Development and Validity Testing of an Assessment Tool for Domestic Elder Abuse

Background: The older adult population is increasing in number, and elder abuse is expected to become a more pressing problem. Developing tools to assess the presence and severity of elder abuse is important to both effectively prevent this abuse and provide increased support for families. Purpose: This study was intended to test the validity of an Assessment Tool for Domestic Elder Abuse (ATDEA). The items that constitute this tool were derived from a literature review. Methods: Two rounds of self-administered questionnaire surveys were conducted with nurses working at home-visit nursing stations. Round 1 was used to evaluate the face validity, and Round 2 was used to test the content using the content validity index (CVI). Results: Two hundred forty nurses participated in the two studies. In Round 1, 56 nurses evaluated 38 items derived from a literature review, resulting in the development of a 36-item ATDEA. In Round 2, 184 nurses evaluated the content validity of the 36-item ATDEA. The Item-CVI (I-CVI) scores ranged from .61 to 1. Twenty-eight of the items met or exceeded the I-CVI threshold of .78, whereas the eight items assessing self-neglect did not. The overall Scale-CVI score for the assessment tool was .90, which met the threshold of .90. Conclusions/Implications for Practice: The results of validity testing established the preliminary validity of this assessment tool. In addition, as self-neglect is known to damage the well-being of older adults, six of the eight items assessing self-neglect were retained in the ATDEA despite their failure to meet the threshold I-CVI of .78. The remaining two self-neglect items were not included in the ATDEA because of excessively low I-CVI scores (< .70). Thus, the final version of the ATDEA includes 34 items. The authors recommend that nursing professionals use the ATDEA as a checklist to assess the presence of elder abuse and to discern the subtypes and severity of this abuse. When evaluating elder abuse, the higher the degree of severity, the greater the urgency to provide support.

unreasonably restricted or suspended. It is necessary therefore to achieve a comprehensive understanding of what constitutes inappropriate care and human rights violations against older adults. When mistreatment becomes evident, it is also important to assess its severity so that the situation may be rectified. Among the commonly used scales, the Indicators of Abuse (IOA) Screen (Reis & Nahmiash, 1998), which is administered by a trained professional and differentiates between potentially abusive and nonabusive situations at the client's home, is used to assess abuse severity and focuses on the mental and psychosocial characteristics of both the caregiver and the care recipient. The IOA Screen uses a 5-point scale that ranges from 0 = nonexistent to 4 = yes/severe. Except for the IOA, no other instrument that is able to assess the degree of abuse has been developed through rigorous research.
For both moral and practical reasons, it is necessary to develop an assessment tool that covers all of the subtypes of domestic elder abuse and is based on current concepts that assess both the presence and degree of severity. To develop the Assessment Tool for Domestic Elder Abuse (ATDEA), a literature review was conducted via a keyword search of ICHUSHI Web (a Japanese medical database), China National Knowledge Infrastructure (a Chinese-language journal database), and PubMed using the keywords ''elder abuse'' and ''elderly abuse'' as well as a manual search of books and Web sites related to elder abuse. This search elicited 38 variations of elder abuse (Yi, Honda, & Hohashi, 2015), and a 38-item initial version of ATDEA was developed to cover seven subtypes of elder abuse, including four items on physical abuse, eight on psychological abuse, seven on neglect, three on economic abuse, four on sexual abuse, four on social abuse, and eight on self-neglect. This study was thus designed to evaluate the face validity and content validity of ATDEA to ensure the validity of this tool in clinical practice.

Methods
Two rounds of surveys were conducted. Round 1 evaluated the face validity of ATDEA, and Round 2 tested the content validity of ATDEA based on the results of Round 1.

Design and Study Sample
A home-visit nursing station is a facility where nurses provide nursing and care services, encourage self-reliance, and support recuperation in people afflicted with disease or disabilities who live at home. Home visits help nurses assess the presence and circumstances of domestic elder abuse. This study used a questionnaire survey to investigate the elder abuse experiences of nurses working at home-visit nursing stations.
In terms of sample size, Ishii and Tao (2002) indicated that 20 participants are needed in studies of the wording and expression of items to test face validity, and Polit and Beck (2006) recommended more than six participants to meet the content validity index (CVI) minimum threshold of .78. Therefore, to fulfill these criteria and ensure the precise testing of ATDEA validity, the authors recruited participants from all of the home-visit nursing stations in a certain ward of a city in Japan. After searching for information on home-visit nursing stations from the Web site of the Prefecture Home-Visit Nursing Station Liaison Committee, research request and questionnaires (five questionnaires per station) were mailed to 65 home-visit nursing stations in one city (Round 1) and 175 home-visit nursing stations in another eight cities (Round 2). Home visit nurses who met the inclusion criteria of having a registered nurse license and an experience related to the care of older adults, and who did not fit the exclusion criterion of no experience dealing with elder abuse, were asked to complete a self-administered questionnaire. The data were collected in SeptemberYOctober 2015 for Round 1 and from November 2015 to February 2016 for Round 2.

