Quality of care in long-term care institutions: A scoping review of literature

This review sought to answer the question “What theoretical models have been used to define and evaluate the quality of care provided to older adults in long-term care institutions?” The employed study methodology followed the recommendations of the Joanna Briggs Institute Reviewers Manual. Using MESH terms and keywords such as “elderly,” “long-term care,” and “theoretical,” four electronic databases (i.e., PubMed, Scopus, Web of Science, and LILACS) were searched for articles in Portuguese, English, and Spanish, without a time frame restriction. Titles and abstracts were independently screened by two reviewers, followed by a full-text review. A total of 1,211 articles met our inclusion criteria, 80 were selected for full reading, and 21 were included for qualitative synthesis. The theoretical models cited in the studies included the multidimensional model (n = 10); the structure, process, and results model (n = 8); a theory centered on the person (n = 2); and one centered on the work environment (n = 1). Few articles used a conceptual model as the basis for assessing long-term care institutions, which is important for the construction of instruments and indicators that assess the quality of care.


Introdution
Population aging is a worldwide phenomenon that occurs in conjunction with an increasing number of older adults with functional impairment and dependence on daily activities. These older adults need continuous and long-term care that is not always offered by their families. In this context, institutionalization has become an alternative for families (Joshua 2017; Burke & Werner 2019).
Globally, there are different types of long-term care institutions (LTCIs) for older adults, with different denominations but common characteristics. These institutions are collective residences for older adults, with or without the need for assistance in activities of daily living (ADLs) or with behavioral problems. The function of LTCIs is to supply homes where the resident can have protection, citizenship, freedom, family support, and the chance to preserve functional status (Onder et al. 2012;Sanford et al. 2015;WHO 2017;Trinkoff et al. 2020).
To verify the quality of care offered by LTCIs, many types of assessments have been developed and applied to guide the choices of residents or the investments by local governments (Spasova et al. 2018;Scheffelaar et al. 2019;Milte et al. 2019;Wagner et al. 2020;Sion et al. 2020), and the use of multiple dimensions and indicators for assessment is recommended (Gilissen et al. 2017;Dyer et al. 2019).
To construct evaluation instruments, it is vital to have a theory to direct the research, guide what will be evaluated, and guide how something will be measured. The choice of the theoretical framework to achieve this requires in-depth knowledge of the topic being studied (Adom et al. 2018;Collins et al. 2018;Varpio et al. 2019). According to the World Health Organization (WHO), a definition of quality must be effective, efficient, accessible, patient-centered, fair, and safe (WHO 2006). The mapping of theoretical models used by researchers to assess the quality of care in LTCIs systematizes the theory employed and contributes to the comparison of them and the choice of the theoretical framework for the design of future studies. Therefore, the objective of this study was to map the theoretical frameworks used to define and assess the quality of care in LTCIs.

Methodology
The present study is a scoping review conducted according to the methodological guidelines of the Joanna Briggs Institute Reviewers Manual (Peters et al. 2015). This type of research map key concepts, summarize theories and clarify the conceptual limits and gaps in the scientific literature (Levac et al. 2010;Tricco et al. 2016;Munn et al. 2018). The Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist was employed in this review.
The following methodological steps were followed: (1) identification of the research problem, (2) development of the eligibility criteria, (3) definition of the search strategy, and (4) qualitative synthesis of the results (Armstrong et al. 2011;Peters et al. 2015).

Research problem identification
This research was conducted due to the need to identify theoretical models to evaluate the quality of care in the context of LTCIs. The PCC framework (population, concept, and context) structured the construction of the research question.
Meanwhile, the population included institutionalized older adults, the concept refers to the theoretical models used to guide the assessment of the quality of care, and the context was the LTCIs themselves (Peters et al. 2015). Therefore, the research question was defined as follows: What theoretical models have been used to define and assess the quality of care for older adults in LTCIs?

Development of eligibility criteria
The eligibility criteria for inclusion in our analysis were (1) studies in English, Portuguese, and Spanish; (2) original studies with a quantitative or qualitative approach; (3) studies that presented a theoretical model for assessing the quality of care in an LTCI, and (4) studies conducted in an LTCI. There was no time limit restriction for studies. Editorials, book chapters, dissertations, and conference abstracts were excluded.

