Each year in the United States, between 1 and 2 mil-lion elderly adults experience abuse, including
physi-cal, psychologiphysi-cal, financial/material abuse, as well as
violation of personal rights and neglect.1 Indeed, 9% of older adults report verbal mistreatment, 3.5% financial
mistreatment, and 0.2% physical mistreatment by a family member. African Americans may be at higher
risk for financial mistreatment and Latinos at lower
risk for verbal abuse.2
Elder abuse has been acknowledged as a significant
component of family violence in the United States, but most cases of elder abuse are not reported to any adult protective service agency.3 Since abused patients
are often assessed by their primary care physicians, opportunity exists for the detection of elder abuse in a clinic setting. Yet despite the almost universal mandate
for physicians to report elder abuse, physicians account for only 2% of all reporting agents.4,5 When surveyed,
physicians express a lack of understanding of
report-ing mechanisms as the most significant obstacle.6 In
addition, physicians may also be unable to recognize key risk factors for abuse.5,7,8 As a result, some surveys
suggest that more than half of physicians have never
identified a case of elder abuse, and more than 60% of
clinicians do not ask their elderly patients about abuse.8
Although screening for elder abuse has been recom-mended,9 and screening tools have been developed to
assess elder abuse, these tools have not been universally reliable or effective in terms of improving elder abuse outcome.10
The Accreditation Council for Graduate Medical Education (ACGME) does not clearly delineate the nature and degree of exposure to elder abuse expected during residency training in specialties such as fam-ily medicine or internal medicine.11 Current residency
training educational efforts have included formal lectures and Web-based resources as well as clinical exposure to elder abuse issues.12,13
Elder Abuse Education in Primary Care Residency
Programs: A Cluster Group Analysis
Deborah B. Wagenaar, DO, MS; Rachel Rosenbaum, OMS III, MS, RD;
Sandra Herman, PhD; Connie Page, PhD
From the Department of Psychiatry (Dr Wagenaar), College of Osteopathic Medicine (Ms Rosenbaum), and Center for Statistical Training and Consult-ing (Drs Herman and Page), Michigan State University.
Background and Objectives: Elder abuse is a serious issue, affecting up to 10% of community-dwelling older adults. This project sought to understand if elder abuse under-reporting was related to physi-cian specialty and residency training. Methods: A 17-item survey focusing on program demograph-ics, current curriculum, priority of elder abuse education, clinical and didactic experiences, and triage resources was mailed to directors of all Michigan primary care residencies approved by the Accreditation Council for Graduate Medical Education. Results: Clustering by program type, family medicine programs were primarily represented in the high-intensity group (standard residual -1.6), while internal medicine programs were primarily represented in the low-intensity group (standard residual 2.5, X2=14.80, df=6). Emergency medicine comprised the mid-intensity groups. The high-
intensity group scored very high or high on areas including knowledge of elder abuse facts, elder abuse curriculum and training, and awareness of community resources related to elder abuse.
Conclusions: Family medicine residency programs appear to address elder abuse more comprehen-sively than do emergency medicine and internal medicine programs. These programs may need to consider additional curriculum and clinical experience in elder abuse.
Is physician education the answer to heighten awareness and help with elder abuse reporting? If so,
are there specific physician groups in need of educa -tional, screening, or triage assistance in dealing with elder abuse patients? In this study, a cluster group analysis methodology was chosen to identify groups of Michigan primary care residency programs that had similar degrees of intensity of elder abuse training in the residency curriculum. The intensity of elder abuse
education was defined in terms of hours of didactic
and clinical training and the number of topics included
related to elder abuse definition and identification, re -porting mechanisms, and community supports.
Methods
The survey instrument and research methods for this project were approved by the Michigan State University Institutional Review Board.
