In the United States, the elderly population is expected to grow from 35 million in 2000 to a predicted 86 million (1 in 5) people by 2050. With longer lifespans the Population Reference Bureau predicts that by 2060 there will be about 2.3 million individuals over the age of 65 residing in nursing facilities. Unfortunately, despite the widespread population they service, long-term care facilities (LTCFs) have not received the level of scientific attention and study that other areas of the healthcare field have.
The ethics of informed consent (IC) is a major challenge to conducting research in LTCFs. LTCF residents are considered a vulnerable population, primarily because at least 50% of them suffer from some degree of cognitive impairment. Therefore, questions related to decision-making capacity are common. Mezey et al and Brod et al determined that moderately cognitively impaired individuals can respond to questions about end-of-life care decisions and quality of life. Furthermore, Buckles et al suggested that very mildly and mildly demented individuals understood the information provided to them during the consent process, whereas the involvement of a responsible caregiver was necessary for residents with moderate dementia. Effectively, an individual's capacity to consent should be assessed separately for each research protocol.
In the US, an individual should demonstrate four abilities to be deemed able to consent:
- (1) be able to understand choices and their consequences,
- (2) be able to understand relevant information,
- (3) be able to understand associated risks and benefits, and
- (4) be able to communicate a choice.
To the best of our knowledge, there is no golden standard used to determine capacity to give IC. In this study, we evaluated the LTCF personnel's assessment of resident ability to provide IC to clinical research, and studied potential limitations to resident autonomy.
2.1 Study design
This study is part of a larger minimally-invasive observational clinical study on the impact of vaccine-preventable infections among LTCF residents in RI, USA. Our sample included 284 long-term care residents. Upon receiving approval from the Institutional Review Board, each LTCF was contacted and given a full description of the study, along with copies of the study protocol and accompanying documents. Following a letter of cooperation, LTCFs provided a list of long-term residents, and their evaluation as consentable or non-consentable individuals. LTCF specific characteristics, including the percentage of long-term residents and Medicare rating for each facility can be found in Supplemental Digital Material, Table 2, http://links.lww.com/MD/E259.
A self-administered questionnaire (Supplemental Digital Material, Table 1, http://links.lww.com/MD/E258) was given to LTCF directors of nursing and social workers, in order to explore the process of the creation of the list with long-term residents. Three researchers answered potential questions at their assigned facilities. This improved the inter-subjectivity of the results provided by different facilities. No restrictions were placed on the residents’ age of eligibility. When interviews were completed, we obtained Brief Interview for Mental Status scores for all 284 recruited residents included in the study. BIMS was the only consistent tool assessing cognitive capacity across all LTCFs. The BIMS assessment has been validated among nursing home residents under the minimum data set (MDS) 2.0 and 3.0 which instructed nursing home staff to administer the BIMS assessment to all their residents who could communicate in any sort of way, followed by what the MDS considers as the gold standard for assessment of cognitive function, namely the modified mini mental status (3MS) exam. In our study, residents who were cognitively unable to be interviewed by the LTCF staff were not included in our analyses. BIMS scores range from 1 to 15. Scores of 1 to 7 indicate severe impairment, 8 to 12 indicate mild impairment, and 13 to 15 indicate intact cognitive abilities.
2.2 Consent process
2.2.1 Residents perceived by LTCF staff as consentable
We visited residents in a private area. During this visit, we verbally reviewed all elements of the Institutional Review Board -approved consent, including the study's purpose, methods, risks, and benefits. Furthermore, we asked for permission to obtain information about their medical records through a medical records release form.
2.2.2 Residents perceived by LTCF staff as non-consentable
We placed telephone calls to all proxies. During this time, a researcher explained the protocol, including the study goals, methods, risks, and benefits. The researcher then met the proxy in a private setting at the facility to obtain consent.
2.3 Data analysis
The analyses performed aimed at describing the resident population and their cognitive status based on both BIMS scores and LTCF personnel assessment. All data were acquired from the individual nursing homes, and there were no missing data.
Mean resident age, length of stay (LOS) and BIMS scores with 95% confidence intervals were calculated for all facilities in order to examine differences among LTCFs (Tables 1 and 2). The most recent admission date was used to calculate LOS for residents who had left the LTCF for longer than 6 months and then returned. A Student's t test was performed to determine if there was a statistically significant difference in the mean BIMS scores of residents perceived by LTCF staff as either consentable or non-consentable.
