Development and Spanish Translation of the Weinberg Center Risk and Abuse Prevention Screen (WC-RAPS)

Abstract

Barriers for enhanced detection, identification, and reporting of elder abuse include the paucity of appropriate, valid, easily-administered screening tools. This article describes the qualitative methods used in the development of the Weinberg Center Risk and Abuse Prevention Screen (WC-RAPS), and of its Spanish version. Focus groups and cognitive interviews were instrumental in identifying problematic items, underscoring potential response errors and informing about putative causes for divergent interpretations of item-intent. Seven of the eleven original items were modified, a double-barreled item was segregated into two, one item deleted, and three additional items included to create the final 13 WC-RAPS items. The multi-step approach implemented for the Spanish conversion evidenced deviation from the original intended meaning for one item. The readability for English and Spanish versions was also assessed. Screening for elder abuse, if implemented systematically can be instrumental in identifying unrecognized abuse and preventing reoccurrence.

Introduction

Increased support for elder abuse research in the behavioral social and medical sciences has heightened efforts to better estimate its prevalence, and identify risk factors, and their sequelae. State-level mandatory reporting laws have also promoted community awareness of the issue (U.S. Government Accountability Office, 2011). Nonetheless, elder abuse continues to be considered a prevalent, under identified, and rising public health concern at the national level. Moreover, population growth projections suggest a potential for increased incidence of elder abuse (Ortman, Velkoff, & Hogan, 2014), with concomitant challenges for the health care, social service, and criminal justice systems.

Research efforts geared toward evaluation of the effectiveness of community-based elder abuse interventions are scarce, (Ayalon, Lev, Green, & Nevo, 2016; Baker, Francis, Hairi, Othman, & Choo, 2016; Pillemer, Burnes, Riffin, & Lachs, 2016). Thus, some of the elder abuse interventions extant have been modeled on successful child abuse initiatives and/or domestic violence programs (Wolfe, 2003). A recent state of the science article on prevention of elder abuse highlights screening efforts as one of the applications of “lessons learned” from those fields (Teresi et al., 2016). Barriers for enhanced detection, identification, and reporting of elder abuse include the paucity of appropriate, valid, easily administered screening tools (Cohen, 2011; Schofield 2017). The absence of screens targeting the long-term support and services (LTSS) population highlights the need for development of an elder risk and abuse screen for this growing and ethnically diverse group. The WC-RAPS was developed to assess risk and abuse among applicants to a variety of LTSS programs targeted for abuse prevention, such as institutional and community shelters for elder abuse (Reingold, 2006), home care, nursing care, legal representation and crisis counseling.

Elder abuse screening can serve two main preventive purposes: 1) to anticipate the risk for abuse by identifying current risk factors, and 2) to flag present, existing abuse. In both instances, screening is aimed to target suitable interventions either to prevent victimization (abuse from occurring) or reducing the risk of revictimization (should abuse be substantiated). Literature reviews on elder abuse screening instruments extant have documented: a) the intended purpose of the selected measures (e.g., to document signs via examples of abuse events or physical evidence), and/or risk factors), b) the intended targeted domains included (e.g., physical-, psychosocial-, financial abuse, neglect), c) the psychometric qualities of the measures (when available), d) the application methods used (e.g., self-reports, professional inventories, clinical assessments), and e) the targeted settings in which the screenings are to be conducted (e.g., clinical and or institutional vs. community-based; see Abolfathi Momtaz, Hamid & Ibrahim, 2013; Cohen, 2011; Fulmer, Guadagno, Dyer, & Connolly, 2004; National Center on Elder Abuse, 2016; Schofield, 2017). Recommendations have been made for further evaluation of elder abuse screening measures with respect to their validity, accuracy, transferability of the application across settings and across culturally-diverse populations, as well as the screening effectiveness (Abolfathi Momtaz et al., 2013; Beach, Carpenter, Rosen, Sharps & Gelles, 2016; Cohen, 2011; Fulmer et al., 2004; National Center on Elder Abuse, 2016; Schofield, 2017).

