Implementing a financial decision-making scale in APS financial exploitation investigations: Use of the PARIHS conceptual framework.

BACKGROUND AND OBJECTIVES: Adult Protective Services (APS) agencies investigate cases of financial exploitation, and a critical aspect of such investigations is often the assessment of decision-making abilities. This study examined APS workers' implementation of a 10-item financial decision-making screening tool, the Financial Decision Tracker (FDT), across a 34-month period: pre COVID-19, throughout COVID-19 restrictions, and for 1 year following the restrictions. RESEARCH DESIGN AND METHODS: Using the PARIHS implementation science conceptual framework, we examined aspects of context, facilitation, and evidence to determine how well APS workers were trained, certified, and skilled in using the FDT. Using individual and group interviews, we assessed factors often related to successful implementation (context, facilitation) and measured the number of scales used, the types of decisions under investigation, and how the tool's scoring system aligned with the APS workers' final ratings (evidence).
RESULTS: Overall, implementation was sustained throughout the 34-month period. However, during COVID-19 restrictions, usage dropped to a rate 58% below that prior to and following the restrictions. A total of 839 scales were administered. Individuals with no decision-making deficits were slightly older (78.7 vs 77.1 years; t(837)=2.54, p=0.01) and had completed high school at a significantly higher rate (69% vs. 59% χ2(1)=5.20, p=0.023) than those who had decision-making deficits. DISCUSSION AND IMPLICATIONS: Overall, the implementation trial can be considered a success. The FDT was used so often and effectively that it is now an APS best practices tool, meaning that in audits, correct use of the FDT will be assessed.

In the United States, adult protective services (APS) agencies provide protective social services to older adults (typically those age 60 or 65 and older) as well as vulnerable adults (typically, those with serious disabilities). APS agencies are the adult equivalent to Child Protective Services and play a critical role in combating the abuse of older adults and vulnerable adults. In financial exploitation cases, the assessment of decision-making abilities is often a critical part of APS investigations. Indeed, the ability to assess financial decision-making capacity is designated as a core competency of APS workers, as outlined by the National Adult Protective Services Association (2013). Many financial exploitation cases pertain to a financial decision or transaction—for example, the older adult decides to buy a grandchild a car, gift money to someone, or send money to a potential scammer. The majority of older adults, even those with mild dementia, are able to state what they want to do and why they want to do it. The challenge for APS is to determine whether the older adult was able to make an informed financial decision. Informed financial decision-making is based on the legal standards of incapacity. In most states (each has its own legal standards), financial decision-making incapacity is defined as the lack of sufficient understanding of the decision and its consequences. For decades, this has been translated into the elements of choice, understanding, appreciation, and reasoning (Appelbaum & Grisso, 1988). APS workers must therefore assess the client’s understanding and appreciation of financial decisions and decide how to proceed if decision-making difficulties are identified. These are the challenges our work aims to address.

Financial decision-making in gerontology is a subfield unto itself, but in this work and in APS cases, “informed” financial decision-making is based on the legal standards of incapacity. In most states (each has its own legal standards), financial decision-making incapacity is defined as the lack of sufficient understanding of the decision and its consequences. For decades, this has been translated into the elements of choice, understanding, appreciation, and reasoning (Appelbaum & Grisso, 1988). How does an APS worker assess understanding and appreciation in financial decision-making, and what should be done if there are decision-making deficits? These are the challenges our work aims to address.

These assessments often incorporate decision-making capacity knowledge. Even though understanding decision-making capacity’s relation to financial exploitation is one of the 23 core competency areas for APS (Liu & Ross, 2021), most APS workers do not possess this skill. This study examines how the implementation of a 10-item financial decision-making screening tool improves these abilities and evaluates the sustainability of the tool in practice. The implementation study was guided from the outset by an implementation science conceptual framework, and implementation was examined pre-COVID-19, through the COVID-19 restrictions for APS workers, and after the restrictions had been removed. The literature with regard to financial exploitation, financial decision-making, and frameworks for evaluating the implementation of evidence-based tools are reviewed later.

