Recruitment and participant characteristics – staff.

The 13 studies that recruited staff [26,32,33,37,38,43,46–49,51,53,54] did so by selecting institutes or homes ranging from 2 [43] to 1,600 [47]. Of those who recruited staff cohorts ranged from 53 [46] to 7,000 [37,38]. Twelve studies reported a response rate, ranging from 15% [43] to 92% [26] (Table 1). Staff participant characteristics were reported in eleven studies; ten reported the majority being female (>82%) [26,37,38,43,46–49,51,53,54]; with an age range between 16 to 74 in seven studies, [53] and 13.7 mean years experienced as reported in five studies [26,46,51,53,54]. Current nursing home experience ranging from 1.1 [36] to 8 years [53]. Of the thirteen studies that utilized reported abuse by staff, four also collected data on one or more resident characteristics [43,47,49,53] (Table 1).

Recruitment and participant characteristics – residents.

Three studies [44,49,52] examined older adult abuse as reported by residents. One study selected 10 ‘nursing homes with 640 eligible ‘nursing home residents invited 200 to participate, with 82% doing so [52], while another study selected 24 ‘senior homes’, screened 1,807 ‘clients’, with 27% meeting the study criteria [49]. The third study collected data from 71 admitted ‘inpatients’ from ‘elderly homes’ or ‘nursing homes’ [44]. No response rates were recorded. The majority of the participants were female (>65%) [44,49,52] with a mean age of 82 (±7.5) (range: 60 – 99) [44,49,52]. Activities of Daily Living (ADL) was used to report resident dependency levels and years of residency was reported in one study as ‘length of stay’, with two-thirds reporting less than four years [49]. No staff characteristics were reported in the studies recruiting residents (Table 1).

Recruitment and participant characteristics – relatives.

Relatives were recruited in a telephone survey [21]. There were 450 participants from the general community who had ‘a family member’ ‘receiving long-term care services’. No response rate was recorded. The only study characteristics collected were description of residents, as the ‘family member’ being female (73%). No staff characteristics were reported (Table 1).

Recruitment and participant characteristics – registry.

Six studies utilized existing registries to report institutional abuse examining sexual abuse over a six-month period [41,42], complaints of neglect over an eight-year period [40] or as an admitted patient due to neglect over a five-year period [39]. The two studies examined forensic medical reports for incidences of ‘female sexual abuse’ [45] or ‘abuse’ [50] from ‘nursing homes’ or ‘institutional settings’ over a 10 [50] to 15 [45] year period resulting in small prevalent cohorts, ranging from 28 [45] to 59 [50], respectively (Table 1).

Of the six studies, three reported a mean age of 79.7 years [39,41,50]. Resident’s dependency was reported as an ADL status, ‘being mildly to totally dependent’(62%) [41], a category [‘mild/moderated’ or ‘severe’ (‘loss of autonomy’ [highest, severe 56%])] [50], a gradient [‘ability to function independently’ 0 – 10 scale (aggregate mean 4.8 [Barthel Index])] [39] or as ‘dependency or require assistance with ADLs’ (18%)] [45]. Years of residency was only reported in one study, at the time of registered abuse, with a median of 17 months [49] (Table 1).

Methodology instruments used to measure older aged abuse.

Overall, we identified 16 instruments used to measure older aged abuse in long term institutes over the last 18 years. In Table 1, we constructed columns examining the methodological approaches, based on who reported the abuse (highest contact with resident, staff to lowest registries) with the following methodological features such as 1) method of administering data (such as mode, distribution and collection) 2) number of items and name of the identified instruments [including source], 3) observed and/or experienced abuse with type of abuse reported 4) recall period and 5) validity of tool. In the S3 Table, we have provided a more in-depth analysis to examine the homogeneity of the commonly defined types of abuse (as defined by WHO).

Instruments used to measure older age abuse - as reported by staff, and/or residents.