Survey Development
To assist with development of the survey tool, a focus group was developed to offer content expertise and assess survey items. Members represented medicine, law, social work, education, and statistics. The focus group members reviewed both the elder abuse literature and the education standards established for primary care specialties by the ACGME. Based on this review, the focus group developed survey questions about the programs to be surveyed and on elder abuse detection, barriers to education, and community resources related to elder abuse.
The final 17-item survey instrument included a defi -nition of elder abuse, followed by questions about the residency programs to be surveyed, including number of residents, resident gender, program specialty, and program location. Questions on the curriculum included
approach to learning (didactic/clinical/computer), cur -riculum content, lecture feasibility and attendance requirements, content role in the overall residency cur-riculum, and awareness of elder abuse resources in the community. Additional questions on the need and type of residency materials program directors would desire
was included. Clinical contact was the final component
of survey questions and included items on the frequency and severity of elder abuse in the clinic population, elder abuse screening activities, and directors’ experience working with adult protective services.
Subjects and Survey Procedures
The 43 survey subjects were the directors (or their designees) of primary care residency programs in the
state of Michigan. We define primary care residencies
as including family medicine, internal medicine, emer-gency medicine, preventive medicine, and transitional-year programs. Mailing addresses and contact
infor-mation for the directors were identified using ACMGE
rosters. Programs were contacted to confirm mailing
information. Two mailing waves of the survey were completed between August and December 2006, and nonrespondents received phone and mail prompts.
Data Analysis
We used cluster analysis to identify homogenous subgroups of residency programs where the within-group variation was minimized and the between-within-group variation was maximized. The clustering variables included program type, requirement to attend elder abuse lectures, presence of a curriculum, and impor-tance of elder abuse in training. Clusters were formed using a hierarchical agglomerative approach in which each case was treated as a unique cluster; then the two cases that were most similar based on the clustering variables were combined to form a new cluster; a third
case was next compared to the first cluster and if it was more similar to the members of the first cluster than to any other case, it was joined to the first cluster. But,
if it was more similar to another case those two cases form a second cluster. This process was repeated until all cases were added to a cluster, and all the clusters
had been combined to form one final cluster.
Ward’s Method with squared Euclidean distances was used to determine the similarity of programs us-ing two trainus-ing intensity and three content coverage variables. In Ward’s Method, variation is minimized within clusters, and the between-cluster variation is maximized. Variables were transformed to z scores before the analysis.
A four-cluster solution was selected after examining the means of the cluster variables for solutions of three
to six clusters. Mean values on the five-cluster variables
were used as the centers for a four-cluster k-means clustering procedures with iterations. This produced
the final cluster assignments. Results
Demographics
Our survey response rate was 62.8% (27/43). Re -sponse rates by specialty, calculated as the number of surveys sent out by specialty divided by the number of surveys received by specialty, were as follows:
fam-ily medicine 11/18 or 61%; internal medicine 10/17 or 58.8%; emergency medicine 6/8 or 75%. Mean program
size was 25 (SD = +20.7) residents. Generally, for both family medicine and internal medicine, a larger number of smaller than larger programs responded, but the difference in average size of program for respondents
and nonrespondents was not statistically significant (t
value=0.87, P<.392, df=22). The majority of programs were community based (65%, n=26), followed by mixed
university/community programs (20%, n=8), and then
Cluster Descriptions
Figure 1 depicts the spatial relationship of the clusters based on training intensity (sum of z scores for the two training intensity variables) and content coverage (sum of z scores for the three content coverage variables), with clusters described by scores on each of the educational intensity variables. Scores were considered very low if the mean z scores were < -1, which means the score was one standard deviation below the mean of that variable, low if the mean z score was between -0.75 and >-0.99, medium low if the mean z score was between -0.26 and -0.75, average if the mean z score was between -.025 and +0.25, medium high if the mean z score was be-tween +0.26 and +0.75, high if the mean z-score was between +0.75 and +1.0, and very high if the mean
z score was > 1.0.
Low-intensity programs had the lowest scores on
all cluster variables and had the least intense training.