APS, Financial Exploitation Prevalence, and Assessment

Older adult abuse can include physical, sexual, emotional, or psychological abuse; neglect; abandonment; or financial exploitation. In reaction to this growing problem, APS agencies first emerged at local and state levels and only recently received greater support from the federal government. Thus, most APS agencies developed in the absence of federal coordination and without the benefit of comprehensive research in the field of an older adult or vulnerable adult abuse; such research is a more recent phenomenon. Today, APS agencies exist in every state and are normally administered at the local or county level; two thirds of states position their APS agencies within their Department of Social Services. In the remaining states, APS agencies operate within a state department on aging or health.

The need for a method to assess decisional abilities regarding finances is highlighted by the prevalence of financial exploitation among older adults. According to several random sample surveys, the rate of financial exploitation in this population is between 3.5% and 7.3% (Acierno et al., 2010; Anderson, 2013; Laumann et al., 2008). The financial exploitation of older adults is also an expensive societal problem, with losses estimated to be about $2.9 billion each year in the United States (MetLife, 2011); this may be a significant underestimate (Anderson, 2013). Financial exploitation assessment tools, however, are barely more than a decade old (Ernst et al., 2014). The first such tool, the Older Adult Financial Exploitation Measure, uses a self-report method (Conrad et al., 2010) and contains no items that assess decision-making. Two barriers to greater implementation of such tools in APS practice include the paucity of researchers who understand the role and skill sets of APS workers (Ernst et al., 2014) and the lack of fit between available tools and the needs of the APS service delivery system (Onken, 2022). Also, finding the right balance of time needed for training and the ability to use the tools is often difficult.

Clinical Tools to Assess Financial Decision-Making

APS professionals have limited tools available to them for assessing decisional capacity, and many require extensive training and take more time to administer than is feasible in the fast-paced work schedule of an APS practice. The empirical literature on financial decisional abilities has focused almost exclusively on the assessment of financial capacity and cognitive abilities. For example, one of the most comprehensive financial capacity instruments is the financial capacity inventory (FCI; Marson et al., 2000)—a standardized, performance-based measure of financial capacity that has undergone several revisions (Griffith et al., 2003; Marson, 2016; Triebel et al., 2009) and, in its current form, contains 18 tasks that cover nine domains—for instance, basic monetary skills, financial concepts, checkbook management, and financial judgment. The FCI measure benefits from its breadth in the domains covered and its theoretically driven basis. However, this tool is limited in its ability to assess the decision-making process for a real-world transaction the older adult is undertaking. Instead, like other tools, the FCI asks older adults to demonstrate their decisional abilities using hypothetical vignettes (Boyle et al., 2012, 2013). Furthermore, the administration of lengthy capacity measures requires extensive and specialized training to perform accurately and may be perceived as threatening by older adults.

However, a handful of shorter assessment tools are available. The susceptibility to scams questionnaire (STS; James et al., 2014) is a five-item self-report questionnaire in which the examinee rates their agreement with a statement such as, “If something sounds too good to be true, it usually is.” STS items assess the personal characteristics and behavioral indicators most commonly related to risky financial decision-making. Administration of the STS benefits from being extremely brief and is usually nonthreatening to an older adult. However, this measure relies on self-report of the older adult’s beliefs about their behavior and does not directly assess financial decision-making with respect to an actual decision the older adult is considering or the experience of financial exploitation.

Another brief measure that more directly evaluates decisional abilities is the Assessment of Capacity for Everyday Decision-making (ACED; Lai et al., 2008). The ACED is a semistructured interview that is not specific to financial decision-making but can be used for that purpose. The measure assesses decision-making across the four criteria outlined by Appelbaum and Grisso (1988): choice, understanding, appreciation, and reasoning. This short tool consists of 15 items and can highlight specific deficits in decision-making abilities to help professionals and caregivers understand how to best support the older adult in successfully making decisions. The ACED benefits from its brevity and adaptability to a variety of decisional situations. However, because it is a semistructured interview, it is difficult to use as a risk-assessment tool. It also demonstrates some low interrater reliability correlations that may render it less suited to widespread implementation in settings such as APS.

Abrams et al. (2019) examine the implementation of the Interview for Decisional Abilities (IDA) by APS offices in several states. The IDA is another semistructured interview designed to broadly evaluate an adult’s decisional abilities in using the services offered by APS. Like the ACED, the IDA can be flexibly used in a variety of decisional questions, including financial decision-making. Both the IDA and the ACED are grounded in Appelbaum and Grisso’s (1988) decision-making model. Administration of the IDA is a three-step process, in which the interviewer assesses the client’s ability to understand a specific decision, appreciate the potential risks and benefits of various options, and reason through to a decision. At each stage, the interviewer rates the client’s ability as “yes,” “no,” or “maybe.” This tool is not meant to provide a specific risk score for decisional abilities, but rather to promote a dialogue with the client about risk. Training and rollout procedures for implementing the IDA can take several years. Abrams et al. highlight the need to further explore the implementation of the IDA in order to inform best practices about the use of this and similar tools in APS settings.