Of the 16 instruments, 11 were used to measure abuse as observed and/or experienced by the staff member, resident or relatives. The review found the three most commonly used tools measured staff abuse; Malmedal et al’s 42 acts of inadequate care instrument (2009) [47,52,53]; Castle’s (2012) 28-item questionnaire measuring how often staff observed and/or perpetrated abuse [37,38,48] and Drennan et al.(2012) [50,53,54] national survey on interactions and conflicts within nursing home settings. There were variations among these instruments ranging from modes of delivery, either presented as a questionnaire or survey, with differences in definition and types of abuse and discrepancies in recall periods.

Malmedal et al.(2009) [53] original 42-item questionnaire was used to measure staff ‘observed’ or ‘committed’ ‘physical, psychological (emotional), financial and neglect’ acts (unintentional and intentional acts) of inadequate care’ within a four-week recall period, using a four-point Likert-type frequency scale ranging from ‘never’ to ‘more than once a week’ was used in two other studies. This questionnaire has been tested for face validity only, indicating the tool was easy to follow and comprehensive, evidence pertaining to the other items of validity were not reported.

Habjanič and Lahe (2012) [52] further modified Malmedal’s (2009) [53] question which asked ‘residents’ face-to-face if they had ‘ever experienced’ ‘mental’, ‘physical’ and ‘financial abuse’ using a ‘six months’ recall period to record the ‘number of incidences’, rather than using the Malmedal et al.(2009) [53] Likert scale (Table 1). In 2017, Blumenfeld Arens et al.[47] used Malmedal et al.(2009) [53] to ask staff if they only ‘observed’ ‘elder abuse’ (not as the original definition, ‘inadequate acts’ of ‘physical, psychological (emotional) and neglect, but not financial abuse, using with the same recall period and Likert-type frequency scale. None of the studies measured correlation ecoefficiency.

Castle and Beach’s (2013) [38] 46-item questionnaire measured the ‘number’ of times staff ‘observed’ ‘physical’, ‘psychological (verbal)’, ‘financial (material exploitation)’ or ‘sexual abuse’ in ‘last three months of their prior place of employment’ and was used again in a study by Castle the following year [37]. The questionnaire has been tested for face and content validity (using the Fleisch–Kinkaid Scale), indicating this tool is measuring the degree which abuse is measuring abuse accurately. Recently, Botngård et al.(2020) [48] changed Castle’s questionnaire to examine staff’s incidences of ‘observed’ and ‘perpetrated’ abuse, measuring additional types of abuse, ‘overall’ and ‘neglect’, within their ‘current place of employment’ in the ‘last 12 months’ All studies utilized the original frequency using a three-point Likert-type scale as ‘never’,’once’ or ‘repeatedly’ (Table 1). None of the studies reported on correlation ecoefficiency.

Gil & Capelas (2022) [51], and Neuberg (2017) [54] utilised the long-established questionnaire by Drennan et al.(2012) [55], a 25-item national survey of staff-resident interactions and conflicts within residential care settings. Between the two papers, there were variations with types of abuse and whether it measured witnessed [54] and/or committed [51] abuse. In Neuberg et al. [54] study, the survey was pretested in a validation pilot study and achieved a reliability coefficient was >  0.7, deeming the instrument to be reliable. Overall, there is still some heterogeneity among these instruments, they are still in their early constructs, more studies and methodology testing are required conducted to validate these instruments. See Table 1, for further details.

Instruments using data registries to measure older age abuse.