Qualitative methods have been recommended (Krause, 2006) and implemented for measure development in efforts to improve the quality of survey items (Christodoulou, Junghaenel, DeWalt, Rothrock, & Stone, 2008; Nápoles-Springer, Santoyo-Olsson, O’Brien, & Stewart, 2006; Ramirez, et al., 2013). Addressing response bias, these methods are particularly helpful to gain insight into respondents’ interpretation and conceptualization of items, which may or may not be consonant with the original intent of the items. Qualitative methods are also instrumental in assessing and addressing conceptual equivalence in the context of comparative research, by highlighting the potential underlying putative sources of divergent responses from members of groups that represent populations with diverse socio-demographic characteristics (e.g., race, ethnicity, age, education level; Nápoles-Springer et al., 2006; Ramírez, Ford, Stewart, & Teresi, 2005; Ramirez, Teresi, Ogedegbe, & Williams, 2016; Teresi, Stewart, Morales, & Stahl, 2006).

The aims of this article are to describe: a) the set of qualitative methods used in the development of the Weinberg Center Risk and Abuse Prevention Screen (WC-RAPS), and b) the development of the Spanish version.

Conceptual Model

A heuristic elder abuse prevention intervention research model that posits elder abuse as a stressful life experience served as the theoretical framework for this effort (See Teresi et al., 2016). This model is well-suited to address elder abuse screening, considered conceptually as a primary prevention effort, in the context of applied research and practice. The model suggests that primary prevention interventions (such as abuse screening) may operate as moderators impacting elder abuse outcomes (directly and indirectly), via the reduction of abuse-related stress and of negative health consequences on the older adult. Conceptually and for assessment purposes, risk factors and elder abuse are distinct constructs (See Pillemer et al., 2016), both relevant for elder abuse screening efforts (Beach et al., 2016). Research-based studies to examine the effectiveness of screening as a favorable intervention contributing to elder abuse prevention have been recommended (Cooper, Manela, Katona, & Livingston, 2008; Nelson, Bougatsos, Blazina, 2012; Nelson, Nygren, McInerney, & Klein, 2004).

Methods

Qualitative Methods for Item Development

Development of the Item Pool:

A multi-step, qualitative methods approach was used in the development of dichotomous, close-ended items for the elder abuse screen. These steps were implemented iteratively. First, a literature review was conducted for the creation of an item pool, consisting initially of 30 items. These items were arrayed in a table by instrument and author.

Panel Review:

A panel of experts reviewed the items for relevance and applicability to the measurement of abuse and risk in an LTSS setting. A final set of 11 items was developed as the first version of the screen by a panel of elder abuse experts (comprised of two graduate-level social workers, an attorney, and a graduate-level nurse), based on content, face validity, and considering parsimony and time-burden for ease of administration. Items were contrasted against similar items from existing elder abuse measures, and wording modifications were implemented as necessary to guard against potential intellectual property issues.

Pilot Testing:

These 11 items were piloted by four clinical practitioners (two community-based social workers and two nursing home rehabilitation social workers) with 20 older adult volunteers from senior citizen day programs.

Focus Groups:

Two focus group meetings with these practitioners were held to discuss observed implementation problems and potential solutions. The panel of experts made decisions about item selection, modification, and adjustment based on feedback obtained and information from the prior steps.

Cognitive Interviews:

The selected, modified set of items was then tested by an elder abuse specialist social worker via cognitive interviews using concurrent structured probes with eight residents of an independent-living facility (seven females and one male) for clarity of content, interpretation, and comprehension of item intent. The cognitive interview participants were identified based on specified selection criteria, i.e., no- to mild cognitive impairment with ability for verbal communication. Word probes were implemented to assess comprehension of specific words or phrases (e.g., “managing your money,” “refused to help you,” and “touched you inappropriately”), and paraphrase probes for item comprehension (e.g., “Are you afraid of anyone that you know?”; “Does a family member depend on you for care, shelter, or financial support?”).

Recommendations and Final Review:

A final review of the items for inclusion, adaptation, and rewording was conducted by the panel of experts taking into account information derived through the above comprehensive discourse. The final product was a set of 13 items (see the Results section below). All decisions by the panel at each step were reached by consensus. Three items were added, one as a concurrent criterion variable: contact with adult protective services, other a result of separating a double-barreled item into two, and the third was a new risk of abuse indicator (described below).

Spanish Translation Method

The multi-step approach implemented for the development of the Spanish translation of the WC-RAPS was modeled after a protocol for multinational translations (Acquadro, Conway, Hareendran, & Aaronson, 2008), and the one used by Eremenco, Cella, and Arnold (2005) in the conversion of health status questionnaires.