Whereas the ACED and IDA are based on a sound theoretical framework and were created by experts, a number of questions remain about their utility within APS practices. Can they be used in practice by all APS workers, or are they primarily tools for teams that specialize in financial exploitation? Given that the instruments do not create risk scores, how can fidelity and replicability across different APS raters be examined? Finally, and perhaps most importantly, given the length and complexity of the training coupled with the growing caseloads of APS workers, how widely can they be implemented? In their article on implementation, Abrams et al. presented no empirical findings on the number of cases in which the IDA was used, or on the types of results obtained. There is a lack of empirical studies on the tools used within APS, and the literature on those tools also lacks a broader focus on implementation science.

The Financial Decision Tracker: An Alternative Approach

The Financial Decision Tracker (FDT) was created to provide a risk score based on 10 multiple-choice financial decision-making items (see Campbell et al., 2019; Lichtenberg et al., 2016; and Lichtenberg et al., 2017, for further details). Development of the FDT was guided by two conceptual frameworks: person-centeredness and decisional abilities. These frameworks affirm the importance of assessing the older adult’s understanding of the financial decision in question, with the requirement that the older adult be able to communicate four important elements of their decision: choice, understanding, appreciation, and reasoning. The FDT differs from the ACED and IDA in two ways: (a) it uses a multiple-choice set of items in order to derive a risk score and (b) it is specific to financial decision(s) the older person is making. The FDT contains 10 questions directly related to the elements of informed financial decision-making. The 10-item FDT was validated and cross-validated in studies in which APS workers administered a paper version of the scale, made their ratings, and then entered the scale responses into the website olderadultnestegg.com, which resulted in the creation of risk scores. APS, legal, and financial professionals were trained on the FDT and used it on 212 cases (Lichtenberg et al., 2017). Cases in which financial exploitation was substantiated had significantly higher risk scores than unsubstantiated cases. On a sample of 105 APS cases, this finding was cross-validated (Campbell et al., 2019). In addition, the cutoff score originally derived was supported in the cross-validation study. The FDT, along with narrated online training and certification, became available on https://olderadultnestegg.com, the Older Adult Nest Egg website, in 2018.

The FDT contains 10 items, and an overall risk score was assigned using seven of the items. For example, if the financial decision poses a high risk or entails significant changes to previously established bequests, a higher risk score would be assigned than in cases of minimal financial risk or no changes to bequests. The other three items are descriptive and neutral—for instance, there is no way to determine whether a new will is riskier than a new investment or gift. The FDT is a structured, multiple-choice interview intended to be administered in a standardized fashion. In introducing the FDT to the older adult, the interviewer reads a one-sentence explanation aloud: “I am going to ask you a set of questions to better understand the financial transaction/decision you are making or have already made. Please answer these as best you can and feel free to elaborate on any of your answers.” Questions are to be read aloud as they are written. If the older adult responds before the choices are offered, and a rating can be made, the interviewer can make the rating without reading all of the choices. If necessary, however, the interviewer should read the full list aloud and ask the person to choose one.

The FDT is a rating scale, and therefore the APS interviewer’s judgment is critical. Scoring occurs in two steps: (a) on each item, the older adult’s response is recorded by entering the answer(s) using the online tool (https://olderadultnestegg.com); (b) on each item, the interviewer will be asked whether they agree that this is the most accurate response. Upon entering all answers in the online system, the Older Adult Nest Egg website will produce a risk score and suggested level of concern about the individual’s informed decision-making. The interviewer will then make a final rating of the level of concern regarding informed decision-making. For information on scale psychometric properties and convergent validity, see Lichtenberg et al. (2016, 2017) and Teresi et al. (2017).

