TABLE A-2Elder Mistreatment Studies

StudyMethodsSelected Findings
Childs et al. (2000)Design: Descriptive
Measure: (1) SVWS; (2) EAA BIS-R
Sample: Nonrandom: 422 young and 201 middle-aged adults
Theory: N/A
Middle-aged respondents viewed psychological behavior more harshly than younger respondent.s Both middle-aged women and young men were less tolerant of middle-aged perpetrators.
Data support relativistic nature of elder abuse.
Coyne et al. (1993)Design: Descriptive survey
Measure: Demographics; Zarit Burden interview; Zung Self-Rating Depression Scale
Sample: 1,000 caregivers who called a telephone help line for dementia; 342 respondents
Mean age of caregiver 56.1; 54.5% were adult children caring for parents; 37.1% caring for spouses; 8.4% cared for other relatives. 11.9% reported they had been physically abusive toward dementia patients.
Abusers had been providing care for more years; patients functioned at a lower level; caregivers had higher burden and depression scores.
Dyer et al. (2000)Design: Case-control study
Intervention: Comprehensive geriatric assessment
Measure: Standard geriatric assessment tools
Sample: 47 older persons referred for neglect and 97 referred for other reasons
45 cases of abuse or neglect identified.
37 were self-neglect.
Elder mistreatment cases were more likely to be white and male.
Higher prevalence of depression and dementia.
Ertem et al. (2000)Design: Descriptive
Method: Meta-analysis
Sample: 10 studies
10 studies: 4 cohort, 1 cross-sectional, and 5 case-control.
The RR of maltreatment in children of abused parents were significantly increased in 4 studies (RR 4.75-37.8).
In 3 other studies the RR was less than 2.
Significant validity issues.
Fulmer and Gurland (1996)Design: Descriptive
Measure: CTS, FRS and NMAP Survey, Beck Depression Scale, BDBS
Sample: 125 elder-caregiver dyads; 51 dyads with cognitive impairment and 74 dyads with no cognitive impairment; mean age of the elder 78 years
Theory: Risk and vulnerability
Cognitive impairment risk factor for elder mistreatment.
CTS higher for CI patients.
FRS higher for CI patients.
CI patients more dependent.
CI patients had higher BDBS.
CI patients had higher Zarit Burden scores.
Fulmer et al. (1999)Design: Descriptive
Method: Analysis of a probability sample of ADHC clients in New York State. Social workers served as informants.
Sample: 9 sites drawn through random sampling
Prevalence of elder mistreatment 12.3%.
Apprehensive behavior was highest reported behavior; with this item removed, prevalence 3.6%.
Social workers noted concern regarding elders who appeared frightened in the presence of their home caregiver.
Fulmer et al. (2000)Design: Descriptive
Measure: EAI, MMSE
Sample: 180 emergency department patients over the age of 70 with MMSE of 18 or greater
36 patients eligible for study.
7 patients screened positive for neglect.
Nurses were able to screen for elder neglect with greater than 70% accuracy; true positive 71%, false positive 7%.
Huber et al. (2001)Design: Descriptive
Method: Analysis of cross-sectional 6-state ombudsman database
Sample: 23,787 complaints
5 most frequent complaints were (1) loss of dignity and respect; (2) accidents; (3) physical abuse; (4) call lights unanswered; (5) poor personal hygiene.
Race and gender differences noted.
Hudson (1991)Design: Descriptive
Measure: 3-round Delphi survey
Sample: 63 elder mistreatment experts
Agreement on a 5-level taxonomy.
11 theoretical definitions proposed by panel.
Hwalek et al. (1996)Design: Descriptive
Method: Database analysis
Measure: Risk of Future Abuse instrument
Sample: State of Illinois Abuse, Neglect and Exploitation Tracking System; 2,577 cases from October 1989 to December 1991. 552 substantiated reports used for this study
73% of victims were women.
Mean age 77 (60-99). Caucasian 73%; widowed 54%; living at home 76%.
Caregiver substance abuse more likely to involve physical or emotional abuse.
Jogerst et al. (2000)Design: Descriptive
Method: Analysis of county-level data between 1984 and 1993 to test association between county characteristics and rates of elder abuse
Sample: 99 counties in Iowa
Analysis: univariate correlational analysis and stagewise linear regression
Community characteristics that had a positive association with rates of reported or substantiated elder mistreatment were:
  1. population density;
  2. children in poverty;
  3. reported child abuse.
Jones et al. (1997)Design: Descriptive
Method: Random sample survey
Sample: 3,000 members of the American College of Emergency Physicians; 705 completed surveys (response rate 24%)
52% of respondents described elder mistreatment as prevalent but less than spouse or child abuse.
