Background: Studies have demonstrated that measures of lower quality of care and associated adverse health effects are more prevalent in for-profit nursing homes compared to not-for-profit facilities. However, these studies omit persons who receive care in the community setting, and exclusively focus on isolated clinical signs that may obscure the true effect size, since these clinical signs rarely occur in isolation. Objective: In this study, we use the Clinical Signs of Neglect Scale (CSNS), which is an aggregate measure of clinical signs of neglect and substandard care, to evaluate the association of residence type on health outcomes among individuals living in both private community residences and for-profit and not-for-profit long-term care facilities. Methods: In a multicenter, retrospective data analysis of 1,149 patients identified from an inpatient hospital registry, we assessed the relationship between residence type (community dwelling, not-for-profit, and for-profit facilities) and clinical signs of neglect. Adjusted parameter estimates and 95% CIs were estimated with linear regression in 3 models using different reference groups. Results: The most serious clinical signs were consistently more prevalent among residents of for-profit facilities, as were measures of poor institutional quality. Relative to low-functioning community-dwelling patients, the mean difference in CSNS scores was higher among patients residing in not-for-profit facilities by 1.99 (p = 0.012) and 3.55 (p ≤ 0.001) among patients in for-profit facilities. In a separate model, the mean difference in CSNS scores among patients living in for-profit facilities compared to not-for-profit facilities was 1.90 (p = 0.035). Conclusions: Using an aggregate measure, our findings support prior studies demonstrating an association between residence type and adverse health outcomes for disabled elderly.

Neglect is a common form of reported elder mistreatment [1, 2], and occurs among individuals living in private residences as well as institutions [1, 3, 4]. While a primary definition for neglect [5] includes institutional caregivers who fail to provide services that maintain the well-being of residents, most cases of neglect occurring in institutions are unlikely to be attributed to substandard care, but rather ascribed to the dying process or the result of comorbidities [6, 7]. Neglect in an institutional setting, in the form of substandard care, directly relates to quality-of-care issues. Many studies have shown that low quality of care as measured by staffing, capacity, training, and deficiencies is strongly associated with adverse health outcomes such as pressure ulcers, hospitalizations, and mortality [7‒9], and a large proportion of residents go through adverse events that threaten their health [6]. Furthermore, measures of lower quality of care and associated adverse health effects are more prevalent in for-profit facilities compared to not-for-profit facilities [7, 10, 11]. For-profit facilities have profit margin goals [4] and on average pay much higher salaries to their management teams [4, 12], which translates to lower expenditures on staff and services within these institutions [4, 12, 13].

However, the studies that evaluate the relationship between residence type and poor health outcomes have exclusively focused on individual clinical signs (e.g., pressure ulcers) [7, 10]. The problem with this approach is that these clinical signs rarely occur in isolation, but rather patients experiencing potential neglect present with a complex constellation of clinical signs that varies in severity [14, 15]. By focusing on individual health outcomes, the true effect size of the association between facility type and adverse health effects may be obscured. In this study, we use the Clinical Signs of Neglect Scale (CSNS) [15], an aggregate measure of clinical signs of neglect, to evaluate the association of residence type on health outcomes.

This study was conducted within the framework of a large multicenter, retrospective analysis of victims of elder mistreatment in the Greater Chicago metropolitan area. For the elder neglect component of the study, we included patients admitted to 5 of the largest Chicago area hospitals between the years 2007 and 2011 who were 60 years and older. These patients lived in home settings as well as not-for-profit and for-profit institutions. The age limit for inclusion in this study is based on both the Illinois Elder Abuse and Neglect Act and an operational definition set by a committee of elder mistreatment experts [5, 14]. These 5 hospitals provide medical care to approximately 10% of all the inpatients in the State of Illinois [16] and serve socioeconomically diverse communities. The University of Illinois at Chicago IRB approved this work (# 2013-0714).