Implementation Science

Implementation science examines the translation of evidence-based practices into widespread usage. To do so, it uses scientific conceptual models and methods to discern processes that are not typically governed by rationality. If the adoption of evidence-based practices were straightforward and rational, it would consist of adopting passive methods to disseminate evidence-based practices. Onken (2022) described three critical challenges to successful implementation efforts. First, implementation requires direct and effective strategies; thus, passive methods most often fail. Second, the complexity of an evidence-based tool or intervention might need to be decreased and its simplicity increased. Finally, the ability to train people to successfully deliver the assessment is a critical component. We will next explore widely established conceptual frameworks and describe our efforts to implement an evidence-based risk-assessment for cases of suspected financial exploitation of older adults.

Kitson et al. (1998) proposed a conceptual framework for understanding implementation successes and failures. Their Promoting Action in Research Implementation in the Health Services (PARIHS) framework is one of the most popular and widely studied models. The core of their model is that the implementation of evidence-based practice has three components, (a) evidence (i.e., the quality of the research and widespread acceptability of its utility, as well as its translation into standardized training); (b) context (i.e., leadership, support, and receptivity to innovation/change within the organization); and (c) facilitation (i.e., the skills and attributes of an individual who directly supports the process of implementation). Kitson et al. (2008) revisited the model and focused on the challenges of PARIHS. The authors sought to further define the three concepts while acknowledging the challenges entailed in utilization. Evidence, they argued, includes research, clinical experiences, and patient preferences—and implementing evidence involves negotiation, a shared understanding of benefits, and a team effort. They also point out that some contexts are more conducive than others to implementation, and highlight the importance of facilitation. Harvey and Kitson (2016) state that the framework, which is multidimensional in scope, emphasizes the complex nature of implementation. Rarely, they note, is their implementation direct; it usually involves “tinkering” in order to best fit the organization’s needs. For this, they added an “i” for innovation to the framework—which makes it an i-PARIHS framework. Rycroft-Malone (2004) further stated that the addition of innovation encourages clarity with respect to defining evidence and its intended use; this includes investigation of issues related to leadership, the feedback systems that evaluate the context’s quality, and identifying both internal and external facilitation.

Reviews of implementation science frameworks have reached some common conclusions. Nilsen (2015) found that most studies provide limited information on how to carry out implementation; they believe that the PARIHS model is well suited to be a determinant framework (i.e., to document how implementation can succeed). Lundmark et al. (2021) concluded that whereas alignment between the goals and abilities of a program in which the tool or intervention will be implemented is extremely important, few studies focus on alignment. Finally, Birhen et al. (2020) noted that most implementation science studies focus on the initial uptake of evidence-based practices and fail to examine sustainability; their declining use following initial implementation is well documented.

Purpose of the Present Study

The purpose of the study was to describe the implementation of training, certification, and scale usage of the FDT, an empirically validated 10-item financial decision-making scale. Specifically, all aspects of the PARIHS model (evidence, context, and facilitation) were documented and assessed regarding how well adoption of the PARIHS model translated into the appropriate use of the FDT scale and how well the risk-scoring system compared with the APS worker’s conclusions.

Implementation procedures and questions for this study:

(1) Context

(1a) Determining how the need for risk scales for APS workers will enhance the APS leadership’s interest in testing the implementation of training and scale usage.

(1b) How well does creating partnerships with APS leadership and area supervisors enhance the adoption of FDT?

(1c) At the conclusion of the implementation trial period, will the state adopt the FDT as a best practice instrument?

(2) Facilitation

(2a) Can implementation be facilitated at multiple levels, including at the senior administrative level of APS and the sector supervisor level, and by the principal investigator?

(2b) Does bidirectional feedback enhance motivation regarding the adoption of FDT in practice across APS workers and across the APS leadership in the state?

(3) Empirical Outcomes

(3a) Almost all APS staff will become certified to administer the FDT after completing the online narrated training on the scale’s purpose, administration, and use.

(3b) Expected agreement between the FDT risk score and the APS worker’s rating will be 85% for completed scales. FDT risk scores will differentiate those cases in which there is concern about decision-making abilities from those cases in which there is no concern about decision-making abilities.

