procedures, participants will be re-consented and notified of the changes during their next scheduled visit. Furthermore, the trial registry entry (ClinicalTrials.gov, NCT05958654) will be updated to reflect any amendments, ensuring that the public and scientific community have access to the most recent protocol version.

Upon completing the proposed study, we will have adapted the DETECT screening tool for HBPC. Additionally, we will have completed a stage III efficacy trial and gathered preliminary data regarding DETECT-RPC’s benefits and harms. Access to the final trial data set will be controlled, and only researchers who have received IRB approval and provided a detailed research proposal will be granted access through the National Archive of Computerized Data on Aging. Our findings will be disseminated to the broader scientific community through formal presentations, peer-reviewed journal articles, community meetings and continuing medical education and credentialing for HBCP clinicians.

Ethics statements

Patient consent for publication

Not applicable.

Footnote

X @brad_cannell

Contributors BC, MDL, JB, KLH and CP conceived the study protocol and obtained funding for the study. NS drafted the initial version of the manuscript. All authors made substantial contributions to translating the study’s funding proposal into the manuscript, including revisions reflected in the final manuscript. BC is the guarantor.

Funding This work was supported by was supported by the National Institute on Aging of the National Institutes of Health grant number R61AG078523. Dr. Lees Haggerty’s effort was also supported by National Institute on Aging grant number K01AG076992. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ National Research Council. Elder mistreatment: abuse, neglect, and exploitation in an Aging America. National Academies Press, 2003: 10406.

² Connolly MT, Brandl B, Breckman R. The elder justice roadmap: a stakeholder initiative to respond to an emerging health, justice, financial and social crisis. U.S. Department of Justice, 2014.

³ National Center on Elder Abuse. Frequently asked questions. 2017. Available: https://ncea.acl.gov/faq/index.html

⁴ Lachs MS, Pillemer KA. Elder Abuse. N Engl J Med 2015; 373: 1947–56. doi:10.1056/NEJMra1404688

⁵ Sooryanarayana R, Choo WY, Hairi NN. A review on the prevalence and measurement of elder abuse in the community. Trauma Violence Abuse 2013; 14: 316–25. doi:10.1177/1524838013495963

⁶ Johannesen M, LoGiudice D. Elder abuse: a systematic review of risk factors in community-dwelling elders. Age Ageing 2013; 42: 292–8. doi:10.1093/ageing/afs195

⁷ Brozowski K, Hall DR. Aging and risk: physical and sexual abuse of elders in Canada. J Interpers Violence 2010; 25: 1183–99. doi:10.1177/0886260509340546

⁸ Acierno R, Hernandez MA, Amstadter AB, et al. Prevalence and correlates of emotional, physical, sexual, and financial abuse and potential neglect in the United States: the National Elder Mistreatment Study. Am J Public Health 2010; 100: 292–7. doi:10.2105/AJPH.2009.163089

⁹ Cooper C, Selwood A, Blanchard M, et al. Abuse of people with dementia by family carers: representative cross sectional survey. BMJ 2009; 338: b155. doi:10.1136/bmj.b155

¹⁰ Wiglesworth A, Mosqueda L, Mulnard R, et al. Screening for abuse and neglect of people with dementia. J Am Geriatr Soc 2010; 58: 493–500. doi:10.1111/j.1532-5415.2010.02737.x

¹¹ Pickering CEZ, Yefimova M, Maxwell C, et al. Daily Context for Abusive and Neglectful Behavior in Family Caregiving for Dementia. Gerontol 2020; 60: 483–93. doi:10.1093/geront/gnz110

¹² Perel-Levin S. Ageing and life course unit. Discussing screening for elder abuse at primary health care level by Silvia Perel-Levin. 38. World Health Organization, 2008.

¹³ Lachs MS, Pillemer K. Elder abuse. The Lancet 2004; 364: 1263–72. doi:10.1016/S0140-6736(04)17144-4

¹⁴ Mouton CP, Rodabough RJ, Rovi SLD, et al. Psychosocial effects of physical and verbal abuse in postmenopausal women. Ann Fam Med 2010; 8: 206–13. doi:10.1370/afm.1095

¹⁵ Dong X, Simon M, Evans D. Decline in physical function and risk of elder abuse reported to social services in a community-dwelling population of older adults. J Am Geriatr Soc 2012; 60: 1922–8. doi:10.1111/j.1532-5415.2012.04147.x

¹⁶ Cannell MB, Weitlauf JC, Garcia L, et al. Cross-sectional and longitudinal risk of physical impairment in a cohort of postmenopausal women who experience physical and verbal abuse. BMC Womens Health 2015; 15: 98. doi:10.1186/s12905-015-0258-2

¹⁷ Dong X, Simon MA. Association between elder abuse and use of ED: findings from the Chicago Health and Aging Project. Am J Emerg Med 2013; 31: 693–8. doi:10.1016/j.ajem.2012.12.028