Medium-low programs were relatively low in covering elder abuse topics and were average in the intensity of their didactic and clinical programs. Medium-high intensity programs did well on elder abuse factors and reporting topics and were average on community re-source topics. They lacked intensity in clinical training and were average in terms of didactic training. Finally, high-intensity programs had the highest scores on most number of topics taught and the most intense elder abuse clinical training experiences.
The high-intensity group primarily represented family medicine while the low-intensity group pri-marily represented internal medicine (Fisher’s Exact Test=18.42, P<.01). Emergency medicine and transition-al and preventive medicine programs placed solely in the middle two clusters (medium-low and medium- high intensity) and were not present in either of the high- or low-intensity cluster extremes. Family medicine was Figure 1
Figure 2
Role of Elder Abuse Training
Figure 3
Screening for Elder Abuse
screened for elder abuse within hospital or clinic settings while 88.8% of low-intensity programs did not screen (Fisher’s Exact Test=8.74, P<.05).
Figure 3 demonstrates how screening behaviors related to cluster intensity. It appears that family med-icine programs use screen-ing tools more frequently than do internal medicine programs. Although the use of screening tools has not been standardized for clinical practices, their use may increase diagnostic vigilance or suspicion.
Lecture quality and
value appear to be differ-ent between family medi-cine (high intensity) and internal medicine (low intensity) groups. Nearly all of the programs in the high-intensity group re-quired elder abuse lecture attendance while none of the low-intensity groups requi red at tend ance (Fisher’s Exact Test=17.6,
P<.001). Figure 4 demon-strates lecture attendance requirements among the cluster groups. The mean quality of training lectures (1=poor, 4=excellent) was
significantly different for
clusters (F=6.78, P<.001). The low-intensity group differed significantly from the medium-high and high-intensity groups.
Lecture quality was rated
on the basis of being prac-tical, comprehensible and retainable.
Discussion
Results of this survey suggest that it is possible to identify specialties that are providing comprehensive elder abuse education as well as those that may need to provide additional training. Family medicine programs tend to use an approach that includes formal lectures and didactics as well as broad clinical exposure to el-characterized as having greater lecture topic variety,
more clinical experiences in elder abuse, and greater use of screening tools for practice. Family medicine programs placed a higher degree of importance on the topic of elder abuse in their curriculum than did internal medicine residency programs (Fisher’s Exact Test=7.63,
der abuse patients. Internal medicine programs in this survey demonstrated a lack of elder abuse education; additional resources targeted toward internal medicine residency programs may be warranted. Developing lectures and educational materials that seamlessly transition into internal medicine grand rounds or case conference format should be considered to address the lack of formal didactics. Examples of effective programs are available.13
Elderly patients who experience abuse are more likely to have worsened functional status and pro-gressive dependency and poorly rated self-health and increased social isolation, leading to a multitude of medical issues.14 As abused older patients often pres-ent in primary care offices, it is of concern that many
primary care programs lack education in elder abuse presentation or care. Reconsidering a requirement for
specific clinical experiences and didactic topics should
be considered for all primary care residency programs by the ACGME to bolster exposure to the problem of elder abuse. Requiring attendance for both faculty and residents at elder abuse lectures may emphasize this topic to young clinicians. In addition, educational op-portunities should be culturally sensitive to the needs of older adults from diverse cultural backgrounds and life experiences.
This survey also highlights the need to develop effective, reliable, and expedient screening tools for elder abuse. In this survey, family medicine programs reported using elder abuse screening tools more than did internal medicine, but there is no elder abuse screen-ing tool that has been shown to be valid and reliable in all primary care settings. Such a tool is needed.