CSNS Description

For the purpose of quantifying clinical signs, we used the CSNS [15]. Table 1 lists the clinical signs included in the CSNS. The CSNS solves one of the key problems of this complex constellation of signs, in that it does not treat each sign of potential neglect equally. Clinical signs weakly associated with neglect such as malnutrition, dehydration, and urinary tract infections are assigned lower weights compared to clinical features more strongly indicative of neglect such as pressure ulcers, dehydration in patients with gastrostomy tubes, or broken catheters/tubes. The scale also differentiates between places of residence, assuming a higher level of care in nursing facilities than in private community residences. In addition, from a statistical perspective, the use of a single scale as compared to using 20–30 different variables to characterize clinical signs of neglect improves the ease of interpretation and parsimony of multivariable models.

Table 1.

Clinical signs of neglect among community- and facility-dwelling patients at the time of index hospitalization, Illinois elder neglect study

 Clinical signs of neglect among community- and facility-dwelling patients at the time of index hospitalization, Illinois elder neglect study
 Clinical signs of neglect among community- and facility-dwelling patients at the time of index hospitalization, Illinois elder neglect study

The CSNS was based on a composite of clinical signs identified in the literature (content validity) and reviewed by an expert panel using a DELPHI process (consensual validity). The weighted CSNS used in this analysis ranges from 0 to 60. However, a patient with all of the clinical signs simultaneously would likely be in an unsurvivable state. The maximum value observed in the patient population in this study was 30. Generally, a one-point increase in the CSNS represents the diagnosis of an additional clinical sign of neglect or the presence of a moderate/serious clinical sign. Based on a prior report [15], a score of 5 or greater for the weighted scoring scheme provided a high sensitivity (89.7–88.9%) and adequate negative predictive value (69.2–75.0%) when compared to clinical recommendations made by 2 geriatricians. For this reason, CSNS values of ≥5 in this pre-screener were determined as an appropriate cutoff for further screening/evaluation using more in-depth tools to determine neglect.

Identifying Potential Study Subjects

The selection of subjects was conducted with the objective of ensuring that the study included a diverse range of patients along the full CSNS continuum – patients ranging from independent elderly individuals to those dependent on others for assistance, as well as dying patients. Patients presenting with multiple clinical signs of neglect are relatively rare, so sampling was divided into 2 segments. First, to identify patients presenting with clinical signs of neglect, not attributable to comorbidities [15], we screened for all (1) patients admitted for (a) poisoning as a result of a medication error and (b) injuries caused by excessive heat or cold among those affected by conditions that restrict activities of daily living including dementia, disorders of the central nervous system, disorders of the peripheral nervous system, depression, adjustment reaction, osteoarthropathy, and neurological impairments caused by cerebrovascular accident, as well as (2) patients presenting with conditions indicating physical decline that could be the result of neglect including (a) cachexia, (b) failure to thrive, (c) protein-calorie malnutrition, (d) dehydration, (e) pressure ulcers on weight bearing parts of body, (f) fecal impaction, (g) complications from poor management of a Foley catheter, colostomy, and enterostomy, (h) muscle or tendon contracture, and (i) aspiration pneumonia [5, 15, 17‒19]. We excluded patients if they were affected by a comorbid condition known to be associated with the indicated diagnosis (e.g., cachexia with a metastatic cancer diagnosis; a detailed listing of exclusion criteria are described in Friedman, 2017) [15].

Second, to ensure the selection of patients on the lower end of the CSNS continuum, we randomly selected patients from all admissions to any of the 5 hospitals who were 60 years and older that (1) did not have any clinical evidence of neglect, (2) had no reported conditions that limited activities of daily living, or (3) were diagnosed with conditions indicating physical decline but had a comorbid condition that was part of the exclusion criteria for the first group of subjects. This second group made up approximately a third of the study population (n = 339). The unweighted CSNS scores in the full study population were near normally distributed (mean = 2.72; SD = 2.33; skewness = 1.02; kurtosis = 0.96; Shapiro-Wilkes W = 0.90).

Record Abstraction

A structured data abstraction form was used, with double entry of 10% of records, to systematically gather information on (1) demographics, (2) functional independence, (3) residential history at the time of the index hospitalization as well as after discharge, (4) physiological intake measures, (5) history of mistreatment from patient and caregiver reports, (6) all billing information including ICD-9-CM diagnosis codes, and (7) complete narrative fields.

