A comprehensive intervention following the clinical pathway of eating and swallowing disorder in the elderly with dementia: historically controlled study

Eating problems in patients with advanced dementia are recognized as one of the greatest risk factors for their survival [1]. Eating problems including disturbed eating behavior, dysphagia, nausea, and anorexia, are associated with various pathological conditions such as dementia, stroke, sarcopenia, mental illness, and systemic diseases that disturb many physical functions [2‐5]. Moreover, these pathological conditions have various underlying etiologies [6‐8], all of which can cause decreased oral intake. However, the etiology of eating problems in each patient is often unclear because most patients with advanced dementia are facing the end-of-life and are usually recommended to receive only palliative or hospice care [9]. Decreased oral intake in patients with a disorder in the anticipatory phase may not only be associated with advanced dementia but also with reversible conditions such as anorexia caused by infectious diseases, acute organ failures, acute exacerbation of chronic diseases, as well as reversible dementia [10]. However, these patients’ eating abilities may improve, leading to better prognosis, and targeted interventions might reinstate them from facing the end-of-life. Keeping oral intake at the level of independence from artificial hydration and/or nutrition (AHN) might help maintain quality of life even at the end-of-life. Although AHN is not usually recommended for advanced dementia, it can assist with nutritional intake predominantly caused by a potentially reversible condition [11].

Meta-analysis showed almost no evidence for the effectiveness of specific interventions to directly support food and drink intake in people with dementia [12]. The single interventional approach applied for each study included in the meta-analysis might have been inadequate for improving oral intake [12] because the etiologies of decreased oral intake were heterogeneous and had not been properly assessed. Inadequate assessment of decreased oral intake may lead to inaccurately concluding that patients remain in an irreversible state, resulting in them receiving inappropriate AHN such as tube feedings. A national Japanese survey showed that approximately 200,000 dysphagia patients in geriatric long-term settings were dependent on tube feedings despite not having undergone adequate evaluation of swallowing function [13].

Comprehensive geriatric assessment (CGA) has been shown to be effective for improving functional status and prognosis in elderly patients [14‐16]. In particular, CGA may clarify the etiologies of eating problems and provide individualized intervention with the potential to improve dementia patients’ decreased oral intake. Therefore, we developed a new medical care system based on CGA for severe eating problems in patients with dementia and investigated whether it could improve their clinical outcomes.

The present study showed that the CGA with a multidisciplinary team could determine the etiologies of ESDED in detail, which could help independence from AHN and improve AHNFS without reducing OS. This result indicates that focusing on the etiology of ESDED is important for improving clinical outcomes.

Many studies have verified the effectiveness of different interventions for oral intake in various care settings. However, Abdelhamid et al. reported finding no definitive evidence on the effectiveness, or lack of effectiveness, of specific interventions; furthermore, study samples were small and observation periods were short [12]. More patients with underlying conditions of acute or reversible diseases should have been included in previous studies. These reversible conditions in the elderly with eating problems and dementia are usually difficult to find because they typically cannot convey their symptoms. Therefore, the heterogeneity in etiologies of ESDED might have hampered the effect of specific simple intervention in previous studies. An assessment that estimates the causes or etiologies of ESDED can be considered essential to improving oral intake in ESDED. Especially when evaluating and managing dysphagia, a multidisciplinary approach is recommended; support for team approaches has already been reported [22‐24]. Interventions such as postural adjustments, direct support of oral intake or modification of foods/liquids were not always regarded as a basic therapeutic treatment but were traditionally considered to be encompassed in symptomatic treatments or compensatory management [25]. Furthermore, it is difficult to make swallowing rehabilitation effective in patients with advanced dementia, such as those in our study populations, as shown by the low RR of the control group (Table 2). Conversely, assessment based on CGA could clarify the etiology of decreased oral intake, which helped us to individualize care and treatment. In addition, the interventions could become comprehensive with broad insights provided by a multidisciplinary team. Several advantages to multidisciplinary team interventions by refining health professionals’ roles and responsibilities within transition models in acute care settings have been reported, including: reducing emergency department re-admission rates, and reducing mortality and functional decline of older people [26]. As a result, our unique intervention method is more effective than specific interventions with a single procedure.

Although the methods for direct interventions for patients utilized in this study varied, the evaluation procedure for each patient (the CGA method) was uniformly performed by using CP. CPs are used broadly in medical care and are sometimes utilized as standardized protocols for optimizing and streamlining patient care [27]. Intervention using a CP was proven to be beneficial with a multidisciplinary team approach [28], and therefore we considered it was suitable for eating problems in patients with dementia. Additionally, the CGA method of this study is reproducible.