Method

Validity Studies

Engagement with APS began in 2015 when the lead author partnered with the policy leaders of the state APS office. The policy leader, in turn, identified APS supervisors interested in piloting the scales. Between 2015 and 2017, a select and small frontline staff responsible for administering the FDT underwent training on the appropriate use of the scale. Following a trial period, APS supervisors shared feedback regarding their staff’s use of the FDT in daily practice, as well as recommendations for improving the functionality of the paper-and-pencil version of the FDT. Supervisor feedback improved this version of the FDT by providing scoring reminders and bullet point direction reminders on the one-page form. A total of 300 scales were collected and analyzed, and the results of the findings were published in three papers on the reliability and validity of the FDT (Campbell et al., 2019; Lichtenberg et al., 2016, 2017).

Implementation Trial

Having established an evidence base for the tool, it was possible to design an implementation trial. The trial began in April 2019, and data were collected through the end of 2021.

Procedures

The electronic version of the FDT was launched in 2018. During this time, the online training modules and certifications also became available. APS supervisors were the first to undergo training and obtain certification, followed by APS caseworkers in early 2019. The online training was augmented by in-person training, along with coaching both before and after each group was trained.

Measures

Demographic measures of age, education, gender, and race were collected for the older adult clients. Education was combined into three groups: elementary and middle school only, some high school, and completed at least high school. This descriptive information is important for determining whether the scale was implemented across different demographic groups.

Context

This was assessed through conversations with the APS state director and sector supervisors when discussing how the scale might benefit APS workers across the state.

Facilitation

Facilitating elements were documented through web-based and in-person training by the FDT author, feedback sessions with APS staff, and interviews with APS supervisors after several months of implementation.

FDT

In this sample, Chronbach’s alpha for the FDT was 0.79. The FDT was administered orally by the APS worker as an interview, and ratings were recorded offline. The older adult’s responses and the APS worker’s individually rated items were then entered into the olderadultnestegg.com scoring system.

(1) Type of decision was documented by the APS worker in question #1 of the FDT.

(2) Financial decision tracker risk score. Based on an algorithm of seven of the 10 items, the FDT produces a single overall risk score.

(3) Concurrence of website risk score with the APS worker’s final rating was determined by reviewing all cases on the website.

Statistical Procedures

t Tests and chi-square analyses were carried out for measures across the entire sample, and for each of the three phases of the study.

Results

Four major elements related to context were identified throughout the study:

(1) An audit of Michigan APS completed prior to the creation of the FDT indicated a lack of risk-scoring tools’ use in cases, and especially financial exploitation. Thus, according to senior leadership, the timing of FDT introduction was a good fit for APS. This contextual issue was reinforced by section supervisors, who wanted their staff to have access to risk-assessment tools.

(2) Access to the Older Adult Nest Egg website for training and certification, calculating risk scores, and receiving recommendations enabled statewide implementation. Because the training was done at a time convenient for each APS worker and was narrated and self-paced, all 400 APS workers and supervisors were successfully trained and certified to use the FDT. Thus, the ability to be trained virtually using self-paced, narrated training was a necessary component in being able to train and certify all APS workers.

(3) Within the first year of the implementation trial, an APS liaison was assigned to expand the implementation of the FDT. This provided an opportunity for the first author to have access to both APS senior leadership and broader APS staff for input. This development turned out to be both a contextual and facilitation component. Structurally, having a line of communication with the liaison allowed for much easier access to senior management but also wider access to APS workers in the field.

(4) Michigan APS is divided into five geographic sectors. The continuity of all APS sector supervisors and their support provided fertile ground for implementation. In contrast, some states organize their APS services by county—and because there are so many counties in most states, attempting to work with each county supervisor would be a massive effort.

These elements of context were vital for enabling the training and certification of APS staff and establishing lines of communication with the five sector supervisors.

Eight major facilitation elements emerged as the implementation progressed.

(1) The first author traveled to each APS sector to provide two trainings—one just prior to implementation and one a month after implementation began. Using a new risk-assessment tool requires considerable trust from not only supervisors but also APS professionals in the field. The commitment to older adults’ well-being and to providing useful tools to APS workers was greatly enhanced by face-to-face training.

(2) The first author was able to review cases on the olderadultnestegg.com system and requested clarification via e-mail with the APS worker and supervisor for cases in which the tool may not have been properly administered. There is a learning curve with any rating scale, and by having the ability and allotting time to review cases in a timely manner, we were able to identify common mistakes and devise solutions to reduce those mistakes.

(3) The strong commitment of sector supervisors demonstrated their support for use of the tool for all APS staff; also, supervisors encouraged staff to generate ideas regarding how best to implement the FDT. Every 2 weeks, sector supervisors received an e-mail listing the dates, respective APS workers, and the conclusion of the FDT. This facilitated sector supervisors’ ability to discuss the tool in the context of specific cases.