The remaining six studies utilized government registries or databases. Four studies utilized existing government registries such as the Registry Arizona Secretary of State & Arizona Department of Health Service offices [40] or the Adult Protective Services (APS) (National Adult Protective Services Association [NAPSA], 2021) (3) in conjunction with a survey (Sex Abuse Survey [SASU]) [41] and/or with hospital records with the use of Clinical Signs of Neglect Scale (CSNS) [39,41,42] to report ‘isolated or ongoing’ investigation of ‘citations and allegations’ [40] or a ‘suspected, reported, unsatisfactory, partial or substantiated resolution case of abuse’ [41,45,50] or used the to identify ‘clinical signs of elder mistreatment or elder neglect’ [39]. While two studies utilized clinical forensic medicine reports [45,50] of ‘current or past medical observations and/or victim complaints of suspicion of physical or psychological abuse’ [50] or ‘alleged incidence of sexual assault among women only’ to report incidences of abuse [45]. These studies varied with recall periods ranging from six months [41,42] to 15 years [45]. Some of these studies required validated professional staff to perform examinations [50], a consortium of experts to develop clinical validated scales [39,41,42] or independent research reviewers [44] ensuring reliability, validity and reliability of findings, however, registers are commonly known for their practical limitations such as incompleteness or inaccuracy with data collected and difficulties with data dredging (Table 1).

All abuse.

Out of the 22 studies, ten studies measured the overall incidence of abuse (measuring one or more types of abuse as defined by WHO) [26,32,33,43,44,48–51,53] (Table 1), with the highest overall prevalence reported over a four week period reported in one study, 91% ‘observed’ abuse by staff, while ‘committed’ abuse by staff was at 87% [53]. Two studies reported abuse by staff over a 12-month period resulted in lower rates of ‘observed’ abuse ranging from 55% [51] to 76% [48] and for ‘perpetrating’, from 54% [26]to 60% [48].

Two studies reported by ‘residents’ overall abuse over a 12-month period, retained lower rates, than above. ‘Experienced’ abuse ranged from 11% [49]to 31% [44], while ‘observed’ was at a lower rate of 5% in one study [49]. No studies examined overall abuse reported by relatives or the community via a registry. Five studies (23%) reported all five types of abuse as defined by WHO [21,26,44,48,51] (See Table 1). The following sections will examine prevalence based on types of abuse as defined by WHO, physical, psychological, financial, sexual and neglect.

Physical abuse.

The most commonly measured form of abuse was physical abuse, also defined as ‘physical violence’ [26], ‘mistreatment [21], ‘maltreatment’ [44] or ‘acts of physical character’ [53], measured in 15 studies [21,26,32,37,38,43,44,47–54]. An accumulation of 81 items were identified to describe acts of physical abuse, with each study using three [47] to 11 [44] items to describe physical abuse. The most commonly used verbs to describe physical abuse was ‘hitting’ (8) [21,37,44,48,49,52] or ‘kicked’ (7) [21,37,48,49,52], with variations in definition, recall periods and persons reported. One study relied on staff to define physical abuse [32] or did not disclose items measured [33] (S3 Table: Definitions of types of abuse).

The highest rate of physical abuse reported was ‘witnessed’ by staff (44%), in the act of ‘restraining/hold back a resident’ ‘over a recall period of four weeks’ [53], and the highest ‘committed’ abuse was 33% from the same act of ‘restraining/hold back a resident’ as reported in the same study. When the same questionnaire was used in Blumenfeld Arens et al. [47] in 2017, the study questionnaire, measured witness physical abuse over a 4-week period, resulting in a lower rate of 1.4%.

Studies examining physical abuse over a 12-month period, Gil and Capelas (2022) [51] and Neuberg et al. (2017) [54] using the same questionnaire [54], resulted in different levels of physical abuse by staff. Neuberg et al. (2017) [53] reported over 12 months, 42% of staff observed ‘force feeding the resident’ in the last 12 months, whereas Gils and Capelas (2022) [51] recorded 14% observed staff committing ‘at least 1 of the 6 behaviours of physical abuse’. The remaining studies utilised various measurement tools, with the same recall period of 12-months resulting in observed rates ranging from 6% [32] to 30% [49], while committed abuse were even lower ranging from 1.7% [32] to 12.3% [26].