Death Certificates

Death records through December 31, 2013 were collected from the National Death Index (NDI), which because of the lag in availability of NDI data was the most recent data available at the time of the data request in September 2015. NDI matched our study subjects with death records using the following core information: social security number, date of birth, first and last name, state of residence, sex, race and marital status. Most persons living in nursing homes are there because of decompensation in their health and functional independence that happens toward the end of life, and persons at the end of their lives are challenged by nutritional and health conditions that make them more susceptible to conditions included in the CSNS. To control for potential confounding associated with the process of dying, days of follow-up were calculated subtracting the index hospitalization date of discharge from the date of death or end of follow-up (December 12, 2013), whichever came first.

Adult Protective Services Investigations Data

Adult Protective Services in the Illinois Department on Aging provided us with data on any investigation conducted through the study period for all patients. The data from the Illinois Department on Aging included all available fields in their database and was used to augment patient, family, and medical professional reports of mistreatment recorded in the medical records. For this study, only cases with reports of mistreatment occurring prior to or within 60 days following the index hospitalization were classified as having a history of mistreatment, which excludes unrelated incidents occurring after the index hospitalization.

CMS Data

For all patients living in long-term care facilities, we gathered data for each facility that the patient resided in immediately prior to and after the index hospitalization as reported on the annual certification surveys, which includes detailed provider information regarding capacity, staffing, and deficiencies. The publicly available provider and deficiency files were downloaded [20] and the data were linked to each patient based on the year of the index hospitalization and facility number. Information on the nature of the corporate structure of each facility was included in the provider file (not-for-profit vs. for-profit; part of corporate chain). An aggregate 3-year variable was created for each patient that summed the number of deficiencies listed in the separate annual deficiency files for total violations in the years prior, of, and after a patient’s index admission. Similar 3-year aggregate variables were calculated for all levels of deficiencies including deficiencies relating to substandard quality of care.

Residential Status Prior to Index Hospitalization

The study group was divided into 4 residential categories based on reported residence prior to the index hospitalization: (1) high-functioning patients living in a private community residence, including senior independent living facilities, who were independent or only required partial assistance across all activities of daily living and were not bedfast, (2) low-functioning patients living in a private community residence, including senior independent living facilities, with 1 or more activities of daily living in which the patient required total assistance or was reported to be bedfast, (3) patients living in not-for-profit long-term care facilities, and (4) patients living in for-profit long-term care facilities. A total of 22 patients were excluded from the analysis because they were homeless (n = 7), living in a hospice (n = 1), jail (n = 1), or had unknown living situations (n = 12).

Additional Covariates

Cumulative activities of daily living (ADL) scores were calculated based on 10 functional items as measured by a physical therapist or a registered nurse during the first 24 h of hospitalization [21]. For each item, the patient was assessed as independent (2 points), required partial assistance (1 point), or required total assistance to complete a specific activity (0 points). Rubin’s multiple imputation was used to impute missing data for 16.2% of subjects with incomplete ADL data [22]. To characterize comorbidities, the Elixhauser Comorbidity Index was used [23]. Because the Elixhauser Comorbidity Index only includes Alzheimer’s Disease for dementias, a separate variable was created to identify any patient diagnosed with any other form of dementia. Low-weight patients were identified using calculated body mass index scores less than 18.5 based on weight and height in their medical records. ICD-9-CM billing codes and abstracted narrative data were used to identify patients (1) with gastric or jejunal feeding tubes, (2) who were bedfast or chairfast, (3) who were respirator dependent, (4) who used restraints, and (5) who were visually impaired.