Forward and Back Translations:

Two independent forward translations were performed by native Spanish-speaking health professionals. A reconciled translation was then developed by an experienced native Spanish-speaking bilingual translator. This reconciled version was back-translated independently by a native English-speaking educator fluent in Spanish, unfamiliar with the original English version.

Final Review and Adjudication:

The reconciled-, the back-translated, and the original (English) versions were compared and reviewed by three bilingual health professionals for harmonization. The final adjudication was performed by two reviewers: an English-dominant graduate-level public health elder abuse expert, and a native Spanish speaking, Ph.D.-level bilingual professional, experienced in measure translations. Discrepancies were resolved by consensus.

Readability

The readability (the difficulty level of written text) of the WC-RAPS was assessed for the final English (13-item) version and the translated Spanish version. The overall, and the item-level readability of the English version was assessed by the Flesch-Kincaid Reading Ease Index (FREI) and the Flesch-Kincaid grade level (F-K) methods available in Microsoft Word 2016. The overall, and item-level readability of the translated Spanish WC-RAPS was assessed by the Fernández Huerta formula (Blanco Pérez & Gutiérrez Couto, 2002; Fernández Huerta, 1959), the Spanish adaptation of the Flesch-Kincaid (Muñoz Fernández, 2017a). The grade level index for the Spanish version was obtained from a reference table provided by the same source (See Muñoz Fernández, 2017b) which is based on inferences made on the Fernández Huerta index. This approach is different from that for the computation of the F-K grade level, which is based on direct item assessment.

Results

Development of the final item set

Based on the clinical expertise of the panel of experts, modifications to the items were implemented iteratively, based on the information gathered via the focus groups and the cognitive interviews. Feedback obtained from focus groups conducted with clinical staff as well as the results of cognitive interviews highlighted problematic items due to lack of clarity in terms of language and/or item intent. The potential sources of response bias were also ascertained.

The item: “Do you have access to your money?” was found to be misinterpreted to inquire about the relative ease of visiting a bank branch and/or using an ATM machine rather than to tapping risk for financial abuse. This question was revised to: “Do you have trouble managing your money (paying bills, accessing cash, etc.)?”, for clarity of intent and to facilitate interpretation. The intention of the item: “Are you able to go out of your house when you want?” was deemed unclear. It was often interpreted to be assessing the respondent’s physical ability to ambulate independently, rather than the actual freedom to come and go without interference and/or coercion from another individual (i.e., a potential abuser). The item was reworded to: “Does anyone prevent you from seeing friends or family?” in order to enhance item comprehension. The question: “Are you free to use the telephone or computer privately?”, intended to assess the respondent’s ability for unrestricted communication (i.e., without interference and/or coercion by a potential abuser), was flagged as confusing. Respondents evidenced difficulties in answering it by inaccurately interpreting it to inquire about restrictive features in their physical living environment. That is, some of the respondents based their negative-direction responses on the layout of their living space, which hindered privacy in using a phone and/or computer. The question thus, was reworded to: “Does anyone prevent you from using a telephone or a computer?”, in order to address the source of confusion. In a similar fashion, explanatory or descriptive phrases and/or words were added or substituted in remaining items for simplicity, clarity and/or to facilitate comprehension of item intent. For instance, “Has anyone ever refused to help you?” was modified to read: “Has anyone ever refused to help you with household chores, medications, appointments, etc.?”. “Does your caregiver depend on you for shelter or financial support?” was adapted to: “Does a family member depend on you for care, shelter, or financial support?”. “Are you afraid of anyone in your life?” was revised to read “Are you afraid of anyone that you know?”. “Has anyone touched you without your permission now reads “Has anyone touched you inappropriately without your permission?”. The double-barreled question addressing close relatives’ alcohol and substance abuse, and mental illness: “Do you live with anyone or have close family members who abuse drugs and alcohol or have a psychiatric illness?” was segregated into two separate items, one inquiring about close relatives’ alcohol and substance abuse, the other addressing mental illness. Another double-barreled item, “Has anyone called you names or threatened to hurt you?”, was retained, however. Its content was deemed sensitive due to its reference to physical violence thus, a “softening” statement (i.e., calling names) might help to elicit an affirmative response, when appropriate. The item “Have you ever been contacted by Adult Protective Services / APS?” was endorsed by the panel of experts for inclusion as a more objective indicator of abuse or risk of abuse and was also used as an indicator of concurrent criterion validity. In most cases, this self-report variable was later confirmed through contact with APS.