Three studies reported physical abuse reported by residents either in the last six [52] to 12 months [44] resulted in lower rates of ‘observed’ abuse, from 1% [49] to 2% [49] and 8% for ‘experienced’ abuse [44]. Cohen’s study [44] found only three residents attained a score of three or more on the signs of physical scale. Compared to this, relatives reporting abuse in telephone interviews in the last 12 months had higher rates of abuse at 74% [21], while physical signs or evidence of physical abuse from forensic medical reports (FMR) from registries, were lower at 55 cases over a 10-year period [49], however these tend to be extreme cases of abuse (S3 Table).

Psychological abuse.

The second most common measured form of older age abuse in long term institutes was psychological abuse. Fourteen studies [26,32,37,38,43,44,47–49,51–54] addressed psychological abuse. Three studies defined this type of abuse as ‘psychological abuse’ [32,44,48,49,51] while the remaining six defined as ‘emotional’ [21,43,47,52], ‘mental abuse’ [26,53] or as a combination of ‘psychological and verbal abuse’ [37] or ‘emotional or psychological and verbal mistreatment’ [21].

There were in total 47 items, with each study using three [47]to 14 [49] items to classify psychological abuse. The most common terms used to describe psychological abuse were of ‘intent’(5) [49] or ‘threat’(4) [37,49,52,53]. A total of four studies did not disclose items or descriptions of types or examples of abuse asked [26,43,54] (S3 Table).

Psychological abuse was reported by staff (11) [26,32,37,38,43,47–49,51,53,54], residents (2) [44,52] and relatives (1) [21]. The highest rates of ‘observed’ and ‘committed’ acts of psychological abuse was ‘entering a room without knocking’ of abuse by staff, 64% of staff committed the act, while 84% observed other staff in the last four weeks [53]. Three studies examining psychological abuse ‘committed’ by staff over the last 12 months found higher incidents ranging from 23% [26] to 46% [49], with variations in instruments utilized to measure this form of abuse. ‘Observed’ abuse by staff was reported in five studies, with incidents ranging from 30% [49] to 62% [21], with variation in instruments used making it difficult to provide an average rate. Two studies utilised Dennan et al. [54] instrument however there was a 20% difference between the act of shouting at resident in anger [50] (33% [50] from 16 care home settings versus 55% from the nursing home and extended care units [54]).

Residents reported ‘experienced’ psychological abuse ranged from 10% over a 12-month period [49] to 56% [52] over a six-month period, however, reported ‘observed’ abuse was lower at 4% [49]. Uniquely, Cohen et al. (2010) [44] reported distribution of disclosed abuse and found “very low complaints for psychological abuse” (13%). Telephone interviews among family members (relatives) reported 84% ‘observed’ ‘verbal mistreatment’ by nursing staff ‘in the previous year’ [21]. No studies measuring psychological abuse used registries. All studies examined specific psychological acts, making it difficult to aggregate the incident rate due to variations as shown above (S3 Table).

Financial abuse.

Eleven studies defined financial abuse either as ‘material exploitation’ [21,37,38,48] and/or ‘financial exploitation’ [26,44], ‘financial abuse’ [32,51,52], ‘acts of financial character’ [21,48,53]. An accumulation of 24 items was identified to describe acts of financial abuse, with each study using one [53] to seven [44] items to describe financial abuse. Most common term used to describe financial abuse was ‘signing documents’ (6) [37,44,48,52] (S3 Table).

Most of the studies examining the rates of financial abuse were reported by staff (8) [26,32,37,38,43,48,51,53], followed by residents (2) [44,52]or relatives (1) [21]. The highest level of financial abuse reported in this review were observations of staff from relatives of older adults residing in nursing homes, 71.9% [21]. This was followed by reported ‘experienced’ financial abuse by residents in one study, at 32.8% [52] over the last 6 months. Lower rates of financial abuse were reported by staff, for ‘observed’ incidents ranged from 2.1% [48] to 3.3% [51] in care and nursing homes.