The BI at baseline, a physical function indicator, was significantly higher in the intervention group than in the historical group. Moreover, the proportion of those with stroke, an underlying disease that could affect the onset of ESDED, was significantly lower in the intervention groups (Table 1). However, neither BI nor stroke was a significant risk factor for independence from AHN according to logistic regression analysis (Fig. 3). Therefore, the higher BI and low frequency of stroke in the intervention group could not have contributed to the higher RR. Rather, sex, serum albumin concentrations at baseline, and the CGA strongly affected independence from AHN.

This study has some limitations. First, the factor that directly improved RR was unclear in each case. However, our new medical care system based on CGA has strong potential to improve the ability of eating and functional prognosis in patients with ESDED, and thus improve some survival measures (Fig. 4). Because all medical staff and caregivers in the two compared groups were almost identical with regard to their professional expertise and organizational structures, the new medical care system alone could be considered as having clear favorable outcomes for ESDED. Other outcome measures also support the effectiveness of our new medical care system (Table 3). These outcome measures in our study have not previously been reported in this field. We selected the primary and secondary outcome measures because the most valuable outcome measures for our study population were thought to be quality of life and prognosis. Although body weight is often selected as one of the index values concerning nutrition and outcome measures in patients with eating problems, the weight data could not be collected, especially in the control group because many of the patients were bedridden.

Second, the present study could not be carried out as a randomized controlled trial (RCT). Previous RCTs showed no evident effectiveness of a multidisciplinary team approach with CGA for elderly patients [16]. Considering the negative results from previous RCTs, we assumed that various medical care routinely provided by a trained multidisciplinary team in the control group might impair the strength of the interventions in the intervention group; thereby reducing the difference of outcome measures between the two groups in RCT. Furthermore, RCT could not be ethically approved for us to refrain from providing the best care or every possible method for treating ESDED in the control group because their prognosis was known to be poor. Therefore, we conducted this study as a historically controlled trial. Most studies evaluating the effect of a multidisciplinary team approach on dysphagia also treated retrospective datasets [22, 23, 29], pointing out that it was difficult to conduct a RCT using a multidisciplinary team approach [29].

Third, a selection bias may exist in this study because the attending physicians declined to enroll some participants (Fig. 2). The most common reason for refusal appeared to be that the physicians were concerned about their medical treatment being disturbed by ESAT or the study protocol. It is important to highlight that the physicians did not select patients for enrollment based on potential degree of benefits that could be obtained from this study. Therefore, it is unlikely that the physicians’ refusal to enroll some patients significantly affected the preferable results in the intervention group.

Fourthly, the decision to withdraw from AHN was dependent on each attending physician because we did not set a target value for intake. A universal target value of oral intake was difficult to set accurately for all subjects, despite established caloric targets being known to be associated with improved clinical outcomes [30‐32]. Rather, we monitored survival rate to examine the validity of each withdrawal of AHN. If required AHN were inappropriately withdrawn, survival rate would reduce. In our study, however, overall survival was not shown to be significantly different between the two groups despite the intervention group achieving a higher RR from AHN (Fig. 4). Thus, AHNFS could not be analyzed precisely. However, we clearly demonstrated that CGA with multidisciplinary interventions contributed to improving the functional status of eating, resulting in a better independence rate from AHN. More studies with high evidence levels are required to conclusively determine the usefulness of comprehensive intervention for ESDED. We believe that higher quality assessments and interventions based on CGA will bring more favorable outcomes and reduce inadequate AHN for ESDED.

Ethical issues associated with AHN in elderly patients with advanced dementia are becoming hot topics in aging societies [33, 34], where physicians have been struggling with end-of-life care for these patients [35]. One embarrassing issue that remains to be solved is difficulty in judging reversibility of decreased oral intake in these patients, although their eating problems indicate poor prognosis [1]. Physicians may thus perform futile life-prolonging treatments with AHN or provide inappropriate end-of-life care without efficient interventions [13]. In addition, insufficient evidence exists regarding the outcome of AHN in older people with advanced dementia [36, 37]. However, we proved through this study that by performing the best assessment with the best procedure and that by providing appropriate care for each patient helped us to precisely judge reversibility. Furthermore, reversibility could improve the functional status of eating and help ESDED cases survive without the need for AHN. These two effects might assist in decision making for indications of AHN.

A CGA with multidisciplinary interventions could improve the functional status of eating in elderly patients with severe eating problems and dementia to help them become independent from AHN. Therefore, we should not regard patients with ESDED as facing the end-of-life before adapting CGA for them.