In summary, seven out of the eleven original items had to be re-written or modified, a double-barreled item was segregated into two separate items, one of the original items was deleted, and three additional items were developed in order to create the final set of 13 items for the WC-RAPS (See Table 1).

Table 1.

Item modification for WC-RAPS, Spanish translation, readability indices, and similar items from other elder abuse screening measures

Original WC-RAPS items	Final WC-RAPS items based on results from qualitative methods. Spanish translation of items.	Readability (English) FREI/Flesch-Kincaid grade level	Readability (Spanish) Fernandez-Huerta (grade level)	Similar existing items/Measure
Do you have access to your money?	RH2. Do you have trouble managing your money (paying bills, accessing cash, etc.)?	60.7/7.7		Has anyone managed or is anyone managing your money without your consent? /Geriatric Mistreatment Scale (GMS) (Giraldo-Rodríguez & Rosas-Carrasco, 2013)
	¿Usted tiene problemas administrando su dinero (pagando facturas, teniendo acceso a dinero en efectivo, etc.)?		45.74 (select courses before university)
Has anyone ever refused to help you?	RH3. Has anyone ever refused to help you with household chores, medications, appointments, etc.?	43.9/10.3		Has anyone prevented you from getting food, clothes, medication, glasses, hearing aids or medical care from being with people you wanted to be with?/Elder Abuse Suspicion Index (EASI) (Yaffe, Wolfson, Weiss & Lithwick, 2008)
	¿Alguna vez alguien se negó a ayudarle con las tareas domésticas, medicamentos, citas, etc.?		62.3 (7th or 8th grade)	Has anyone ever failed to help you take care of yourself when you needed help?/ American Medical Association-Screen for Various Types of Abuse and Neglect (AMA-SVTAN). (AMA, 1992).
				Has anyone kept you from getting clothes, footwear, etc.? Has anyone kept you from receiving the medications you need?/ Geriatric Mistreatment Scale (GMS) (Giraldo-Rodríguez & Rosas-Carrasco, 2013)
Has anyone taken things or money that belong to you without your permission?	RH4. Has anyone taken things or money that belong to you without your permission?	63.4/7.6		Has anyone tried to force you to sign papers or to use your money against your will?/ Elder Abuse Suspicion Index (EASI) (Yaffe et al., 2008)
	¿Alguna vez alguien ha tomado cosas o dinero que le pertenecen a usted sin su permiso?		85.88 (5th grade)
				Has anyone taken things that belong to you without your O.K.?/Vulnerability to Abuse Screening Scale (VASS) (Schofield & Mishra, 2003)
				Has anyone taken anything that was yours without asking?/ Has a stranger ever spent (insert name’s) money or sold (his / her) property without (his / her) permission/American Medical Association-Screen for Various Types of Abuse and Neglect (AMA-SVTAN) (AMA, 1992).
				Has a stranger ever spent your money or sold your property without your permission/National Elder Mistreatment Study) (Acierno et al., 2010)
				Has anyone taken any of your belongings without your permission?/Geriatric Mistreatment Scale (GMS) (Giraldo-Rodríguez & Rosas-Carrasco, 2013)
Does your caregiver depend on you for shelter or financial support?	RH5. Does a family member depend on you for care, shelter, or financial support?	56.9/8.5		Does your caregiver depend on you for shelter or financial support?/Questions to Elicit Elder Abuse. (Carney, Kahan, & Paris, 2003).
	¿Hay un miembro de su familia que depende de usted para cuidado, refugio, o apoyo financiero?		78.68 (6th grade)	Are you helping to support someone?/ Hwalek-Sengstock Elder Abuse Screening Scale (H-S/East) (Neale, Hwalek, Scott, & Stahl, 1991).
Has anyone called you names or threatened to hurt you?	RH6. Has anyone called you names or threatened to hurt you?	95.1/2.4		Have you been upset because someone talked you in a way that made you feel shamed or threatened?/ Elder Abuse Suspicion Index (EASI) (Yaffe et al., 2008).