Staff reporting ‘committing’ financial abuse were at a lower rate 0% [53] to < 1% [26,48] over the last four weeks to 12 months, while two studies examined staff ‘observed’ financial abuse found 10%, of staff took ‘assets’ from nursing home residents or ‘destroying belongings’ of resident residing in assisted living institutes, 26% [38]. Interestingly Castle’s questionnaire used in two studies, in two similar settings and recall periods, found the incidents from ‘taking residents assets’ were similar, 10% [37] versus 11% [37]. No studies measuring financial abuse used registries (S3 Table).

Sexual abuse.

Eleven studies reported the prevalence of sexual abuse, described either as abuse [21,26,32,38,41–46,48,49,51] with variation in definition of this form of abuse ranging from as an act of ‘assault’ [45], ‘misconduct’ [21], ‘violence’ [26], ‘unlawful or unwelcome sexual behaviour’[46] or ‘sexual nature without consent’ [51], to an outcome of signs of ‘forensic evidence’ [42] or ‘victimization (women)’ [41].

Number of items describing sexual abuse ranged from one [51] to 11 items [41]. Among the eleven studies, in total were 34 items that identified abuse including, as an act of exposure to (4) [37,41,42] (hands off) to oral-genital contact (3) [37,41,42] (hands on). Evidence of signs of sexual abuse included a torn underwear to infected [44].

Most of the studies relied on reports by staff (4) [26,37,48,49], or registries (3) [41,42,46], followed by direct reporting from residents (2) [44,49]and one by relatives [21]. Reports from relatives had the highest reported level of sexual abuse at 40% [21]. Registries reporting an incidence of sexual abuse performed by staff ranged from 15.6% to 25% [41,42,45] however these cases were over a ten-to-15-year period. The lowest reported incidences of this type of abuse were reported by staff as ‘observed’ resulted in ≤  7% or ‘committed’ < 1% [26,32,37,38,43,48,49,51], while two studies reporting no sexual abuse reported by residents [44,49]. There were two studies showing some consistency with findings, utilising the same questionnaire in different institutionalised settings, found staff observed 69 nursing home staff and 61 assisted living staff ‘exposed private body parts to embarrass resident’ in the last three months [37,38].

Neglect.

Similar to psychological abuse studies, neglect is equally the second highest form of abuse investigated in this review among older adults residing in long term institutions [21,26,32,39,40,43,44,47–51,53,54].

The definition of neglect varied with 3 [40] to 11 items [39] describing these acts from ‘physical and mental neglect’ [26], to ‘clinical sign of neglect’ [39,44,50], or collectively categorised as ‘personal, environmental, medical’ [40] to specific items described care neglect such as ‘not changing the position of bedridden person’ or ‘ignoring resident when they called’ [49,51,54]. Only two studies utilised the same instrument to measure neglect [37,38]. Four studies did not provide or specify items that were measured for this type of abuse [26,43,54].

Neglect ‘observed’ by relatives retains the highest rate of failure to provide basic needs to residents (86.9%) [21]. Four studies reported neglect ‘committed’ by staff over the last 12 months, with results varied from 1% [49] to 46.9% [48], compared to nine studies reporting ‘observed’ acts of neglect ranging from 9% [49] to 57.8% [48]. These variations are due to different instruments and definitions used to measure neglect. Surprisingly, four studies used the same instruments, however disseminated findings differently, with one study reporting if ‘observed’ or ‘committed’ one of the ten items listed for neglect, while the other reported 10 items distinctly with respected incident rates [47,51,53,54].

The highest prevalence of neglect was 24%, attained from the face-to-face interviews conducted by hospital staff [44] among inpatient residents, while another study when interviewing residents on ‘observed’ or ‘experienced’ neglect conducted in facilities were ‘unmentioned’ [49]. Registries reported 20% of severe cases of neglect, however again, this was over a 10 or more-year period [39,50], while another study reported a total of 1,196 total neglect allegations, with 535 substantiated, over an eight year period, making it difficult synthesis findings [39]. Other abuse items not classified by WHO are also included in the S3 Table.