Sample characteristics
All of the participants were requested to provide their age, gender, and years of work related to the care of older adults.
Evaluation of the items (Round 1) A free description column was set up to ask for revision recommendations and nurses' views on each item of the 38-item initial version of ATDEA. Examples of recommendations and views include problems with item phrasing, overlapping items, and inappropriate items, among others.

Validity of the items (Round 2)
A 4-point scale (1 = not relevant to 4 = highly relevant) was used to assess whether the item was appropriate to the assessment of abuse as well as whether the item pertained to the classified subtype of abuse.

Round 1
The phrasing of items was discussed and revised based on evaluation by nurses. Nurses' opinions as to why items were unworkable were sorted by the subtypes of elder abuse, and each key point was treated as a single code. On the basis of the similarity of content, all codes were organized and synthesized to create categories, and the number of occurrences of each category was calculated. The process of categorization was conducted repeatedly until general consensus was reached among the nine researchers to ensure the rigor of the analysis.

Round 2
To evaluate the content validity, each item was coded as ''relevant'' for the assessment tool if the nurses scored the item as 3 = quite relevant or 4 = highly relevant, which was consistent with the protocol described by Polit, Beck, and Owen (2007). The CVI was calculated according to Lynn's quantification method of content validity (Lynn, 1986). The Item-CVI (I-CVI), which is seen as the content validity of an individual item, was computed as the number of content validators who answered ''relevant'' for an item divided by the total number of content validators. The I-CVI should be 1 in the case of five or fewer content validators. In cases with six or more content validators, the threshold may be relaxed to no lower than .78. The Scale-CVI (S-CVI), which represents the content validity of the overall scale, refers to the average of the I-CVIs for all items. The average congruency percentage for the S-CVI should be .90 (Polit & Beck, 2006;Shi, Mo, & Sun, 2012). In this study, ATDEA was presumed valid when items achieved an I-CVI of no lower than .78 and an S-CVI of .90 or above.

Ethical Considerations
This study was approved by the institutional review board of the researchers' university (Approval no. 390). An anonymous, self-administered questionnaire was used; all nurses were informed that participation was voluntary and that no names or personal details would be included in the study. No incentives were offered for participation. Returning the completed questionnaire to the researchers constituted agreement to participate.

Round 1: Evaluation of Face Validity
Sample characteristics Valid responses were obtained from 56 participants. (The response rate could not be determined because of uncertainty in the total number of home visit nurses.) All of the 56 participants were female; 19 (33.9%) were in their 40s, and 18 (32.1%) were in their 50s ( Table 1).

Opinions of the items
Except for sexual abuse, 18 categories were extracted based on the subtypes of elder abuse and named as ''reasons why items are seen as unworkable'' (Table 2). Four categories under ''physical abuse'' were extracted: (a) coercive feeding is necessary for the elderly's nutrient retention, (b) restraint is needed for the elderly and caregiver safety, (c) short-term physical restraint for treatment may be possible, and (d) assessment of physical abuse depends on the degree of the acts. Three categories under ''neglect'' were extracted: (a) consideration of the family's economic situation is essential, (b) consideration of the elderly's will and values is needed, and (c) unintentional neglect occurs because of a caregiver's lack of knowledge. Two categories under ''psychological abuse'' were extracted: (a) necessary to consider family relationships and recognition of the elderly, and (b) in the caregiver's mind, treating the elderly like a child may be acceptable based on sentiments toward his or her own parents. Three categories under ''economic abuse'' were extracted: (a) the assessment of economic abuse depends on the elderly's judgment ability, (b) management of assets by the elderly himself or herself may be difficult when considering his or her cognitive ability, and (c) assessment of economic abuse depends on whether it economically disadvantages the elderly. Two categories under ''social abuse'' were extracted: (a) appropriate restriction is essential based on the elderly's cognitive ability, and (b) appropriate restriction for the elderly's safety is needed. Finally, four categories under ''self-neglect'' were extracted: (a) consideration of the values the elderly and the way he or she lived is necessary, (b) consideration is needed for the elderly's decisions and judgment, (c) consideration is essential for the psychosomatic state of the elderly, and (d) we may lack cognition/knowledge concerning self-neglect.
Participants provided opinions regarding the way items were phrased, such as ''Add a word [intentionally] to the item of 'psychological abuse' to indicate 'Intentionally made to feel anxiety','' ''The two items of 'Intentionally ignored' and 'Estranged from his/her family' under 'psychological abuse' overlap,'' and ''The meanings of 'Insulted, ridiculed' and 'Humiliated in front of others' under 'psychological abuse' are somewhat similar.'' Phrasings and descriptions were revised based on this feedback, and the two items of ''psychological abuse'' that were identified as overlapping with other items were deleted. Thus, the final assessment tool comprised 36 items under the following categories: physical abuse (4), neglect (7), psychological abuse (6), sexual abuse (4), economic abuse (3), self-neglect (8), and social abuse (4).