	R6. ¿Hay alguien que le ha llamado por “malos nombres” a usted (le ha dicho insultos) o le ha amenazado con hacerle daño físico?		94.18 (4th grade)	Have you been upset because someone talked you in a way that made you feel shamed or threatened?/ Elder Abuse Suspicion Index (EASI) (Yaffe et al., 2008).
				Has anyone ever scolded or threatened you?/American Medical Association-Screen for Various Types of Abuse and Neglect (AMA-SVTAN). (AMA, 1992).
Are you able to go out of your house when you want?	RH7. Does anyone prevent you from seeing your friends or family?	66.1/6.2		In the past 12 months, has NAME OF ALLEGED ABUSER prevented you from contacting family, friends, or community resources?/Older Adult Psychological Abuse Measure (Conrad, Iris, Ridings, Langley, & Anetzberger, 2011).
	¿Alguien le impide ver amigos o familiares?		74.87 (6th grade)
				My partner tried to keep me from seeing or talking to my family./The Psychological Maltreatment of Women Inventory (PMWI). (Tolman, 1999).
Are you free to use the telephone or computer privately?	RH8. Does anyone prevent you from using a telephone or a computer.	52.8/8.3		In the past 12 months, has NAME OF ALLEGED ABUSER prevented you from having contact with the outside world via telephone, newspapers, television, or radio, etc.?/ Older Adult Psychological Measure (OAPAM) (Conrad et al., 2011).
	¿Alguien le impide usar el teléfono o la computadora?		69.05 (7th or 8th grade)
Are you afraid of anyone in your life?	RH9. Are you afraid of anyone that you know?	92.9/2.2		Has anyone made you afraid, touched you in ways that you did not want, or hurt you physically?/ Elder Abuse Suspicion Index (EASI) Yaffe et al., 2008)
	¿Usted le tiene miedo a alguien que conoce?		97.76 (4th grade)
				Are you afraid of anyone at home?/ American Medical Association-Screen for Various Types of Abuse and Neglect. (AMA-SVTAN). (AMA,1992).
				Has anyone made you feel afraid?/Geriatric Mistreatment Scale. (Giraldo-Rodríguez & Rosas-Carrasco, 2013)
Has anyone touched you without your permission?	RH10. Has anyone touched you inappropriately without your permission?	18.9/12.6		Has anyone made you afraid, touched you in ways that you did not want, or hurt you physically?/ Elder Abuse Suspicion Index (EASI) (Yaffe et al., 2008)
	¿Alguien le ha tocado inapropiadamente sin su permiso?		59.96 (high school)
				Has anyone ever touched you without your consent?/ American Medical Association-Screen for Various Types of Abuse and Neglect (AMA-SVTAN) (AMA,1992).
Do you live with anyone or have close family members who abuse drugs and alcohol or have a psychiatric illness?	RH11. Do you live with anyone or have close family members who abuse drugs or alcohol?	67.5/7.5		Does anyone in your family drink a lot?/Hwalek-Sengstock Elder Abuse Screening Scale (H-S/East) (Neale et al., 1991).
	¿Vive con alguien o tiene familiares cercanos que abusan de drogas o alcohol?		80.21 (5th grade)
Has anyone forced you to do things you did not want to do?				Has anyone forced you to do things you didn’t want to do? /Hwalek-Sengstock Elder Abuse Screening Scale (H-S/East) (Neale et al., 1991).
	RH12. Do you live with anyone or have close family members with a psychiatric illness?	71.7/6.7
	¿Vive con alguien o tiene familiares cercanos con una enfermedad mental?		70.43 (6th grade)
	RH13. Have you had multiple hospitalizations in the last 6 months? How many? #____	59.7/6.4
	¿Ha tenido múltiples hospitalizaciones en los últimos 6 meses? ¿Cuántas? #____		50.18 (high school)
	RH14. Have you ever been contacted by APS (adult protective services)?	52.8/8.3
	¿Alguna vez ha sido contactado por APS (Servicios de protección para adultos)?		77.96 (6th grade)
WC-RAPS Total		57/8	58.66 (high school)