We carried out a logistic regression analysis to examine the association between the LTCF evaluation of ability to consent and BIMS scores. Age, gender, the facility and number of years spent in the LTCF were included in the logistic regression model to account for confounding factors. Crude odds ratios did not account for confounding factors. LTCF#9 provided only consentable residents, and was not included in the logistic regression.
Finally, a logistic regression analysis was performed to examine the association between LTCF determination of ability to consent and an existing diagnosis of dementia. The regression model was adjusted for age, gender, the facility and number of years spent in the LTCF factors.
Residents who could not be interviewed by LTCF staff using the BIMS examination were not included in the logistic regression analyses. All calculations were performed using STATA 15.1 (StataCorp LLC, TX). The significance threshold was set at P < .05.
3.1 Resident characteristics
Our sample included 284 residents, 113 were perceived as able to consent by LTCF personnel, and 171 were perceived as unable to consent. 72.5% were women, which is in accordance with the results of the 2004 National Nursing Home Survey. The mean age was 83.4 years, and 98.6% were English speaking. The mean LOS in the LTCF was 2.8 years. BIMS scores were available for 232 residents, because 52 were cognitively unable to complete the BIMS test. Of the residents with BIMS scores, 121 were cognitively intact, 56 were mildly impaired, and 55 severely impaired. Tables 1 and 2 summarize all demographic data for the population we studied.
3.2 LTCF determination of resident ability to consent
We received lists from 10 LTCFs. Despite all 10 facilities reported use of BIMS to assess resident cognitive capacity as part of the federally mandated minimum data set (MDS) for LTCFs, 9 out of 10 facilities indicated that the BIMS score was based on subjective knowledge and interaction with the resident. LTCF#1 solely used BIMS scores to determine resident ability to consent.
3.3 Residents perceived as able to consent by LTCF staff
Among residents perceived as consentable by LTCF staff, BIMS scores were on average over 13 in 7 facilities (Table 1). Thus, in these facilities, on average cognitively intact residents were considered consentable. However, the range indicates that some residents with a BIMS score of mild or severe impairment were also considered consentable. In 3 other facilities, residents’ BIMS scores ranged from 10.0 to 12.23. In these facilities, on average mildly impaired individuals were perceived as able to consent, as well as some severely impaired and cognitively intact individuals.
Interestingly, 2 LTCFs (LTCF#1, LTCF#4) elected to not assess resident ability to consent without the family's permission. Therefore, all residents’ families were called asked if they wanted to meet with us to provide consent. Some families wanted to meet with us first, whereas others instructed the LTCF staff to assess resident ability to consent. When families met with us some allowed the resident to consent, but some did not. Thus, individuals able to consent were not provided the autonomy to do so. BIMS score ranges (Table 1) indicated that on average cognitively intact residents (BIMS≧13) were perceived as consentable in LTCF#1, but also some mildly and severely impaired individuals. Moderately (8≧BIMS≦12) and severely (1≧BIMS≦7) impaired residents were perceived as able to consent by LTCF staff in LTCF#4. Figure 1 further illustrates these points, where 14% mildly and 5% severely impaired individuals were perceived as able to consent. These observations are in agreement with the responses LTCF personnel provided on the self-administered questionnaire.
3.4 Residents perceived as unable to consent by LTCF staff
Among residents perceived as non-consentable by LTCF staff, residents who because of their severely impaired cognition could not be interviewed (31%) were not included in the statistical analyses.
BIMS scores ranged from 8.0 to 12.23 in 5 facilities, with a lower and upper limit of 2 and 15 respectively, for those who could be interviewed. Thus, on average, residents who based on their BIMS score were mildly impaired (24%) were perceived as unable to consent. Furthermore, some intact (16%) and severely impaired residents (29%) were perceived as unable to consent (Fig. 1). Supplemental Digital Material, Table 3, http://links.lww.com/MD/E260 details the reasons why 16% of residents with an intact BIMS score were perceived as unable to consent. In the remaining 5 facilities, BIMS scores ranged from 6.56 to 7.83, with a lower and upper limit of 2 and 14 respectively. Essentially, on average, severely impaired individuals were considered unable to consent by LTCF staff. However, some mildly impaired and cognitively intact individuals were also perceived as unable to consent.
3.5 Association of LTCF determination of ability to consent and BIMS score
A Student's t test showed that there was a statistically significant difference between the mean BIMS scores of residents perceived as consentable and non-consentable by LTCF staff (P < .001). Consentable subjects had a mean BIMS score of 13.7 (95% CI, 13.2–14.2), compared to 8.4 (95% CI, 7.7–9.2) for non-consentable residents. We stratified the subjects by facility and found no statistically significant differences between mean BIMS scores of residents perceived by LTCF staff as consentable or non-consentable among facilities.