Round 2: Evaluation of Content Validity
In total, 207 home visit nurses participated in this validation study. (As noted previously, because the total number of home visit nurses at those 175 facilities was unclear, calculation of the response rate was impossible.) Responses from five participants who did not provide demographic information, 13 who did not respond to the questionnaire, and five who gave inappropriate answers (evaluating the validity of the act itself instead of the content of the item) were removed from further analysis. Consequently, valid responses were obtained from 184 participants.

Evaluation of the Assessment Tool
Overall assessment tool The S-CVI of the assessment tool was .90, and eight of the 36 items did not meet the I-CVI threshold of no lower than .78. Of the 28 items that met or exceeded the I-CVI of .78, 26 had I-CVI scores of .90 or above, and two had I-CVI scores between .78 and .90.

Items
The I-CVI scores of the 36 items based on the seven subtypes of elder abuse are presented in Table 3.

I-CVI no lower than .90
Twenty-six items had I-CVI scores 9 .90. Among these, two items under physical abuse (''subjected to physical violence [e.g., hitting, beating, punching]'') and economic abuse (''forced into issuing or changing the content of documents concerning assets'') earned the highest I-CVI score of 1. Seven items had an I-CVI score of .99, including one item of neglect (''provided with an inadequate/insufficient diet''), two items of psychological abuse (''subjected to a torrent of verbal abuse [e.g., bellowing, cursing, derogatory remarks]'' and ''intimidated, threatened [e.g., threatening of the elderly such as making hostile gestures]''), three items of sexual abuse (''inappropriately sexually touched without his/her consent,'' ''forced to discuss sex-related topics,'' and ''forced to watch sexual images or videos''), and one item of economic abuse (''assets [e.g., savings, property] disposed of/used without his/her consent'').

Subtype of Elder Abuse
Category (n of codes) Physical abuse 1. Coercive feeding is necessary for the elderly's nutrient retention. (5) 2. Restraint is needed for the elderly and caregiver safety. (4) 3. Short-term physical restraint for treatment may be possible. (2) 4. Assessment of physical abuse depends on the degree of the acts.
(2) Neglect 1. Consideration of the family's economic situation is essential. (4) 2. Consideration of the elderly's will and values is needed. (3) 3. Unintentional neglect occurs because of a caregiver_s lack of knowledge.
(2) Psychological abuse 1. Necessary to consider family relationships and recognition of the elderly. (4) 2. In the caregiver's mind, treating the elderly like a child may be acceptable based on sentiments toward his or her own parents.
(1) Economic abuse 1. The assessment of economic abuse depends on the elderly's judgment ability. (4) 2. Management of assets by the elderly himself or herself may be difficult when considering his or her cognitive ability. (3) 3. Assessment of economic abuse depends on whether it economically disadvantages the elderly.
(2) Social abuse 1. Appropriate restriction is essential based on the elderly's cognitive ability. (7) 2. Appropriate restriction for the elderly's safety is needed.
(3) Self-neglect 1. Consideration of the values of the elderly and the way he or she lived is necessary. (6) 2. Consideration is needed for the elderly's decision and judgment. (6) 3. Consideration is essential for the psychosomatic state of the elderly. (4) 4. We may lack cognition/knowledge concerning self-neglect.