(4) A large feedback session organized by the APS liaison led to improved processes for APS workers. Twenty of the most frequent users of the FDT and five nonusers were assembled through Zoom for a feedback session. The APS workers were asked to comment on whether the tool added value to their work, and to describe how the tool was used. There was overwhelming support for the use of the tool and its added value to their work. The biggest question during the feedback session was, “When do we use this tool?” The FDT isn’t used, for example, in cases of theft because there was no decision on the older adult’s part. The FDT can only be used when linked to a financial decision or set of decisions. APS workers also did not want to use the tool if the older adult already had a guardian.

In terms of improved processes, a decision tree and FDT usage sheet were created. Stems from the decision tree include: (a) Does this case involve a report of financial loss? (b) Is the older adult their own decision-maker (e.g., no conservator, guardian, rep-payee, or Durable Power of Attorney)? and (c) Does the older adult appear to be a vulnerable adult (a requirement for elder financial abuse in Michigan)? Factors include:

Cognitive impairment

Frail appearance

Dependency/ unable to perform own activities of daily living (ADLs)

Lack of control over own finances

In response to the second biggest question—“What can we do with the results?”—the FDT usage sheet offered guidance for how to use the FDT report with law enforcement, prosecutor’s offices, and probate judges. For example, if the score is above the cutoff for high risk, the following ideas guidance is offered:

(1) Discuss threats and explore ways to protect the individual. Information on the FDT can be useful for health professionals, judges, prosecutors, guardians, and Guardian Ad-Litems.

(2) Ask to speak with the client’s trusted friend or family member.

(3) Recommend a complete mental health evaluation to examine possible depression, anxiety, cognitive deficits, and substance abuse.

(4) Recommend that the client postpone financial transitions.

(5) Share publications from the Older Adult Nest Egg resource page with law enforcement, prosecutors and judges, that discuss the components of informed decision-marking and the evidence base for the FDT, which can be found on at olderadultnestegg.com.

(5) The electronic record used by APS had a specific FDT results section for financial exploitation cases. Site supervisors’ ability to examine APS workers’ documentation of FDT results also facilitated the use of the scale, and thus the FDT should be embedded in the routine processes of APS sectors.

(6) The FDT training and certification process was integrated into the onboarding process for new APS workers. During the study period, many APS offices received funds to increase staff; this meant that many workers were hired well after the implementation trial began. APS offices often experience turnover or expansion. By making FDT training and certification a requirement for newly hired APS workers, the scale’s use was largely ensured.

(7) The first author provided refresher trainings to APS sectors. APS workers have large caseloads and insufficient time to spend on all of them. The job can entail high stress, and the ongoing contact between the first author and APS workers in the field enhanced their enthusiasm for using the FDT and sharing case examples in which the tool was helpful.

(8) APS case studies and feedback were integrated into trainings and widely disseminated. In a few cases, the use of the FDT was associated with saving an older adult as much as $2 million. These were eye-opening for staff. In other cases, the FDT was associated with convincing law enforcement to investigate and was a useful resource for prosecutors.

Empirical Results From Implementation Prior to and Throughout the Pandemic

The results of this implementation trial cover 34 months. We will present data for the entire trial, but also for the three phases: (a) pre-COVID-19 pandemic; (b) during COVID-19 restrictions, when APS workers were not performing face-to-face assessments in-person; and (c) after COVID-19 restrictions had been lifted and face-to-face, in-person assessments resumed.

Across the 34-month implementation trial, APS staff administered 839 FDTs. Just over 5,000 APS financial exploitation cases were investigated during this period. The state is unable to provide the breakdown, but several of the cases were for vulnerable adults younger than age 60, and several cases involved exploitation in which no financial decision was made by the older person. Thus, it is impossible to calculate what percentage of financial decision cases involved an adult 60 years or older.

Table 1 reports the financial decisions investigated for possible financial exploitation. Overall, the three decisions most frequently examined by APS are (a) allowing someone to take over the older person’s finances (35.6%); (b) giving a gift (26.9%); and (c) participating in a scam (22.5%). As can be seen in Table 1, giving a gift and allowing others to take over the individual’s finances occurred significantly more often in cases in which there were no decision-making concerns. In contrast, participating in a scam was more likely to be associated with decision-making deficits. Of the 189 scam cases, 67% were associated with older adults’ demonstrated decision-making deficits. Participation in a scam increased during the pandemic, from 17% of cases in the prepandemic phase to 30% in the final phase. No other type of decision was associated with an increase or decrease in percentages over the course of the 34 months.