Respondents evaluated a mean of 4 ± 8 suspected cases of elder mistreatment in the last 12 months; 50% were reported.
Lachs et al. (1994)Design: Prospective cohort study
Method: Case matching with adult protective services database
Sample: 329 elders investigated in 1985 and 1986
Analysis: Relative risk calculations
68 (2.4%) of database cohort members received ombudsman investigation.
Risk factors for elder mistreatment investigation using logistic regression included requiring assistance with feeding OR 3.5, being a minority elder OR 2.3, over age 75 at cohort inception OR 1.9, and poor social networks OR 1.7.
Lachs et al. (1997a)Design: Prospective cohort study
Method: Case matching with adult protective services database
Sample: 184 cohort members
Analysis: Pooled logistic regression
47 cohort members were seen for elder mistreatment (prevalence 1.6%).
Age, race, poverty, functional disability, and cognitive impairment were identified as risk factors for reported elder mistreatment, with ORs reported.
The onset of new cognitive impairment was also associated with abuse and neglect.
The influence of race and poverty is likely to be overestimated due to reporting bias.
Lachs et al. (1997b)Design: Prospective cohort study
Method: 7-year longitudinal database with identification of 182 victims of elder abuse
Sample: 114 elders seen in 2 emergency departments
114 individuals accounted for 628 visits (median 3, range 1-46).
30.6% resulted in hospital admission.
66% had at least one visit that resulted in an injury-related chief complaint.
Lachs et al. (1998)Design: Prospective cohort study
Measure: mortality among elders for whom protective services were used to corroborate mistreatment and elderly persons for whom protective services were used for self-neglect
Sample: 176 adult protecti,ve services elders
Cohort members seen for elder mistreatment at any time during follow up had poorer survival (9%) than others.
Reported and corroborated elder mistreatment and self-neglect are associated with shorter survival after adjusting for other factors associated with increased mortality in older adults.
Moody et al. (2000)Design: Descriptive
Measure: H-S/EAST
Sample: 100 black, Hispanic, and white elders living in public housing
Principal components FA of 15-item instrument supported the 3-factor structure for a total of 10 items explaining 38% of the variance.
Discriminant function analysis showed that 6 items were as effective as the 9-item model in classifying cases as abused (71.4%).
National Center on Elder Abuse at the American Public Human Services Association [formerly American Public Welfare Association] in collaboration with Westat, Inc. (1998)Design: Descriptive study
Method: Incidence study using sentinel agency reports
Sample: 20 counties in 15 states: nationally representative sample
551,000 elder mistreatment cases in 1996.
Female elders are abused at higher rates than males.
The oldest elders (80 years and older) are abused and neglected at 2-3 times their proportion in the elderly population.
In almost 90% of elder mistreatment cases, the perpetrator is a family member and 2/3 are adult children or spouses.
Victims of self-neglect are usually depressed, confused, or extremely frail.
O'Malley et al. (1984)Design: Descriptive
Measure: Case analysis using OARS
Sample: 24 cases from primary care clinic
Cases divided into three categories: (1) extremely impaired who receive care from individuals responsible for abuse and neglect (N = 4); (2) impaired elders who receive inadequate or intermittent care (N = 9); (3) involved independent elders whose only care needs resulted from threats or violence from relatives (N = 11).
Paveza et al. (1992)Design: Descriptive
Measure: CTS
Sample: Purposive sample from Alzheimer's disease registry: 184 patients
Severe family violence as measured by the CTS was a significant problem: overall prevalence 17.4%.
15.8% of patients had been violent since diagnosis.
5.4% of caregivers reported being violent toward the patient.
Violence by the Alzheimer's disease patient against the caregiver was serious problem.
Pavlik et al. (2001)Design: Descriptive
Method: Analysis of Texas Department of Protective and Regulatory Services, Adult Protective Services
Sample: 62,258 allegations of elder mistreatment in 1997
Neglect accounted for 80% of allegations.
The incidence of being reported to adult protective services increased sharply after age 65.
Prevalence was 1,310 over 100,000 ≥ 65 years of age.
Phillips and Rempusheski (1985)Design: Descriptive
Method: Interviews with grounded theory analysis
Sample: 29 health care providers (16 nurses and 13 social workers)
4-stage model describing decisions of health care providers about elder abuse. Model identifies 3 types of decisions: diagnostic, value, and intervention.
Complexity of decision processes is revealed via 5 pathways.
Phillips et al. (1990a)Design: Adaptation of the QUALPACS
Method: Instrument development
Sample: Piloted with 8 data collectors (4 in each of 2 sites) who interviewed 4 elder-caregiver dyads. A total of 29 elder-caregiver dyads were interviewed
QUALCARE Scale contains six subscales and 53 items. Included in 6 subscales: environmental, physical, medical maintenance, psychological, human rights, and financial.