¹⁸ Dong X, Simon MA. Elder abuse as a risk factor for hospitalization in older persons. JAMA Intern Med 2013; 173: 911–7. doi:10.1001/jamainternmed.2013.238

¹⁹ Baker MW, LaCroix AZ, Wu C, et al. Mortality risk associated with physical and verbal abuse in women aged 50 to 79. J Am Geriatr Soc 2009; 57: 1799–809. doi:10.1111/j.1532-5415.2009.02429.x

²⁰ Schofield MJ, Powers JR, Loxton D. Mortality and Disability Outcomes of Self‐Reported Elder Abuse: A 12‐Year Prospective Investigation. J Am Geriatr Soc 2013; 61: 679–85. doi:10.1111/jgs.12212

²¹ Burnett J, Jackson SL, Sinha AK, et al. Five-year all-cause mortality rates across five categories of substantiated elder abuse occurring in the community. J Elder Abuse Negl 2016; 28: 59–75. doi:10.1080/08946566.2016.1142920

²² Lachs MS, Williams CS, O’Brien S, et al. The mortality of elder mistreatment. JAMA 1998; 280: 428–32. doi:10.1001/jama.280.5.428

²³ Fang B, Yan E. Abuse of Older Persons With Dementia: A Review of the Literature. Trauma Violence Abuse 2018; 19: 127–47. doi:10.1177/1524838016650185

²⁴ Van Royen K, Van Royen P, De Donder L, et al. Elder Abuse Assessment Tools and Interventions for use in the Home Environment: a Scoping Review. Clin Interv Aging 2020; 15: 1793–807. doi:10.2147/CIA.S261877

²⁵ Willis J, Antono B, Bazemore A, et al. Primary care in the united states a chartbook of facts and statistics. Robert Graham Center, IBM-Watson Health, The American Board of Family Medicine, 2021.

²⁶ Ahmad M, Lachs MS. Elder abuse and neglect: what physicians can and should do. Cleve Clin J Med 2002; 69: 801–8. doi:10.3949/ccjm.69.10.801

²⁷ American Medical Association. Preventing, identifying & treating violence & abuse. Available: https://www.ama-assn.org/delivering-care/ethics/preventing-identifying-treating-violence-abuse [Accessed 1 Jun 2021 ].

²⁸ Hoover RM, Polson M. Detecting elder abuse and neglect: assessment and intervention. 2014. Available: https://www.aafp.org/afp/2014/0315/p453.html [Accessed 18 Jan 2021 ].

²⁹ US Preventive Services Task Force, Curry SJ, Krist AH, et al. Screening for Intimate Partner Violence, Elder Abuse, and Abuse of Vulnerable Adults: US Preventive Services Task Force Final Recommendation Statement. JAMA 2018; 320: 1678–87. doi:10.1001/jama.2018.14741

³⁰ Stall N, Nowaczynski M, Sinha SK. Systematic review of outcomes from home-based primary care programs for homebound older adults. J Am Geriatr Soc 2014; 62: 2243–51. doi:10.1111/jgs.13088

³¹ Kim CO, Jang SN. Home-Based Primary Care for Homebound Older Adults: Literature Review. Ann Geriatr Med Res 2018; 22: 62–72. doi:10.4235/agmr.2018.22.2.62

³² Cannell MB, Jetelina KK, Zavadsky M, et al. Towards the development of a screening tool to enhance the detection of elder abuse and neglect by emergency medical technicians (EMTs): a qualitative study. BMC Emerg Med 2016; 16: 19. doi:10.1186/s12873-016-0084-3

³³ Cannell B, Livingston M, Burnett J, et al. Evaluation of the Detection of Elder Mistreatment Through Emergency Care Technicians Project Screening Tool. JAMA Netw Open 2020; 3: e204099. doi:10.1001/jamanetworkopen.2020.4099

³⁴ Platts-Mills TF, Sivers-Teixeira T, Encarnacion A, et al. A Scalable Elder Mistreatment Screening and Response Tool for Emergency Departments. Gener J Am Soc Aging 2020; 44: 51–8.

³⁵ Mosqueda L, Burnight K, Gironda MW, et al. The Abuse Intervention Model: A Pragmatic Approach to Intervention for Elder Mistreatment. J Am Geriatr Soc 2016; 64: 1879–83. doi:10.1111/jgs.14266

³⁶ Montaño DE, Kasprzyk D. Theory of reasoned action, theory of planned behavior, and the integrated behavioral model. In: Health behavior and health education: theory, research, and practice. 4th edn. Jossey-Bass, 2008.