Statistical Analysis

We used SAS software for all statistical analyses (v.9.4; Cary, NC, USA). Because nursing home residents are usually low functioning with multiple comorbidities, it is important to account for this when comparing them to community-dwelling individuals. Therefore, we developed 3 multivariable linear regression models to evaluate the relationship between CSNS scores and residential status prior to the index hospitalization: (Model 1) high-functioning, community-dwelling individuals were used as the reference group, (Model 2) low-functioning, community-dwelling individuals were used as the reference group and high-functioning community-dwelling individuals were excluded, and (Model 3) patients living in not-for-profit facilities were the reference group and we excluded all community dwellers. Statistical evaluation of covariates, as well as a priori knowledge [7‒11], was used to determine inclusion of covariates in the final models. At a minimum, the following covariates were assessed in all models ADL functionality, Elixhauser Comorbidity Index, Dementia, cerebrovascular accident, bedfast, sensory impairments, insurance coverage, days of follow-up (see detailed description above in Death Certificates section), age at time of index hospitalization, race/ethnicity, and history of elder mistreatment. The third model also included the number of residents who were Medicaid recipients and the number of deficiencies over 3 years as reported in the CMS data [20]. A 2-sided p-value less than 0.05 was considered statistically significant. Parameter estimates from the adjusted models are presented, including the 95% CI. No evidence of multicollinearity among the final independent variables was indicated (based on evaluation of standard errors and evaluation of variance of inflation and tolerance tests).

The demographic data for the 1,149 patients included in this analysis are presented in Table 2. Among the community dwelling patients at the time of the index hospitalization, the majority lived in private residences (n = 670; 93.2%) and the remainder lived in senior living or assisted living communities (n = 49; 6.8%). Among the facility dwelling patients, 61 patients lived in 22 different not-for-profit facilities and 369 patients lived in 83 different for-profit facilities. All the facilities were in the Chicago metropolitan area.

Table 2.

Demographic characteristics of community- and facility dwelling-patients at the time of index hospitalization, Illinois elder neglect study

 Demographic characteristics of community- and facility dwelling-patients at the time of index hospitalization, Illinois elder neglect study
 Demographic characteristics of community- and facility dwelling-patients at the time of index hospitalization, Illinois elder neglect study

Among those living in for-profit facilities, a greater proportion were African American/Black (p < 0.001) and had co-insurance of Medicaid (p < 0.001), but a lower proportion had co-insurance from a private insurer (p < 0.001). All of the cases of the uninsured elderly (n = 45) lived in a community dwelling, of which almost half were Hispanic (n = 22; p < 0.001). Low-functioning patients living in the community had the highest proportion of a reported history of mistreatment (18.3%; p < 0.001). Among all patients, 21 cases were reported to authorities for mistreatment by hospital staff during the index hospitalization, of which 6 were residents in institutions. Among the community-dwelling residents, 19.1% of the high-functioning and 51.7% of the low-functioning subjects had documented homemaker, nursing, and/or physician home care services. The proportion of patients who returned to the same place of residence after discharge is as follows: high-functioning community residents, 64.6% (n = 387); low-functioning community residents, 59.2% (n = 71); residents of not-for-profit facilities, 67.2% (n = 41); and residents of for-profit facilities, 76.7% (n = 283).

Health and Functioning

Table 3 presents information on general health and functioning of the patients. At the time of the index hospitalization, low-functioning, community-dwelling patients and facility-dwelling patients had substantially lower mean activities of daily living scores (mean ADL scores respectively: low functioning community dwelling, 3.4; not-for-profit facility dwelling, 4.9; for-profit facility dwelling, 4.7) compared to high-functioning community-dwelling patients (mean ADL, 17.4; p < 0.001). Among the patients dwelling in for-profit facilities, a greater proportion had an assigned medical power of attorney or guardianship, use of restraints, and gastrostomies. Overall, the mean number of comorbidities was highest among the patients living in for-profit facilities, as well as specific neurologic and psychiatric conditions. The proportion of patients who died by the end of the period of follow-up was highest among those living in for-profit facilities (82.9%) followed by low-functioning community-dwelling individuals (77.5%).

Table 3.