Table 1.

Types of Decisions Made by Older Adults in Adult Protective Services Cases for Financial Exploitation

		Interviewer score
	Overall sample	No concerns	Concerns
	(N = 839)	(n = 468)	(n = 372)
Decision type	n	n	n	Chi-square
A. Giving a gift	226 (26.94%)	133 (58.8%)	93 (41.2)	χ 2(1) = 9.36, p = .002
B. Making a purchase	65 (7.75%)	43 (66.1%)	22 (33.9%)	χ 2(1) = 5.55, p = .018
C. Participating in a scam	189 (22.53%)	62 (32.8%)	127 (67.2%)	χ 2(1) = 23.75, p < .001
D. Allowing someone else access to your money	60 (7.15%)	33 (55.0%)	27 (45.0%)	χ 2(1) = 1.67, p = .197
E. Allowing someone else to take over your finances	299 (35.64%)	197 (65.8%)	102 (34.2%)	χ 2(1) = 20.87, p < .001

Note: Data collected from April 12, 2019 to December 31, 2021.

*p < .05; **p < .01.

In Table 2, the concurrence between interviewer rating and the risk score determined by the scoring algorithm can be found. The APS worker agreed with the risk score 92% of the time. The risk score is dependent on the APS worker’s data entry, so this result is not particularly surprising. It is, however, important in an implementation trial, because concurrence indicates two things: (a) that the scores make sense to the APS worker and (b) that the APS worker is following the FDT instructions. A case example provides some description of an APS worker’s use of the scale. An 88-year-old widower was befriended by a 33-year-old woman who shortly after moved in with him. Soon, his monthly expenses increased fivefold, and his adult children filed a financial exploitation complaint with APS. The APS worker used the FDT and found that the 88-year-old had no knowledge that his expenses had increased and no understanding of the affect on his finances or the financial risks involved. He claimed that the expenditures benefited him more than anyone, but a review of his credit card expenses revealed that almost none of the money was spent on items he was associated with. Armed with this high-risk score and the APS worker’s agreement that decision-making deficits were present, the APS worker was able to convince law enforcement to investigate; they discovered that the woman was felon wanted for similar behavior in a neighboring state.

Table 2.

Interviewer Agreement With Risk Score for Overall Sample (N = 839)

	Interviewer agreed with FDT risk rating	Interviewer disagreed with FDT risk rating	Interviewer reduced risk rating compared with FDT risk recommendation	Interviewer increased risk rating compared with FDT risk recommendation
	n	n	n	n
Cases	773 (92.13%)	66 (7.87%)	31 (3.69%)	35 (4.17%)

Note: FDT = Financial Decision Tracker.

Implementation Results Across the Three Phases of the Study

The COVID-19 pandemic and resulting restrictions served to illustrate the strengths and vulnerabilities of the sustainability of FDT implementation. During Phase I—which lasted 1 year—400 FDTs were administered, for an average of 33 cases per month. Informed decision-making deficits were associated with 51%. Phase II lasted for 10 months, and COVID-19 restrictions did not allow APS workers to interview clients in-person. Only 114 scales were administered, for an average of just over 11 per month, and only 35% of cases were found to have decision-making deficits. In Phase III, during which the COVID-19 restrictions were lifted, APS workers interviewed clients face to face and in-person. This phase lasted 1 year; 325 scales were administered, for an average of 27 per month—83% of the average number of cases per month in Phase I. FDT scores differed significantly in the expected direction across decision-making groups in each of the three phases. Interpretation of these findings included that (a) the FDT is likely to be most sensitive when used in face-to-face interactions; nonverbal behaviors, such as body language, may be important for making both the APS worker and the older adult more comfortable about discussing sensitive matters. Although the tool itself should be robust enough to administer by phone, focused discussion of financial decision-making matters remains a very sensitive area and it may be that APS workers were less comfortable in this regard in the context of COVID-19 restrictions. (b) The FDT, despite the drop in administration during the COVID restriction period, demonstrated good sustainability: once the restrictions were lifted, use of the scale resumed to near pre-COVID-19 levels.