Phillips et al. (1990b)Design: Descriptive correlational study
Measure: QUALCARE
Sample: Convenient sample of 249 elder-caregiver dyads
Interrater reliability for 55 observations ranged from 79% to 88%.
Internal consistency: alpha = .097.
Conceptual structure: confirmatory factor analysis indicated 6 significant factors accounting for 64.4% of the variance.
Criterion validity: all correlations between criteria variables and QUALCARE were in correct direction and p ≤ 0.05 level.
Construct validity: 8 of 9 correlations in the predicted direction.
Pillemer and Finkelhor (1988)Design: Descriptive
Method: Stratified random sample survey
Sample: 2,020 community-dwelling elders in metropolitan Boston
63 elder persons were maltreated.
Rate of 32 per 1,000.
95% confidence interval of 25-39 per 1,000.
No minority differences or age differences.
Those in poor health were 3 to 4 times likely to be abused.
Males were more likely to be abused than females.
Pillemer and Finkelhor (1989)Design: Descriptive
Method: Case control
Sample: 46 abuse or neglect victims and 215 random controls
Factors associated with elder mistreatment included abuser factors of deviance, dependence on victim, and life stress.
Victim factors included court help, disability, dependence on abuser, and conflictual relationship (spouse only).
Pillemer and Moore (1989)Design: Descriptive
Measure: CTS
Sample: 577 nursing personnel from 31 nursing homes in New Hampshire
36% of the sample had seen at least one incident of physical abuse in the preceding year.
Most frequent abuse observed was excessive restraint.
Second most frequent type was physical abuse.
81% observed at least one psychologically abusive incident in the preceding year.
10% of respondents reported committing physical abuse.
40% of respondents reported committing psychological abuse.
Pillemer and Suitor (1992)Design: Descriptive
Method: Analysis of quantitative and qualitative data
Sample: 236 family caregivers for dementia victims
Characteristics predictive of violent feelings in caregivers included physical aggression by elder, disruptive behaviors, and a shared living situation.
Structural relationship and caregiver age were related to actual violence: spouses were more likely to be violent than other relatives, as were older individuals.
Violence by elder was positively related to caregiver violence.
Rosenblatt et al. (1996)Design: Descriptive
Method: Analysis of State of Michigan records of reported cases of suspected elder abuse 1989–1993
Sample: 27,371 cases of possible elder mistreatment
17,238 of cases were older than age 65.
Physicians reported only 2% of cases.
Physician reporting rates did not increase over a 5-year period.
Shaw (1998)Design: Descriptive
Method: Grounded theory
Sample: 21 semistructured interviews conducted with six abuse investigators and 15 nursing home staff
The two types of abusive nursing home staff were identified as reactive and sadistic.
Wolf (1986)Design: Descriptive
Method: Analysis of cases from an elder mistreatment intervention project
Sample: 59 elder mistreatment cases compared with 49 cases randomly selected from a nonabuse caseload
Victims and nonabuse clients were similar in age, sex, and health status.
Caretakers for both groups were similar in age and health status.
More perpetrators were males.
A majority of elder mistreatment cases resided with family members versus nonabused persons living alone.
Victims and perpetrators had more psychological and emotional health problems.
Abused elders did not appear to be more dependent.
Wolf and Pillemer (1997)Design: Descriptive
Measure: ADLs, IADLs, CTS
Sample: 73 older women: 22 victimized by husbands and 51 victimized by adult children
Wives more likely to be dependent on husbands for IADLs.
Adult children more likely to be dependent on mothers for housing and finances.
Husbands more likely to use physical violence against wives than adult children against mothers.
Wolf and Li (1999)Design: Descriptive
Measure: DV was number of reports per 1,000 persons age 60 years and older during 1994
Sample: 27 geographical areas in Massachusetts
Rate of reports varied from a low of 2.41 per 1,000 through 9.31 per 1,000.
Higher rates of reporting were associated with lower socioeconomic status, more community training, higher agency service rating scores, lower community agency relationship score.

SOURCE: Adapted from Fulmer (2002).

From: APPENDIX A, Elder Mistreatment Measures and Studies

Cover of Elder Mistreatment
Elder Mistreatment: Abuse, Neglect, and Exploitation in an Aging America.
National Research Council (US) Panel to Review Risk and Prevalence of Elder Abuse and Neglect; Bonnie RJ, Wallace RB, editors.
Washington (DC): National Academies Press (US); 2003.
Copyright © 2003, National Academy of Sciences.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.