Spanish translation

The back-translation in one of the Spanish-translated items evidenced a discrepancy indicative of a deviation from the intended meaning of the original (English) item, thus a potential source for response bias. The connotation of physical harm implicit in the phrase “threatened to hurt you?” [Has anyone called you names or threatened to hurt you?] was not reflected in the Spanish translation. Modifications were implemented accordingly to include the concept of physical harm. The final item in Spanish now reads: “¿Hay alguien que le ha llamado por ‘malos nombres’ a usted (le ha dicho insultos) o le ha amenazado con hacerle daño físico?” No further adjustments were considered necessary when the revised Spanish- and the forward translated versions were evaluated against the final set of the English items for fidelity.

Readability

The Flesch-Kincaid Reading Ease Index (FREI) computed at the item-level for the 13-item English version ranged from 18.9 to 95.1; higher scores are indicative of reading ease. The item-level F-K grade level indices ranged from 2.2 to 12.6. The item-level readability index of the Spanish version assessed by the Fernández Huerta formula ranged from 45.74 to 97.76; similar to the FREI, higher scores reflect reading ease. The FREI- and the Fernandez Huerta’s-based indices were generally consonant. Few notable discrepancies at the item-level between the English and Spanish versions across the respective indices were shown (See Table 1). Among the easiest items in terms of readability for both the English and Spanish versions were: “¿Usted le tiene miedo a alguien que conoce?” [“Are you afraid of anyone that you know?”] and “¿Hay alguien que le ha llamado por ‘malos nombres’ a usted (le ha dicho insultos) o le ha amenazado con hacerle daño físico?” [“Has anyone called you names or threatened to hurt you?”]. Examples of the most difficult items in the English version are: “Has anyone touched you inappropriately without your permission?” and “Has anyone ever refused to help you with household chores, medications, appointments, etc.?” Examples of the most difficult items in the Spanish version are: “¿Usted tiene problemas administrando su dinero (pagando facturas, teniendo acceso a dinero en efectivo, etc.)?” [“Do you have trouble managing your money (paying bills, accessing cash, etc.)?”] and “¿Ha tenido múltiples hospitalizaciones en los últimos 6 meses? ¿Cuántas?” [“Have you had multiple hospitalizations in the past 6 month? How many?”]

The overall FREI and F-K indices for the English version were generally comparable to those of the Spanish version. The readability index of the translated Spanish version using the Fernández Huerta formula (58.66) was similar to the FREI (57.0) for the English version; both reflect relatively moderate levels of reading ease. In terms of the grade level indices, the English version is deemed to be readily comprehensible for individuals at a U.S eighth- grade reading level, while the Spanish version for a high school reading level. (See Table 1). A caveat is that divergent methods were used to generate the grade level estimates for the English and Spanish versions, both at the item-and measure-level. Rather than based on direct item assessment like the English version (F-K grade level index), the grade level for the Spanish version was obtained from a reference table, based on the Fernandez Huerta index (See Muñoz Fernández, 2017b). Additionally, while the F-K grade level index for the English text was provided in decimal numbers (e.g., for RH10= 12.6; for the total scale= 8.0) its equivalent for the Spanish text was available as broad categories, (e.g, “high school level” for both, RH10 and the total scale) (see Table 1).

Discussion

Challenges in the assessment and identification of elder abuse have been recognized and discussed extensively. These challenges are in part a function of the societal lack of awareness and/or resistance to acknowledge the phenomenon, unwillingness of the victims to admit or label behavior as “abusive” (particularly in the context of co-dependent relationships), divergent definitions of what represents abuse/ mistreatment, and/or “intent” to inflict harm. However, arguably, at the center of the barriers to understanding the prevalence and consequences of elder abuse is the scarcity of reliable and valid screening measures (Schofield, 2017). This paucity in turn, has public health implications because lack of measures hinders the implementation of a systematic screening process as a preventive intervention. This paper addresses this gap by applying qualitative methods to the development of the Weinberg Center Risk and Abuse Prevention Screen as well as its Spanish version.

The implementation of a set of qualitative methods used iteratively in the development of the WC-RAPS not only flagged problematic items in terms of clarity, comprehension, and item-intent, but also shed light on the underlying putative sources of the problems. Thus, collectively these methods provided information instrumental to revision, adjustment, and enhancement of the questions, to inform the final selected set of items. The relevance of implementing a standardized methodology in the context of cross-cultural research was also evidenced by highlighting the conceptual adequacy/inadequacy of translated items, which in turn, carries implications for measurement equivalence.