The logistic regression analysis showed that consentable residents were more likely to have higher BIMS scores compared to non-consentable residents (OR: 1.43, 95% CI, 1.27–1.61) (Table 3). For each unit increase in BIMS scores the odds of being consentable increased by a factor of 1.43, meaning that the resident was 43% more likely to be considered consentable by the LTCF. There was no statistically significant finding between age, sex, the facility or the length of stay in the LTCF, and LTCF determination of ability to consent (Table 3). When no confounders were considered, BIMS scores remained significantly positively associated with LTCF assessment of resident ability to consent (OR: 1.48, 95% CI, 1.34–1.64) (Table 3).
3.6 Association of LTCF determination of ability to consent and dementia
A logistic regression analysis showed that a previous diagnosis of dementia was significantly negatively associated with LTCF personnel assessment of resident ability to consent (OR: 0.21, 95% CI, 0.11–0.42) (Table 4). Thus, residents without dementia were 78.9% more likely to be perceived as consentable by LTCF staff. Furthermore, age had a negative and statistically significant association with LTCF assessment of ability to consent (OR: 0.96, 95% CI, 0.93–0.99). Finally, there was a positive and statistically significant association between residents with dementia who live in LTCF#2 and LTCF assessment of ability to consent (OR: 5.14, 95% CI, 1.48–17.82).
3.7 Self-administered questionnaire
The LTCF staff reported that cognitive impairment and family involvement are challenges they face when trying to provide residents with more control over their medical decision making. Furthermore, lack of trust, misinformation and fear of experimentation affect a resident's desire to consent to research. When asked how one can help residents have a voice for themselves, most LTCFs stressed the need to offer more ability to make decisions. However, they were unable to provide concrete ways to achieve this objective. All LTCFs reported fulfillment of state department requirements, as the main reason for participating in a clinical study. Finally, helping residents advocate for themselves was consistently a major challenge. The questionnaire is provided in the Supplemental Digital Material, Table 1, http://links.lww.com/MD/E258.
In the present study, we found that there was a statistically significant relationship between LTCF assessment of resident ability to consent and BIMS scores, as well as with a previous diagnosis of dementia. Therefore, despite 9 out of 10 LTCFs using subjective means to assess residents’ cognitive abilities, their “consentability” assessment was associated with BIMS scores (aOR: 1.43, 95% CI, 1.27–1.61). This result of the regression effectively allows us to suggest that the assessment was accurate overall, where the higher the BIMS score, the more likely the LTCFs were to assess the resident as able to consent. Furthermore, we found a statistically significant relationship between diagnosed dementia and LTCF assessment of resident ability to consent. There was no statistical difference between the recruited LTCFs. However, among residents perceived as unable to consent by LTCF personnel, 17% were cognitively intact based on their BIMS scores.
Ability to consent is a key tenet of the ethical conduct of human research, and spans both ethical and legal dimensions. Cognitive capacity is the core requirement for IC; that is, whether an individual can provide consent, or whether consent must be obtained from a proxy. IC is particularly complex for individuals who are elderly, live in a nursing facility, or suffer from dementia.[4,17] Historically, LTCFs have defaulted to seeking IC from a proxy, especially in cases of patients with dementia.[18–20] This presumption takes away a resident's voice, therefore disregarding their autonomy and privacy. Moreover, involvement of a proxy assumes ability on the part of the proxy to accurately reflect the resident's wishes. As detailed above, LTCF#1 and LTCF#4 decided to consult with proxies before allowing us to approach the residents or their responsible parties. Interestingly, both of these facilities assessed cognitively intact individuals as unable to consent. This could indicate that the decision by LTCF personnel to consult proxies despite the residents’ cognitive ability to consent, took away their right to choose for themselves.
To date, there is no standard used to determine capacity to give IC. Additionally, there is no standard for the process of IC that LTCFs can follow when determining participation in research. Limited guidance exists for projects including individuals with dementia. The US National Institutes of Health produced a detailed document in 2009 describing various ways that would be used to determine capacity to consent. Although, they did not recommend one particular technique, they did point out that capacity should be assessed for a specific research study and not broadly. Thus, it is encouraging that our data showed that LTCF staff did not perceive residents as unable to consent on the basis of a diagnosis of dementia.