Health status and outcomes of community- and facility-dwelling patients, Illinois elder neglect study

 Health status and outcomes of community- and facility-dwelling patients, Illinois elder neglect study
 Health status and outcomes of community- and facility-dwelling patients, Illinois elder neglect study

Institutional Quality Measures

Compared to not-for-profit facilities, the for-profit facilities had significantly higher patient capacity and volume, as well as lower staffing commitments during the year of the index hospitalization (For-profit vs. not-for-profit respectively): mean number of beds (220.9 vs. 183.2; p < 0.001), mean number of residents in certified beds (178.2 vs. 117.7; p < 0.001), proportion of residents who were Medicaid recipients (71.8 vs. 48.7%; p < 0.001), mean total nurse staffing hours per resident per day (Certified nursing assistants + licensed practical nurses + registered nurses: 4.6 vs. 7.2; p < 0.001) and mean total registered nurse staffing hours per resident per day (0.8 vs. 1.8; p < 0.001). The proportion of residents who were Medicaid recipients was lower within facilities belonging to a corporate chain in both not-for-profit facilities (chain vs non-chain: 44 vs. 55.9%) and for-profit facilities (chain vs. non-chain: 63.8 vs. 76.4%). The mean number of total deficiencies (e.g., failure to report acts of abuse, neglect or mistreatment of residents; failure to hire personnel without known histories of abuse; failure to limit use of restraints to only cases required for medical treatment) in the years prior, of, and after a patient’s index admission was substantially higher in for-profit facilities compared to not-for-profit facilities (12.0 vs. 3.2; p < 0.001).

Clinical Signs of Neglect

Table 1 presents the proportion of cases with specific clinical signs of neglect as well as the cumulative crude and weighted scores (Fig. 1). In all cases, patients living in for-profit facilities had significantly higher mean crude and weighted CSNS scores. The mean weighted CSNS scores did not significantly differ when comparing facilities that belonged to a larger corporate chain with those that did not for either not-for-profit facilities (chain vs. single facility respectively: 7.7 and 6.7 mean CSNS) or for-profit facilities (chain vs. single facility respectively: 9.3 vs. 9.0 mean CSNS).

Fig. 1.

Distribution of clinical signs of neglect scale (CSNS) scores by residential type, Box-Whisker plots.

Fig. 1.

Distribution of clinical signs of neglect scale (CSNS) scores by residential type, Box-Whisker plots.

Close modal

In the final multivariable linear regression model, where high-functioning community-dwelling individuals were used as the reference group, the mean difference in weighted CSNS scores was as follows: low-functioning patients living in the community setting –0.81 (95% CI –1.87 to 0.26; p = 0.137), patients residing in not-for-profit facilities 1.19 (95% CI –0.04 to 2.42; p = 0.058) and patients residing in for-profit facilities 2.83 (95% CI 1.98 to 3.67; p ≤ 0.001; R2 = 0.414). In a second multivariable model, excluding high-functioning community dwellers, where the low-functioning community-dwelling patients were the reference group, the mean difference in CSNS scores was as follows: patients residing in not-for-profit facilities 1.99 (95% CI 0.43–3.55; p = 0.012) and patients residing in for-profit facilities 3.55 (95% CI 2.44–4.67; p ≤ 0.001; R2 = 0.258). In a third multivariable model, excluding all community dwellers, where the patients living in not-for-profit facilities were the reference group, the patients living in for-profit facilities had a mean difference in CSNS scores of 1.90 (95% CI 0.23–3.69; p = 0.035; R2 = 0.198).

Patients receiving care in for-profit institutions were diagnosed with substantially more clinical signs of neglect than patients residing in not-for-profit facilities and low-functioning, community-dwelling patients. As reported in prior research, for-profit facilities caring for the patients in this study were significantly inferior across nearly all staffing, capacity, and deficiency measures [7, 10, 11]. Furthermore, the most serious clinical signs were consistently more prevalent among residents of for-profit facilities, including dehydration with presence of gastrostomy, not being provided basic medications to manage chronic conditions, stage 3 or 4 pressure ulcers, and complications with urinary catheters and feeding tubes.

Many studies show that neglect is the most common form of elder mistreatment [1‒4] but is more likely to be overlooked because of its muted nature, although the outcomes of neglect can be as pernicious as physical abuse. The definition for neglect with the greatest consensus [5] clearly includes substandard care provided by institutions and medical professionals. However, what makes neglect difficult to identify is that the clinical signs of neglect are often common clinical features of many different diseases and occur as part of the dying process. In this study controlling for potential confounding associated with the process of dying as well as comorbidities, the association between residence type and clinical signs of neglect persisted.