Discussion

This implementation trial contained four unique elements. First, it was guided by the PARIHS model, beginning prior to the trial, so that all three aspects of the trial—context, facilitation, and evidence—could be examined. Second, the trial began a full year before the COVID-19 pandemic and lasted a full year after COVID-19 restrictions were lifted; this enabled investigation of how the pandemic and the restrictions affected implementation. Third, the length of the trial, 34 months, enabled analysis of the sustainability of the implementation and not simply the initial results. And finally, the tool implemented assessed financial decision-making in APS financial exploitation investigations—an area widely recognized as critical to such investigations but one in which empirical findings are sparse.

The PARIHS model, although simple, is also elegant in its attention to context and facilitation issues. The “fit” of the practice to be implemented begins with the alignment of contextual elements. In this trial, risk-assessment tools—which, according to APS senior management, are plentiful in Child Protective Services—are lacking for older and vulnerable adults. Contextual issues are important across the entire organization. The relative centralization of APS in Michigan, with its five sectors, each containing multiple counties, was another key element in the ability to even launch an implementation trial. Finally, being able to not only deliver training but also capture, in real-time, use of the tool—and provide feedback—was another contextual element vital to this trial.

Despite the alignment of these contextual issues, facilitation efforts were enormously impactful in this trial. Support throughout the statewide APS organization and its willingness to include the first author as a partner in all aspects of training and consultation were critical to the trial’s success. Not only did the sector supervisors and APS liaison support the study; they actively enhanced it. As the result of feedback from these supervisors, the PDF version of the FDT was improved, a decision tree and usage guide were created and disseminated, and successful case studies were widely shared. Facilitation by the first author was a constant throughout the trial. Feedback for supervisors and case workers every few weeks helped to keep the FDT on the radar of all APS staff, who responded positively to initial and updated training sessions provided in-person or remotely by the first author.

Overall, the implementation trial can be considered a success. The FDT was used so often and effectively that it is now an APS best practices tool; this means that in audits, the decision tree will be used to assess how well the tool is being used. In other words, when using the FDT decision tree, if the use of the FDT is indicated, it is expected that the APS professional will administer it and document its use. This enhances the likelihood of long-term sustainability. Also, we were able to elaborate on the elements of context and facilitation that may well generalize to other APS settings across the country. Finally, over 800 scales were administered in a 34-month period, for an average of just under 25 cases per month, and the risk-scoring system was consistent with APS workers’ overall ratings over 90% of the time.

The ability to screen for informed financial decision-making is an essential skill for APS workers. The most prominent issue in cases in which APS is investigating losses due to a financial decision or set of decisions an older person has made is whether the financial decision was informed. Some will argue that if the older person knows what they want to do (choice) and why (rationale), they should simply be allowed to proceed. Legal standards, however, require more from the older decision-maker; they must communicate understanding of what they are doing and an appreciation of the risks/benefits of the decision. The identification of decision-making deficits allows the APS worker to help curb or eliminate financial exploitation. As with any screening tool, the APS worker’s skill is the most important skill involved. The FDT enhances this skill by providing evidence-based questions and a scoring system that enables the efficient identification of decision-making deficits. The FDT can also help the APS worker more easily communicate their findings to other professionals who may need to be involved (e.g., banks, law enforcement, judges, and health providers).

There are several limitations to this study. First, we were not able to follow each case to its end, and thus we do not know how the FDT tool applied to the substantiation of cases, or how many times it was used to communicate with other health professionals, law enforcement, or prosecutors. The implementation trial covers a single state. While other states or counties within states have expressed interest in our trainings and tools, to the best of our knowledge, none have made the tool part of their standard exploitation investigation. We were unable to perform any test/retest assessment or use other methods to rate the quality of how the tool was used across cases. Despite these limitations, this implementation trial makes a significant contribution to the financial exploitation field. We identified the types of financial decisions most associated with these cases, and how the patterns of these decision types changed across the pandemic. We effectively trained and certified hundreds of APS staff and documented widespread use of the tool, accompanied by reports of high levels of satisfaction. The APS implementation trial withstood COVID-19 restrictions and demonstrated that it could be sustained. Finally, we demonstrated the value of using an implementation science conceptual framework from the planning to concluding phases of an implementation trial.