Search strategy
The search was carried out between July 2020 and February 2021 in the following electronic databases: Web of Science, PubMed, LILACS, and Scopus. The search strategy included descriptors indexed in the Medical Subject Headings (MESH) joined by Boolean AND and OR. The descriptors were separated into four blocks (quality of care, LTCI, older adults, and theoretical model) and the search strategy was adapted according to each electronic database (Table 1). A manual search was also carried out by consulting the list of references of the included articles.

Scopus
( ( "quality of care" OR "quality of healthcare" OR "quality of health care" OR "quality indicator" OR "quality measure*" OR "quality assess*" OR "quality evaluation" ) ) AND ( ( ( ( "Nursing Homes" [mesh] OR "Homes for the Aged" [mesh] OR residential AND care AND home*[title/abstract] OR "long AND term AND facilit*" OR "homes, AND nursing" OR "residential AND care" OR "residential AND care AND institutions" OR "long-term AND care" OR "long AND term AND care" OR "care AND home*" ) ) ) AND ( ( elder* OR aged OR geriatric* OR "older AND adults" OR "aging" ) ) ) AND ( ( "theoretical model" ) OR ( "theoretical framework" ) OR ( "theory" ) OR ( "measure" ) OR ( "models theoretical" ) OR ( "theoretical study" ) OR ( "conceptual model" ) AND ( "conceptual framework" ) )

LILACS
( ( "quality of care" OR "quality of healthcare" OR "quality of health care" OR "quality indicator" OR quality measure* OR "quality assess*" OR "quality evaluation" ) ) AND ( ( ( ( "Nursing Homes" OR "Homes for the Aged" OR residential AND care AND home* OR long AND term AND facilit* OR "homes, AND nursing" OR "residential AND care" OR "residential AND care AND institutions" OR "long-term AND care" OR "long AND term AND care" OR "care AND home*" ) ) ) AND ( ( elder* OR aged OR geriatric* OR "older AND adults" OR "aging" ) ) ) AND ( ( "theoretical model" ) OR ( "theoretical framework" ) OR ( "theory" ) OR ( "measure" ) OR ( "models theoretical" ) OR ( "theoretical study" ) OR ( "conceptual model" ) AND ( "conceptual framework" ) ) Source: Authors.

Qualitative summary of results
Relevant studies were initially identified. The software Microsoft Excel® was used to identify and exclude duplicate articles. Two independent researchers (BLCV and ACM) selected the studies. The titles and abstracts were read for an initial screening, and the selected articles were read in full by both researchers. Disagreements about inclusions were discussed between the two researchers until a consensus was reached; if consensus could not be reached, disagreements were discussed with another researcher. According to the guidelines of a scoping review, an assessment of the quality of the selected articles was not carried out.
Information was then extracted from the selected studies using a table in Microsoft Excel® with the following information: author, year of publication, the country where the research was conducted, objectives, sample composition, variables used, data collection method, and dimensions of quality of care used to assess the quality of care in LTCIs. Theories were identified as multidimensional, when indicated in articles with that denomination or when citing the dimensions used to assess quality; structure-process-result, when the author used that term or indicated the study to follow Donabedian theory; person-centered, when it was described as based on consumer preference or satisfaction; and worker-centered, when the evaluation involved the worker's action on residents' well-being.

Results
A total of 1,216 studies were retained; 7 were duplicated. Of the studies, 1,204 were screened by reading the titles and abstracts, and 80 were screened by reading the full texts. Five more studies were selected from the references of the articles selected for a complete reading ( Figure 1). The search sequence, including and excluding criteria, was described in the flow chart ( Figure 1). The selected articles were separated according to the theories used. Table 2 shows the information about the author(s), research site, type of study, objectives, sample composition and variables. After screening and applying the elegibility criteria, 21 articles were included in the review, most of which were cross-sectional (n = 13) and qualitative (n = 6); only two were longitudinal. Most studies were conducted in a single country, most often in the United States (n = 10), the United Kingdom, Denmark, Finland, Israel, Lebanon, and Switzerland. There were five multicenter studies: one involving the Czech Republic, Finland, France, Germany, Italy, Israel, the Netherlands, and the United Kingdom; another involving Denmark, Iceland, Italy, and the United States; another involving the United States and Finland; another involving the United States, Canada, and Iceland; and another involving the United States and Iceland ( Table 2).
The theoretical models used were structure-process-results, multidimensional, person-centered, and work-centered.