However, without adequate identification and reporting of cases to parties responsible for oversight, it will be difficult to effect change that protects patients and leads to remediation of substandard facilities. Screening in hospitals is important because it is generally the only location an elderly person has contact with medical professionals outside of the nursing home or private residence [24]. All too often, a nursing home resident is transferred to a single hospital or group of interchanging hospitals, where substandard care is not considered a contributing factor of the clinical signs. While in the hospital, the patient is temporarily stabilized and in some cases shows improvement. Then the patient is sent back to the long-term care facility only to deteriorate again, later being readmitted to a hospital, frequently with non-descript conditions such as “altered mental status”. The only way to break the cycle is to improve screening so that these patients can be identified prior to discharge and alternative residential options can be explored or better medical care plans can be provided to the long-term care facilities.

Yet there exist substantial barriers to implementing screening tools for elder mistreatment in hospitals including a lack of awareness of the problem, unfamiliarity with the reporting process, fear of lawsuits, lack of institutional protocols for identifying mistreatment, lack of training on related issues, feeling uncomfortable talking about mistreatment because of poor training, and potential conflicts of interest when the physician is employed by both the hospital and nursing home [25]. For these reasons, medical personnel do a very poor job recognizing elder mistreatment, especially neglect [24, 26‒29].

One integral component to improving screening of elder neglect in hospitals is through the implementation of automated pre-screening tools [24] like the CSNS. Pre-screeners filter the large pool of elderly patients treated daily in hospitals and identify only those that require more detailed interviews. Pre-screeners are most useful when there are systemic failures to screen for common events, or a failure to identify less common events that pose substantial risks to a patient’s quality of life or health, as is the case with elder neglect. In a feasibility simulation using over 600,000 elderly patients, the CSNS pre-screener was shown to reduce this large pool of patients by 95% [15]. In other words, using an automated version of the CSNS means that you would only have to conduct in-depth follow-up screening using more conventional time consuming elder mistreatment screening tools on 5% of this patient population. For most facilities, this translated to in-depth screening of 1–3 patients per day [15].

The CSNS was designed to improve discharge planning for at-risk patients and to be used as a tool to efficiently screen for patients who may require more time intensive traditional elder mistreatment screening. It was not designed as a definitive tool to identify cases of neglect for legal intervention. Elder mistreatment cannot be substantiated on physical features alone and generally requires evidence that is not available to clinicians. These investigations must be handled by Adult Protective Services and other enforcement agencies who investigate mistreatment. It is the role of hospital-based clinicians to identify and report suspected cases only; but these clinicians are critical to the process because these patients frequently have limited contact with people outside of their residences [24].

In evaluating the association between facility type and clinical signs of neglect, several limitations should be considered. The referral patterns of patients to hospitals may differ in not-for-profit and for-profit facilities. However, the inclusion criteria were not based on place of residence, and the 5 hospitals have very large catchment areas that cover approximately 10% of all inpatient cases statewide. In the Chicago area (Counties of Cook, Lake, DuPage and Will), there were 338 nursing and rehabilitation facilities that provided different levels of nursing support to persons in need and were monitored by CMS between 2000 and 2015. Among these facilities, 76.4% were registered as for-profit facilities, which is nearly identical to the distribution in our study (83 out of 105 facilities; 79.1%). In addition, the number of deficiencies and nursing hours per patient was strongly associated with the facility type, which has been shown to be associated with adverse health outcomes [7‒9]. Second, residence location and type varies across time. We did not have information on the length of residency prior to or after the index hospitalization at any given location, but we did have address location for the place of residence across multiple hospitalizations. In a subsequent study, we are looking at change in health status as patients move across time between different types of facilities. Third, nursing homes are more likely to treat persons with functional limitations, as well as persons at the end of their lives who are more likely to present with the clinical signs treated as outcomes in this study. This is why we included days of follow-up in our models and the low-functioning, community-dwelling group in this analysis. In fact, the ADLs, low BMI, and comorbidities were worse or comparable in this community-dwelling group. Fourth, some of the clinical signs are interdependent and vary across time. For example, malnutrition and dehydration are associated with the occurrence and worsening of pressure ulcers [30], which may worsen the state of malnutrition, as appetite is suppressed by the onset of pain and the medication to treat the pain.