Limitations
Our study has limitations that may influence the
interpretation of our results. First, the survey results are based on self-reported data, and recall bias may
have influenced our results. Second, our response rate
is modest and may not fully capture curricula in all primary care residency programs; response bias may be present in that only programs with an interest in elder abuse responded. Third, since the survey is lim-ited to Michigan residency programs, results may not be generalized across the nation. Fourth, although we believe that program directors completed the survey, if this task was delegated to another faculty member it
may not accurately reflect program curriculum. Finally,
how residency programs teach and what residents learn about elder abuse may or may not be fully applicable to real-world clinical practice.
Figure 4
Conclusions
As the American population ages, elder abuse may become an increasingly important clinical program. Based on the results of our study, more targeted education about elder abuse is needed in primary care residency programs.
Acknowledgments: Funding for this survey project was provided in full by a grant from the Blue Cross Blue Shield Foundation of Michigan. The authors wish to acknowledge the help of focus group members for this project,
in-cluding Drs Frank Komara, Jed Magen, Laurie Post, and Dorrie Rosenblatt,
for their help and consultation in instrument development.
The content of this study was presented April 30, 2007, at Lansing Community 2nd Annual Research Day, Lansing, Mich (development of
cluster groups) and in March 2008 at the Annual Meeting of the American Association of Geriatric Psychiatrists (descriptive data only).
Corresponding Author: Address correspondence to Dr Wagenaar, Michigan State University, Department of Psychiatry, A 231 E Fee Hall, E. Lansing,
MI 48824. 517-432-2994. Fax: 616-676-2781. [email protected].
References
1. Shields LB, Hunsaker DM, Hunsaker JC. Abuse and neglect: A 10-year
review of mortality and morbidity in our elders in a large metropolitan area. J Forensic Sci 2004;49(1):122-7.
2. Laumann EO, Leitsch SA, Waite LJ. Elder mistreatment in the United
States: prevalence estimates from a nationally representative study. J Gerontol B Psychol Sci Soc Sci 2008;63(4):S248-S254.
3. National Center on Elder Abuse and Westat, Inc. The National Elder Abuse Incident Study: Final Report. Washington, DC: National Center on Elder Abuse, September 1998.
4. Kahan F. Why elder abuse continues to elude the healthcare system. Mt Sinai J Med 2003;70(1):62-8.
5. Nelson HD, Nygren P, McInerney Y, Klein J. Screening women and elderly adults for family and intimate partner violence: a review of the evidence for the US Preventive Services Task Force. Ann Intern Med 2004;140(5):387-96.
6. Taylor DK, Bachuwa G, Evans J, Jackson-Johnson V. Assessing barriers
to the identification of elder abuse and neglect: a community-wide sur -vey of primary care physicians. J Natl Med Assoc 2006;98(3):403-4. 7. Wei GS, Hergers JE Jr. Reporting elder abuse: a medical, legal, and
ethical overview. J Am Med Womens Assoc 2004;59(4):248-54. 8. Jones JS, Veenstra TR, Seamon JP, Krohmer J. Elder mistreatment:
national survey of emergency physicians. Ann Emerg Medicine 1997;30(4):473-9.
9. US Preventive Services Task Force. Screening for family and intimate partner violence: recommendation statement [review]. Ann Fam Med 2004;2(2):156-60.
10. Kennedy RD. Elder abuse and neglect: the experience, knowledge, and attitudes of primary care physicians. Fam Med 2005;37(7):481-5. 11. Accreditation Council for Graduate Medical Education. Residency
Review Committee for Family Medicine. www.acgme.org/acWebsite/ downloads/RRC_PIF120_FamilyMedicineContinued_082007_
u06102008.doc.
12. Hill JR. Teaching about family violence: a proposed model curriculum.
Teach Learn Med 2005;17(2):169-78.
13. Heath JM, Dyer CB, Kerzner LJ, Mosqueda L, Murphy C. Four
models of medical education about elder mistreatment. Acad Med 2002;77(11):1101-6.
14. Dong X. Medical implications of elder abuse and neglect. Clin Geriatric Med 2005;21(2):293-313.