Theoretical models
We identified 10 studies that used the multidimensional model; eight that used the structure-process-result model (more popularly known as the Donabedian model), two were based on the person-centered model, and one was based on the work environment. One study related the two theories that it employed (multidimensional and structure-process-result).
Articles were published between 1990 and 2019.
The articles that used multidimensional models addressed several dimensions of care, including residents and family members, workers and managers, the LTCI`s physical structure, the quality of care, the involvement with the family, and the community. Studies that used the Donabedian model (structure-process-result) assessed the results of the research to verify the quality of care. Meanwhile, studies that adopted the person-centered model considered well-being and quality of life as indicators of good quality. Finally, those that employed the workplace-based approach considered the psychosocial incentive of workers as a key to efficient care at LTCIs.

Dimensions of theoretical models for quality assessment and measurement using indicators
In the structure-process-result model, the dimensions involved several indicators. The structure assessment verified the number of beds (vacancies) at the LTCIs and the qualification of workers (n = 5), as well as the final medical cost per older adult (n = 2). The work process emphasized communication (n = 2) and an adequate number of workers (n = 2). The expected results were defined using indicators referring to the patients' clinical conditions, such as whether they had experienced pressure ulcers, falls, incontinence, infections, cognitive decline, and weight loss.
The dimensions cited in the multidimensional model were care, health status, functional status, well-being, medical services, care planning, work team, family involvement, communication, environment, cost, security, social environment, administrative leadership, culture, positive experience, and effectiveness. The indicators used in these studies were professional performance in providing care (n = 5), family involvement (n = 3), user satisfaction (n = 2), structure (n = 4), recreational activities and social interaction (n = 3), and cost (n = 1).
The person-centered model adopted user satisfaction and family involvement as dimensions. The indicators used in these studies were consumer satisfaction, preferences, regular contact with the family, choice of food, choice of music, regular contact with friends, watching television, having privacy, going outdoors, giving gifts, choosing the time of day to bathe, and activities outside the LTCIs.
The work-centered model only evaluated professional performance. The indicators used were the possibility of development at work, influence, leadership, meaning at work, commitment, and clarity of the worker's role.
Studies by Phillips et al. (1997), Frijters et al. (2013), and Grabowski (2001) specialists, but was without the participation of older adults, which is not highly valued by this model.
Some studies have attempted to connect the variables used with actual improvement in quality of care. The variables that had a positive relationship with quality were the workers' psychosocial resources and favorable work environment (Winsløw;Borg 2008), individualized treatment, directing the focus of care to older adults (Abbott et al. 2018), quality of life (Malley et al. 2019), the work process and work teams (Kajonius & Kazemi 2016), and safety culture (Thomas et al. 2012).

Discussion
Measuring the quality of services offered at LTCIs has been a challenge worldwide. The lack of uniformity regarding the concept of quality for LTCIs and the use of multiple dimensions are just some of the difficulties faced. It is important to integrate the perspectives of older adult residents, family members, and professionals when investigating the concept of quality to build tools for assessing the quality of LTCIs (Nakren et al. 2008), because care in LTCIs involves several subjects as residents, workers and managers (Frytak et al. 2001).
In this regard, the multidimensional model assumes that the quality of care needs to involve multiple dimensions.
Some studies carried out qualitative research seeking to understand the perceptions of quality from specialists, care providers, care users, and family members to establish these dimensions. The most cited dimensions were the environment, family involvement, human resources, home, individualized care, communication, and a central focus on residents, family, and the community (Gustafson et al. 1990;Rantz et al. 1999;Thompson et al. 2012). From this, the QAI (Gustafson et al. 1990) and OIQ were constructed for researchers, users, and regulators interested in observing and evaluating the quality of care in LTCIs, based on a multidimensional theoretical model (Rantz et al, 2000;2002). The studies that used this model were more comprehensive, with multiple dimensions, with regard to the evaluation of care.
The use of the structure-process-result model was justified by the authors that employed the model by the need for an evaluation that would guarantee the delivery of good quality service to residents. It connects structure, process, and results as the key to achieving good quality, and each of these dimensions has several indicators. The structure involves material, human, and organizational resources; the process involves activities related to care itself; the results refer to the effect of care on the patient's condition (Fleishman 1988). The most common criticism regarding the use of the Donabedian model is that it is adapted from the industry that aims to produce objects and not services, without having as a central point the health and wellbeing of people (Unruh & Wan, 2004;Harrington et al. 2005), and that there is not enough detail to truly meet the needs of users or to develop a care plan (Degenholtz et al. 2014). In general, studies that support this model use fewer comprehensive variables than multidimensional models. None of the included studies evaluated family involvement or recreational activities.
Many studies using secondary data and normally available tools are based on this theory, such as the Minimum Data Set (MDS) (Weech-Maldonado et al. 2004;Thomas et al. 2012) and OSCAR (Grabowski, 2001;Weech-Maldonado, et al. 2004;Laberge et al. 2008). Some authors, even without supporting this theory, used the same data. Generally, these tools are used to define financing and to compare the LTCI's quality services (Fleishman 1998;Laberge et al. 2008;Kang et al. 2011;Thomas et al. 2012;Weech-Maldonado et al. 2004).
In the person-centered model, the first study found that the quality of care is based on the person and his well-being, and quality of life is a frequently used indicator (Lowe et al. 2003). The second address the patient's holistic view, and his physical, mental and psychosocial health is considered to meet the individuality and desire of the person (Abbott et al. 2018 Organizations, 2007). In general, there are criticisms from a person-centered organization because it is believed that it is not strongly oriented toward equity in health, since it is believed that to create equitable health systems, it is essential to analyze the perception of users and those involved, as well as the workers who have the potential to identify disparities in health care and who can contribute to increasing the quality of care (Cunningham et al. 2014;David et al., 2020).
The theoretical model that related the work environment was found only in one study (i.e., Winsløw and Borg 2008), being defined as the frequency at which workers perform actions for users, which promotes well-being and quality of life. This raises the hypothesis that a sufficient number of workers with adequate qualification in an environment that favors the psychosocial aspects of these workers leads to the provision of a better quality of care (Lindolpho et al., 2020;Machado et al., 2020). A relationship was identified between psychosocial aspects (influence, possibilities of development, meaning of work, commitment, clarity of functions, predictability, and leadership) and quality of care (Winsløw;Borg 2008). Studies have shown that professional satisfaction is mediated by proactive behaviors (Khatri et al. 2016;Tourangeou et al. 2017;Backhaus et al. 2017) and that career and job satisfaction is related to the quality of service provided to the consumer (Spence et al.