During the past 10 years, the number of government and nonprofit nursing homes has declined nationwide in the United States, while the number of for-profit nursing homes has increased slightly [31]. The small decline in the overall number of nursing homes and available beds is running counter to an increase in the number of people needing services among the disabled and elderly. This puts further pressure on existing nursing homes and reduces the discharge options available for hospital staff and individuals. For-profit facilities help fill this void and are important to the overall capacity needs nationwide, but adequate oversight needs to be in place to guarantee that an increase in capacity does not occur at the expense of quality of care. However, oversight cannot be restricted to for-profit facilities but needs to occur concurrently with improved screening and reporting of suspected cases by all parties, better staffing of enforcement agencies, and training in evaluations of quality assurance and performance improvement programs as well as quality assurance and assessment committees, tightening of adherence to federal law by Central Management Services that ties Medicare and Medicaid reimbursement with quality of care, pressure from insurance providers to limit costly outcomes, and increased investment in quality staff and facilities.

1.
Acierno R, Hernandez MA, Amstadter AB, Resnick HS, Steve K, Muzzy W, et al: Prevalence and correlates of emotional, physical, sexual, and financial abuse and potential neglect in the United States: the National Elder Mistreatment Study. Am J Public Health 2010; 100: 292–297.
2.
Yon Y, Mikton CR, Gassoumis ZD, Wilber KH: Elder abuse prevalence in community settings: a systematic review and meta-analysis. Lancet Glob Heal 2017; 5:e147–e156.
3.
U.S. General Accounting Office, Nursing Homes: More Can Be Done to Protect Residents from Abuse, GAO-02-312 (Washington, March 2002).
4.
U.S. General Accounting Office, Nursing Homes: CMS’s Special Focus Facility Methodology Should Better Target the Most Poorly Performing Homes, Which Tended to Be Chain Affiliated and For-Profit, GAO-09-689 (Washington, August 2009).
5.
National Research Council (US): Panel to Review Risk and Prevalence of Elder Abuse and Neglect. Bonnie RJ, Wallace RB, editors. Elder Mistreatment: Abuse, Neglect, and Exploitation in an Aging America. Washington, National Academies Press (US), 2003.
6.
Office of Inspector General (US): Adverse Events in Skilled Nursing Facilities: National Incidence Among Medicare Beneficiaries. Daniel R. Levinson Inspector General, OEI-06-11-00370 (Washington, February 2014).
7.
Comondore VR, Devereaux PJ, Zhou Q, Stone SB, Busse JW, Ravindran NC, et al: Quality of care in for-profit and not-for-profit nursing homes: systematic review and meta-analysis. BMJ 2009; 339:b2732.
8.
Centers for Medicare and Medicaid Services (CMS): 2001. Report to Congress: Appropriateness of Minimum Nurse Staffing Ratios in Nursing Homes: Phase II Final Report. Baltimore: CMS. https://www.justice.gov/sites/default/files/elderjustice/legacy/2015/07/12/Appropriateness_of_Minimum_Nurse_Staffing_Ratios_in_Nursing_Homes.pdf (accessed August 12, 2017).
9.
Castle NG, Engberg J: Further examination of the influence of caregiver staffing levels on nursing home quality. Gerontologist 2008; 48: 464–476.
10.
Ronald LA, McGregor MJ, Harrington C, Pollock A, Lexchin J: Observational evidence of for-profit delivery and inferior nursing home care: when is there enough evidence for policy change? PLoS Med 2016; 13:e1001995.
11.
Barron DN, West E: The quasi-market for adult residential care in the UK: do for-profit, not-for-profit or public sector residential care and nursing homes provide better quality care? Soc Sci Med 2017; 179: 137–146.