Discussion
The results of the content validity testing support the validity of the 28 non-self-neglect items in the ATDEA. Although the validity of the eight items pertaining to self-neglect was not supported by the measures, the authors believe that their content validity is partially supported based on previous findings (Ruhl, Scheich, Onokpise, & Bingham, 2015) and that these items remain helpful to this type of survey, given that they were derived from a literature review (Yi et al., 2015). Moreover, because self-neglect is known to seriously damage the health, safety, and well-being of the older adult (Gibbons, Lauder, & Ludwick, 2006) and undermine the older adult's capacity for self-care (Mardan, Hamid, Redzuan, & Ibrahim, 2014) and is associated with increased mortality (Dong et al., 2009), it remains vital to assess self-neglect to enable experts to implement appropriate support measures that increase self-care skills and enhance well-being in older adults. Considering these reasons and on the basis of previous research that set the I-CVI threshold as 9 .70 (Nakagami, Yamauchi, Noguchi, Maeda, & Nakagami, 2014), the two items pertaining to self-neglect that had I-CVI scores lower than .70 were deleted. All participants (N = 184) recognized ''subjected to physical violence'' as an abusive behavior because this type of abuse frequently leaves visible wounds, bruises, or bone fractures (Wang, 2006), making it easy to recognize. Perhaps, because the participants were aware that the item ''forced into issuing or changing the content of documents concerning assets'' may involve an illegal or immoral act, it achieved the highest I-CVI score of 1. One item pertaining to neglect, ''provided with an inadequate/insufficient diet,'' and two items pertaining to psychological abuse, ''subjected to a torrent of verbal abuse (e.g., bellowing, cursing, derogatory remarks)'' and ''intimidated, threatened (e.g., threatening of the elderly such as making hostile gestures),'' may be assessed and discovered easily from the older adult's psychosomatic state when the home visit nurse provides nursing, physical, and psychological care. Therefore, these items achieved a high I-CVI score of .99. A study conducted by Lo, Lai, and Tsui (2010) to elicit the perception, knowledge, and awareness of student nurses regarding elder abuse suggests that students recognize both physical and psychological abuse as behaviors denoting elder abuse. The results of this study also support this finding. For two items pertaining to physical abuse, namely, ''restrained physically'' and ''restrained by improper dosage of medicine,'' although analysis of participant opinions suggested that ''Restraint is needed for the elderly and caregiver safety,'' the Ministry of Health, Labour and Welfare, Japan (2001) has specified that, when the status of the older adult does not meet all three requirements of ''urgency,'' ''nonsubstitutability,'' and ''temporality,'' any form of physical restraint, in principle, is regarded as elder abuse. Thus, in the absence of all three of these requirements, a nurse must recognize physical restraint as elder abuse, irrespective of the older adult's cognitive or health status. Regarding ''force-fed,'' which is also categorized as physical abuse, some nurses noted that ''Coercive feeding is necessary for the elderly's nutrient retention.'' Despite retention of minimal nutritional status as important for older adult health, if coercion causes mis-swallowing or oral injury, it should be viewed as physical abuse.
Older adults, even those who are cognitively impaired, deserve the respect of both family members and medical professionals. Considering their vulnerability and support needs, treating them like children may injure their self-esteem or even damage their mental health. ''Treated like a child'' therefore constitutes psychological abuse. The reason that the I-CVI score of this item was lowest among the six items pertaining to psychological abuse was that, ''in the caregiver's mind, treating the elderly like a child may be acceptable based on sentiments toward his or her own parents.'' Neglect refers to failure to take care of the older adult. However, because of the caregiver's or family's economic circumstances, it may sometimes be difficult to financially support or provide necessities to the older adult. Therefore, on the basis of the opinion from the nurses that ''Consideration of the family's economic situation is essential,'' a premise that the caregiver or family has a certain level of economic means is necessary before the two items of ''refused financial support'' and ''refused to provide necessities such as clothing, eyeglasses, hearing aids, etc.'' may be justifiably identified as abusive behaviors.
Cognitive impairment in older adults is a risk factor for elder abuse (Cooper et al., 2006;Johannesen & LoGiudice, 2013;Yaffe & Tazkarji, 2012), and degree of cognitive impairment has been associated with risk of economic abuse (Garre-Olmo et al., 2009). This means that older adults with cognitive impairments may be more vulnerable to economic abuse. However, considering that the cognitively impaired older adult may have difficulty in managing assets, refusing to allow access to funds may in some cases be reasonable. Therefore, taking into account the participant opinion that ''Management of assets by the elderly himself or herself may be difficult when considering his or her cognitive ability'' only when the caregiver or family member has no reasonable excuse, the item ''kept from managing/using his/her own assets without reasonable excuse'' may be justifiably viewed as economic abuse.
Professionals working in community-based comprehensive support centers in Japan have encountered some dilemmas regarding whether to respect the individual's self-determination when using intervention methods to work toward a solution for self-neglect. Examples include balancing the older adult's self-determination with the need to safeguard life and balancing respect for the older adult's self-determination and the professional's mission to support well-being (Hamazaki et al., 2011). In this study, under the category of self-neglect, nurses claimed that ''consideration of the values of the elderly and the way he or she lives is necessary,'' ''consideration is needed for the elderly's decision and judgment,'' and ''consideration is essential for the psychosomatic state of the elderly.'' These statements reflect a phenomenon in Japan: The dilemma over whether the older adult's self-determination ought to be respected not only causes confusion among professionals managing self-neglect cases but also interferes with the assessment of self-neglect. In the United States, where a national elder abuse support center was established in 1987, the issue of self-neglect is being formally addressed, and elder abuse prevention agencies such as the National Center on Elder Abuse and the National Committee for the Prevention of Elder Abuse are already dealing with selfneglect as a subtype of elder abuse. In Japan, however, despite Tatara having presented the concept of self-neglect in 1987, more than 90% of the older adult welfare departments and community comprehensive support centers have expressed the unmet need for countermeasures to self-neglect in the older adult (AI-advocacy Support Network, 2016). Currently, the various services in Japan remain inadequate for dealing with self-neglect. The nurses' comments that ''We may lack cognition/knowledge concerning self-neglect'' mirrors the fact that even professionals do not fully recognize self-neglect and fail to deal with it sufficiently. Because of these considerations, the I-CVI score for self-neglect did not meet the threshold. Despite the lack of content validity, as the presence of self-neglect plays a pivotal role in dealing with this problem, the authors agreed to include six items in the ATDEA and exclude the remaining two items with I-CVI scores lower than .70.
Most of the current screening and assessment instruments for elder abuse have been developed in the context of Western cultures. However, as noted by Yan, Chan, and Tiwari (2015), because of considerable cultural differences, instruments developed in that context may not be suited for use in Asian cultural contexts. Recognizing the need for an assessment tool consistent with Asian social and cultural norms, ATDEA was created to better assess the state of elder abuse as well as to facilitate the provision of adequate support. In developing this assessment tool, the authors reviewed articles and reports from Japan, China, and Western countries to create the assessment items (Yi et al., 2015). The S-CVI for the ATDEA was .90, which met the lowest threshold of S-CVI, suggesting the validity for use of the overall assessment tool. Therefore, ATDEA should be applicable not only in Asia but also in other parts of the world. Because of the differences in the psychosomatic state of the abused older adult and in family situations, the subtypes and severity of abuse experienced by older adults may vary considerably. Consequently, it is important to provide prevention and support measures that accord with the unique abuse experiences of older adults and their families. However, with the exception of the IOA, almost no screening and assessment tool has been developed through rigorous research that is able to assess severity. Consequently, the authors established five levels for the ATDEA to assess the severity of elder abuse.