Due to its broad usage throughout the recruited LTCFs, we chose the Brief Interview for Mental Status (BIMS) examination as an assessment of cognitive function. BIMS is a rapidly and widely administered instrument used as part of the MDS —a federally mandated process that measures the health status of LTCF residents. A recent study showed that BIMS can predict cognitive diagnoses, however, the Brief Cognitive Assessment Tool-Short Form (BCAT-SF) is more sensitive and able to differentiate among all cognitive levels. The Mini Metal Status Examination (MMSE) is a “practical method” to assess cognition. However, MMSE also has a low sensitivity and specificity in terms of capacity. Finally, the Aid to Capacity Evaluation (ACE) is suggested as the best available instrument to assist clinicians in assessing medical decision-making capacity. The ACE is the only instrument accompanied by training materials. We do not recommend the BIMS examination as the standard to assessing capacity to IC.
Standardizing the process of IC in LTCFs can enhance the ability to perform research with LTCF subjects. The implementation of a commonly accepted, objective and easy to use assessment, such as the BIMS examination, can be a step in the right direction. Especially for minimally invasive studies, such as the one described here, being cognitively intact (i.e., BIMS≧13) should afford residents the ability to decide if they would like to participate in a research study. However, the limitations of the BIMS instrument, and current research suggesting that cognitive testing alone is insufficient to assess the presence of capacity, tell us that new and more holistic approaches to assessing individuals’ consent to participate in research are required.
Researchers in future studies could use more composite cognitive assessment tools to assess resident ability to consent. Warner et al establish the use of 4 key elements when assessing capacity to consent —providing salient and comprehensible information, allowing time to understand and retain the information, testing retention and belief, and assessing ability to weight the information. Approaching all residents without having been influenced by the nursing staff’ judgment of residents’ cognitive state, and assessing the aforementioned elements will allow for a more objective assessment of consent. All the while increasing resident autonomy to providing consent in different circumstances.
This study has some notable limitations. First, the study was limited to residents of only 10 facilities in RI, USA, so our sample might not be representative of different populations across the state or the nation. Second, the questionnaire used was not standardized and currently in broad use, which meant providers were not familiar with it, and might have had difficulty responding to all questions. Third, we relied on the subjective cognitive assessment we were provided by the LTCF staff, and did not administer additional tests to assess the ability to consent. Therefore, we did not adhere to the “situational” dimension of consent. LTCF staff were asked to provide lists qualifying residents as able or unable to consent, but we did not visit the residents to determine whether, despite potential cognitive incapacities they might have been able to respond to this particular project. Lastly, our study followed an observational design, and thus we cannot be certain that we have accounted for all potential confounding factors.
5 Conclusions and implications
As the population of adults living in LTCFs is rising, quality of care and research should be top priorities.[29,30] An evidence-based determination of the challenges (infections, living conditions, medication over-prescription, etc.) that residents face, will facilitate the advancement of care in LTCFs and other long-term care settings. Autonomy to consent might be another significant challenge that LTCF residents face. Therefore, the present study hopes to provide insight into the process of IC and the associated hardships, effectively providing directives for future research in similar settings to further resident autonomy and quality of care.
The authors would like to thank the facilities, staff, residents and families who volunteered their time to participate in this research.
Conceptualization: Eleftherios Mylonakis.
Data curation: Katerina Tori, Markos Kalligeros, Fadi Shehadeh, Rajamohammed Khader, Aman Nanda, Robertus van Aalst, Ayman Chit, Eleftherios Mylonakis.
Formal analysis: Katerina Tori, Markos Kalligeros, Fadi Shehadeh, Rajamohammed Khader, Aman Nanda, Robertus van Aalst, Ayman Chit, Eleftherios Mylonakis.
Funding acquisition: Robertus van Aalst, Ayman Chit, Eleftherios Mylonakis.
Project administration: Eleftherios Mylonakis.
Supervision: Eleftherios Mylonakis.
Validation: Eleftherios Mylonakis.
Writing – original draft: Katerina Tori, Markos Kalligeros, Fadi Shehadeh, Rajamohammed Khader, Aman Nanda, Robertus van Aalst, Ayman Chit, Eleftherios Mylonakis.
Writing – review & editing: Katerina Tori, Markos Kalligeros, Fadi Shehadeh, Rajamohammed Khader, Aman Nanda, Robertus van Aalst, Ayman Chit, Eleftherios Mylonakis.
. Smith SK, Rayer S, Smith EA. Aging and disability: implications for the housing industry and housing policy in the United States. J Am Plan Assoc 2008;74:289–306.
. Mark Mather LAJ, Kelvin MP. Aging in the United States. Population Reference Bureau 2015;70:
. Resnick B, Gruber-Baldini AL, Pretzer-Aboff I, et al. Reliability and validity of the evaluation to sign consent
measure. Gerontologist 2007;47:69–77.