12.
Friedman BS, Hattis PA, Bogue RJ: Tax exemption and community benefits of not-for-profit hospitals. Adv Health Econ Health Serv Res 1990; 11: 131–158.
13.
Harrington C, Olney B, Carrillo H, Kang T: Nurse staffing and deficiencies in the largest for-profit nursing home chains and chains owned by private equity companies. Health Serv Res 2012; 47: 106–128.
14.
Bond MC, Butler KH: Elder Abuse and Neglect. Definitions, Epidemiology, and Approaches to Emergency Department Screening. Clin Geriatr Med 2013; 29: 257–273.
15.
Friedman LS, Avila S, Liu E, Dixon K, Patch O, Partida R, Zielke H, Giloth B, Friedman D, Moorman L, Meltzer W: Using clinical signs of neglect to identify elder neglect cases. J Elder Abuse Negl 2017; 29: 270–287.
16.
Illinois Health Facilities and Services Review Board, Illinois Department of Public Health: Annual Hospital Questionnaire: Hospital Profiles and Annual Bed Reports, 2011. Springfield, Illinois Department of Public Health, 2013.
17.
Yaffe MJ, Wolfson C, Lithwick M, Weiss D: Development and validation of a tool to improve physician identification of elder abuse: the elder abuse suspicion index (EASI). J Elder Abuse Negl 2008; 20: 276–300.
18.
Dong XQ, Simon M, Evans D: Cross-sectional study of the characteristics of reported elder self-neglect in a community-dwelling population: findings from a population-based cohort. Gerontology 2010; 56: 325–334.
19.
Phelan A: Elder abuse in the emergency department. Int Emerg Nurs 2012; 20: 214–220.
20.
Centers for Medicaid and Medicare Services: Five Star Nursing Home Quality Rating System, Downloads of Staffing, Provider, Penalty and Deficiency Files. https://www.cms.gov/medicare/provider-enrollment-and-certification/certificationandcomplianc/fsqrs.html (Last accessed October 12, 2017).
21.
Hardy S: Consideration of Function & Functional Decline. Current Diagnosis and Treatment: Geriatrics, Second Edition. Editors: Williams BA, Chang A, Ahalt C, Chen H, Conant R, Landefeld CS, Ritchie C, Yukawa M. New York, McGraw-Hill, 2014, pp 3–4.
22.
Allison PD: Missing Data 2002. Thousand Oaks, Sage Publications.
23.
Elixhauser A, Steiner C, Harris DR, Coffey RM: Comorbidity measures for use with administrative data. Med Care 1998; 36: 8–27.
24.
Platts-Mills TF, Barrio K, Isenberg EE, Glickman LT: Emergency physician identification of a cluster of elder abuse in nursing home residents. Ann Emerg Med 2014: 64: 99–100.
25.
Clarke ME, Pierson W: Management of elder abuse in the emergency department. Emerg Med Clin North Am 1999; 17: 631–644.
26.
Fulmer T, Guadagno L, Bitondo Dyer C, Connolly MT: Progress in elder abuse screening and assessment instruments. J Am Geriatr Soc 2004; 52: 297–304.
27.
Saveman BI, Sandvide A: Swedish general practitioners’ awareness of elderly patients at risk of or actually suffering from elder abuse. Scand J Caring Sci 2001; 15: 244–249.
28.
Cooper C, Selwood A, Livingston G: The prevalence of elder abuse and neglect: a systematic review. Age Ageing 2008; 37: 151–160.
29.
Stevens TB, Richmond NL, Pereira GF, Shenvi CL, Platts-Mills TF: Prevalence of nonmedical problems among older adults presenting to the emergency department. Acad Emerg Med 2014; 21: 651–658.
30.
Posthauer ME, Banks M, Dorner B, Schols JM: The role of nutrition for pressure ulcer management: national pressure ulcer advisory panel, European pressure ulcer advisory panel, and pan pacific pressure injury alliance white paper. Adv Skin Wound Care 2015; 28: 175–188; quiz 189–190.
31.
Centers for Medicare and Medicaid Services (CMS): 2015. Nursing Home Data Compendium 2015 Edition. https://www.cms.gov/Medicare/Provider-Enrollment-and-Certification/CertificationandComplianc/Downloads/nursinghomedatacompendium_508–2015.pdf (Last accessed August 12, 2017).