Clinical Applications in Nursing Practice
Nurses currently tend to limit their assessments to physical abuse only, which is typically more discernable and diagnosable than other subtypes of abuse. Because ATDEA holistically contains seven subtypes of elder abuse, it enables nurses to make appropriate assessments of such subtypes as psychological abuse and self-neglect that would otherwise be difficult to recognize. Thus, ATDEA may also prove useful for the early detection and prevention of elder abuse. Other professionals, in addition to nurses, may use ATDEA because elder abuse is encountered in fields other than nursing care. The authors recommend using ATDEA as a checklist in cases when medical professionals suspect elder abuse. As five levels have been set in the ATDEA, when medical professionals check Level 1 = nonexistent, it represents their assessment of no evidence of elder abuse. Checking Levels 2Y5 indicates evidence of elder abuse at different levels of severity, with higher levels associated with greater urgency for countermeasures.

Limitations of the Study
Both validity and reliability should be evaluated to determine the effectiveness of the ATDEA. However, because of difficulties in recruiting nurses who deal with elder abuse, this study did not examine reliability. Although the six items pertinent to self-neglect that earned I-CVI scores greater than .70 were included in the ATDEA, the difficulties encountered in establishing the validity of self-neglect indicate the need for further study of this issue by knowledgeable individuals. Because ATDEA was developed as a checklist rather than as a scale associated with a total score, construct validity was not examined. Finally, as participants were recruited in only one city ward, their perspective may not fully reflect that of the general population.