³⁷ Borkan JM. Immersion-Crystallization: a valuable analytic tool for healthcare research. Fam Pract 2022; 39: 785–9. doi:10.1093/fampra/cmab158

³⁸ Fernandez ME, Ruiter RAC, Markham CM, et al. Intervention Mapping: Theory- and Evidence-Based Health Promotion Program Planning: Perspective and Examples. Front Public Health 2019; 7: 209. doi:10.3389/fpubh.2019.00209

³⁹ Biros M. Capacity, Vulnerability, and Informed Consent for Research. J Law Med Ethics 2018; 46: 72–8. doi:10.1177/1073110518766021

⁴⁰ Creswell J, Plano Clark V, Gutmann M, et al. Advances in mixed methods research designs. In: Tashakkori A, Teddlie C, eds. Handbook of mixed methods in social and behavioral research. SAGE, 2003: 209–40.

⁴¹ US Preventive Services Task Force. Update on methods: estimating certainty and magnitude of net benefit. US Preventive Services Task Force. Available: https://www.uspreventiveservicestaskforce.org/uspstf/about-uspstf/methods-and-processes/update-methods-estimating-certainty-and-magnitude-net-benefit [Accessed 22 Oct 2024 ].

⁴² Schuchman M, Fain M, Cornwell T. The Resurgence of Home-Based Primary Care Models in the United States. Geriatrics (Basel) 2018; 3: 41. doi:10.3390/geriatrics3030041

⁴³ What is HBPC. Available: https://www.aahcm.org/what_is_hbpc [Accessed 29 Apr 2024 ].

⁴⁴ Prusaczyk B, Cherney SM, Carpenter CR, et al. Informed Consent to Research with Cognitively Impaired Adults: Transdisciplinary Challenges and Opportunities. Clin Gerontol 2017; 40: 63–73. doi:10.1080/07317115.2016.1201714

⁴⁵ Cella D, Riley W, Stone A, et al. The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005-2008. J Clin Epidemiol 2010; 63: 1179–94. doi:10.1016/j.jclinepi.2010.04.011

⁴⁶ Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982; 17: 37–49. doi:10.1016/0022-3956(82)90033-4

⁴⁷ Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606–13. doi:10.1046/j.1525-1497.2001.016009606.x

⁴⁸ Lawton MP, Brody EM. Assessment of Older People: Self-Maintaining and Instrumental Activities of Daily Living. Gerontologist 1969; 9: 179–86. doi:10.1093/geront/9.3_Part_1.179

⁴⁹ Katz S, Downs TD, Cash HR, et al. Progress in development of the index of ADL. Gerontol 1970; 10: 20–30. doi:10.1093/geront/10.1_part_1.20

⁵⁰ Bush K. The AUDIT Alcohol Consumption Questions (AUDIT-C) An Effective Brief Screening Test for Problem Drinking. Arch Intern Med 1998; 158: 1789. doi:10.1001/archinte.158.16.1789

⁵¹ Straus MA. Measuring Intrafamily Conflict and Violence: The Conflict Tactics (CT) Scales. J Marriage Fam 1979; 41: 75. doi:10.2307/351733

⁵² Straus MA, Hamby SL, BONEY-McCOY S, et al. The Revised Conflict Tactics Scales (CTS2): Development and Preliminary Psychometric Data. J Fam Issues 1996; 17: 283–316. doi:10.1177/019251396017003001

⁵³ Cooper C, Maxmin K, Selwood A, et al. The sensitivity and specificity of the Modified Conflict Tactics Scale for detecting clinically significant elder abuse. Int Psychogeriatr 2009; 21: 774–8. doi:10.1017/S1041610209009387

⁵⁴ National Institute on Alcohol Abuse and Alcoholism (NIAAA). National epidemiologic survey on alcohol and related conditions (NESARC) - III [dataset]. 2014.

⁵⁵ Finkelhor D, Shattuck A, Turner H, et al. A revised inventory of Adverse Childhood Experiences. Child Abuse Negl 2015; 48: 13–21. doi:10.1016/j.chiabu.2015.07.011

⁵⁶ Pruchno RA, Michaels JE, Potashnik SL. Predictors of institutionalization among Alzheimer disease victims with caregiving spouses. J Gerontol 1990; 45: S259–66. doi:10.1093/geronj/45.6.s259

⁵⁷ Morycz RK. Caregiving strain and the desire to institutionalize family members with Alzheimer’s disease. Possible predictors and model development. Res Aging 1985; 7: 329–61. doi:10.1177/0164027585007003002

⁵⁸ Given CW, Given B, Stommel M, et al. The caregiver reaction assessment (CRA) for caregivers to persons with chronic physical and mental impairments. Res Nurs Health 1992; 15: 271–83. doi:10.1002/nur.4770150406

⁵⁹ Amirkhan JH. A factor analytically derived measure of coping: The Coping Strategy Indicator. J Pers Soc Psychol 1990; 59: 1066–74. doi:10.1037//0022-3514.59.5.1066