. Gruber-Baldini AL, Zimmerman SI, Mortimore E, et al. The validity of the minimum data set in measuring the cognitive impairment of persons admitted to nursing homes. J Am Geriatr Soc 2000;48:1601–6.
. Mezey M, Teresi J, Ramsey G, et al. Decision-making capacity to execute a health care proxy: development and testing of guidelines. J Am Geriatr Soc 2000;48:179–87.
. Brod M, Stewart AL, Sands L, et al. Conceptualization and measurement of quality of life in dementia
: the dementia
quality of life instrument (DQoL). Gerontologist 1999;39:25–35.
. Buckles VD, Powlishta KK, Palmer JL, et al. Understanding of informed consent
by demented individuals. Neurology 2003;61:1662–6.
. Informed consent
for research on human subjects with dementia
. AGS Ethics Committee. American Geriatrics Society. J Am Geriatr Soc 1998;46:1308–10.
. Appelbaum PS, Grisso T. The MacArthur Treatment Competence Study. I: Mental illness and competence to consent
to treatment. Law Hum Behav 1995;19:105–26.
. Thorogood A, Maki-Petaja-Leinonen A, Brodaty H, et al. Consent
recommendations for research and international data sharing involving persons with dementia
. Alzheimers Dement 2018;14:1334–43.
. Ponraj M, Dubashi B, Harish BH, et al. Cefepime vs. cefoperazone/sulbactam in combination with amikacin as empirical antibiotic therapy in febrile neutropenia. Support Care Cancer 2018;https://www.medicare.gov/nursinghomecompare/search.htmlhttps://www.medicare.gov/nursinghomecompare/search.html
? Accessed January 23 2019.
. 2008;Saliba D, Buchanan J. Development and Validation of a Revised Nursing Home Assessment Tool: MDS 3. 0.
. Thomas KS, Dosa D, Wysocki A, et al. The minimum data set 3.0 cognitive function scale. Med Care 2017;55:e68–72.
. Kuderer NM, Dale DC, Crawford J, et al. Mortality, morbidity, and cost associated with febrile neutropenia in adult cancer patients. Cancer 2006;106:2258–66.
. Ioannidis JP. Informed consent
, big data, and the oxymoron of research that is not research. Am J Bioeth 2013;13:40–2.
. Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the infectious diseases society of America. Clin Infect Dis 2011;52:e56–93.
. Hougham GW. Waste not, want not: cognitive impairment should not preclude research participation. Am J Bioeth 2005;5:36–7. author reply W15-38.
. Hubbard G, Downs MG, Tester S. Including older people with dementia
in research: challenges and strategies. Aging Ment Health 2003;7:351–62.
. Jongsma KR, van de Vathorst S. Beyond competence: advance directives in dementia
research. Monash Bioeth Rev 2015;33:167–80.
. Kim SY, Kim HM, Ryan KA, et al. How important is ’accuracy’ of surrogate decision-making for research participation? PLoS One 2013;8:e54790.
. Beattie E, O’Reilly M, Fetherstonhaugh D, et al. Supporting autonomy of nursing home residents with dementia
in the informed consent
. Warren JW, Sobal J, Tenney JH, et al. Informed consent
by proxy. An issue in research with elderly patients. N Engl J Med 1986;315:1124–8.
. Nakane T, Tamura K, Hino M, et al. Cefozopran, meropenem, or imipenem-cilastatin compared with cefepime as empirical therapy in febrile neutropenic adult patients: a multicenter prospective randomized trial. J Infect Chemother 2015;21:16–22.
. Saliba D, Buchanan J. Development and Validation of a Revised Nursing Home Assessment Tool: MDS 3.0. 2008.
. Mace RA, Mansbach WE, Clark KM. Rapid cognitive assessment of nursing home residents: a comparison of the brief interview for Mental status (BIMS) and brief cognitive assessment tool-short form (BCAT-SF). Res Gerontol Nurs 2016;9:35–44.
. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98.
. Sessums LL, Zembrzuska H, Jackson JL. Does this patient have medical decision-making capacity? JAMA 2011;306:420–7.
. Warner J, McCarney R, Griffin M, et al. Participation in dementia
research: rates and correlates of capacity to give informed consent
. J Med Ethics 2008;34:167–70.
. Maas ML, Kelley LS, Park M, et al. Issues in conducting research in nursing homes. West J Nurs Res 2002;24:373–89.
. Dellefield ME. The relationship between nurse staffing in nursing homes and quality indicators. J Gerontol Nurs 2000;26:14–28.