The overall readability indices for the English and Spanish were close; few notable discrepancies in the ease of language were evidenced at the item level as reflected by the indices. It is possible that these discrepancies may be an artifact of the computation algorithm which considers word length and the number of syllables per word in the calculation, given that the Spanish version of the items generally, tended to be longer (wordier). Additionally, while the F-K grade level index for the English text was provided in decimal numbers, the Fernández Huerta grade level index was only available as discrete categories. This lack of specificity (categories) made the comparability of readability in terms of grade level in particular, challenging. For example, the F-K grade level index for item RH10= 12.6, and for the total scale= 8.0; the respective indexes for the Spanish text (the Fernández Huerta grade level index) were “high school level” for both. Evidently, the Fernández-Huerta (see Muñoz Fernández, 2017b) “high school level” category could potentially encompass F-K grade level indexes between 9 to 12. In such case, an F-K grade level readability score of 8th grade (in English text) might (or might not) represent a meaningful discrepancy from that of Fernández Huerta “high school level” (see Muñoz Fernández, 2017b). Because the other readability index (FREI) for the total English items was only 1.7 points lower than that of the Spanish version by the Muñoz Fernández calculation (57.0 vs 58.66, respectively), and because the same elements (word length and the number of syllables per word) are used in the calculation of both, the FREI and the grade level indexes; arguably, the “high school level” readability index for the total scale reported by Muñoz Fernández could actually reflect that of early high school grades, thus, demonstrating acceptable English-Spanish comparability in terms of readability.

Complementary to the qualitative analyses presented, the need for further psychometric examination of the WC-RAPS in order to establish its psychometric properties and metric equivalence of the scale is acknowledged. The examination of the reliability and concurrent criterion validity of the WC-RAPS is the aim of a companion paper. Because the screen will be used with individuals who differ in terms of background characteristics, e.g., race and ethnicity; it is important to establish that the WC-RAPS measures constructs that are actually applicable across diverse groups. Further analyses using advance psychometric methods to examine measurement equivalence are also recommended.

The final version of the WC-RAPS contains 13 dichotomous, close-ended (Yes/No) items designed to tap existing signs of elder abuse (7 items), and risk of elder abuse (5 items). An additional item measures APS contact. The screening tool requires the assessor to collect additional information such as basic demographic information (name, age, race, ethnicity, gender, living arrangement), and to request consent for its administration. The respondents are offered the option of refusing to answer any question prior to its administration. The protocol provides space at the end of the page for the assessor to note the language in which the interview was conducted, the reason why the interview was not completed (if that was the case), and to describe the respondent’s cognitive capacity (“known impairment, suspected impairment, no impairment”). After completion of the interview the assessor flags if risk and/ or abuse was suspected (based on responses and the overall demeanor during the interview), in which case further investigation to establish “caseness” is warranted. Currently, reliance is on clinical interpretation of responses to decide whether to flag risk or follow-up with further evaluation. Future research could identify optimal cut points on the risk and abuse screens.

Conceptually, (consistent with the theoretical framework), and relevant to clinical and social service practice, well-implemented screening for elder abuse can be instrumental in: identifying unrecognized abuse (detection), barring it from occurring, and/or preventing its reoccurrence. It has been argued that to be an effective preventative intervention, screening for elder abuse should become a standard procedure in health care delivery (American Medical Association [AMA], 1992; Fulmer, Paveza, Abraham, & Fairchild, 2000).

In conclusion, a methodical process grounded on investigation of the circumstances, a systematic assessment, and a substantiation of abuse adjudication, elder abuse screening is fundamental in facilitating the implementation of primary- (e.g., APS, multi-disciplinary teams) and secondary preventive interventions, such as dedicated elder abuse shelters (Heck & Gillespie, 2013; Reingold, 2006, Teresi et al., 2016), which attempt to alleviate re-victimization risk, increase safety, and to ultimately enhance older adults’ quality of life. Further evaluation of the effectiveness of abuse preventive programs is necessary (Ernst et al., 2014) and measurement is essential (Beach, et al., 2016; Teresi et al., 2016) so that improved, evidence-based interventions can be evaluated and promoted for replication.

Limitations

A limitation of this effort is that the equivalence of the English and Spanish version of the WC-RAPS was merely examined in terms of language conversion. The applicability and the cultural relevance of the Spanish version of the WC-RAPS should be further evaluated through cognitive interviews with Spanish-speaking older individuals. Caution is also advised in the interpretations of the grade level readability estimates of the Spanish version because they were not computed directly but were obtained from a reference table, based on the Fernandez Huerta index. Also, the Muñoz-Fernández grade level index was only available as discrete categories lumping all high school grades, which made the English-Spanish comparability of readability in terms of grade level challenging.

Additionally, as noted earlier, further psychometric examination of the WC-RAPS is recommended in order to establish its factor structure, reliability, validity, and measurement equivalence. Both conceptual and metric equivalence are required in order to establish the accuracy necessary for comparative efforts.