2016
). However, there was no evidence of an association between quality and payment for performance (Glickman et al. 2007).
The focus on just a few indicators may not substantially improve quality; several indicators are required for this to occur (Goodson et al. 2008;Thompson et al. 2012;Kontezka, 2019, Song et al. 2019Minayo 2019).
With regard to theoretical models, most are incomplete or have not been created specifically to assess LTCIs. Difficult factors in the assessment are those intrinsic to older adults, such as age and genetic load, which cannot be controlled and affect clinical indicators (Unruh & Wan., 2004;Harrington et al. 2005). Another factor is related to the indicators themselves; generally, little has changed over time, not been validated, and not necessarily related to improvement in the quality of care For a good evaluation, it is necessary to search for concepts with a broad understanding (Bond et al. 2018) and choose the elements that will constitute the evaluation, assuming that the lowest quality is associated with inefficiency (Donabedian 1988). The evaluation, in general, should vary according to time, internal culture and also include qualitative aspects, although difficult to numerically measure (Adom et al. 2018, Collins et al. 2018, Varpio et al. 2019.
This review provides a comprehensive, relevant, and up-to-date view of the theoretical models used to assess LTCIs.
A strong point of this study is that the quality models used to evaluate LTCIs drive advances in evaluation models with positive results. Additionally, multi-base and peer searching strengthen research. There is still a long way to go to improve the quality of care assessment processes in LTCIs, which attends a population that is growing rapidly in the world.
A multidimensional model with comprehensive dimensions including all those involved (consumers, residents, and family members, managers, and workers), with an outstanding assessment of worker and resident satisfaction, would be the best way to better capture the current moment of quality of care and what will be necessary to improve it.

Conclusion
Four theoretical models were used to assess the quality of care in LTCIs: multidimensional, structure-process-result, person-centered, and centered on the work-centered.
There are dimensions and some common indicators between models. The indicators of the model work and personcentered are also part of the multidimensional and Donabedian models.
The theoretical model influences the construction of evaluation systems. The more comprehensive the assessment, the more likely the quality will be achieved.
Multidimensionality must be considered when assessing the quality of care in LTCIs, which must be performed continuously.
More studies must be developed to evaluate the quality of care offered in LTCI based on quality theoretical frameworks, considering the several dimensions of the quality concept. In the same sense, there is a need to develop valid and reliable indicators to measure these quality aspects. In this way, the evaluative studies will improve the quality of care, health, and well-being of institutionalized older adults.

Declarations
Funding -This study is part of PhD thesis developed in Universidade Federal de Minas Gerais (School of Dentistry).