The results of Study 1 helped identify the issues in information sharing, such as the lack of shared information and the discrepancies in the perception of information between facilities; it was found that inappropriate information sharing was conducted. Study 2, which was conducted to identify items that were difficult to share, quantitatively showed that the current and ideal information sharing diverged for many items in both types of facilities. That information other than the current medical information was difficult to share. The difference occurred in a different category than the one assumed (i.e., hypothesis H was rejected). Study 3, which was positioned as a follow-up survey to Study 2, showed that mutual understanding between the two types of facilities was necessary and that information sharing did not consider a variety of information; there were also difficulties in collecting and communicating information in the university hospital. From the results of these exploratory studies, we focused on the information sharing required for interfacility care and the difficulties in providing technological support.
7.1 Lack of Information
The results supporting the three hypotheses, with the exception of h4, were discussed from a statistical point of view. In almost all results of Study 2, the effect sizes ranged from small to large [82]. Thus, despite the small sample size, we found the study to be meaningful. Although h3 was corroborated, there was a possibility of the ceiling effect. Significant differences may be observed in some items by modifying the scale of the scores according to the trend of the data and re-examining it. This ceiling effect was the result of methodological limitations. We thought it would be desirable to compare it with the ideal state to examine the shortfall of the present state. Although it was inevitable that respondents scored the ideal state near the ceiling, we judged that it was not a failure because of the statistical results between the hospital and care facilities in the present state. The large effect size (r > 0.5) [82] of the results below the 5% significance level supported our judgment. For h4, we expected that part of the items would exhibit significant differences, as indicated by h1 and h2; however, a significant difference was observed only in relation to two items. The differences observed were slightly more significant in the results of h1 and h2, in which the number of significant differences between the ideal and real states differed between facilities. The results of h3 and h4 showed that there was a need to adopt a different method when comparing the inter-facility settings than the questionnaire for the intra-facility settings.
Next, we discuss the lack of information. The rejection of hypothesis H cannot be ignored when considering the information needed in transitional care at the time of discharge. Based on Study 1, we thought that the information not conveyed to care facilitates information on caregiving support and QoL support, which was not the case. The results indicated that the difficulty of information sharing could be attributed to structural issues. All information perceived by the nurses being communicated was not shared with the care workers; they were unsatisfied with the shared information. We consider that all factors mentioned in this discussion influenced this rejection.
Study 2 revealed that information other than the current medical information was inadequately transferred from the hospital to the care facility. This result quantitatively confirmed the qualitative results of Study 1 (i1-3 and i1-4). The significant differences shown differed from those of the expected categories. In addition, it was quantitatively verified that scenarios such as the one displayed in Figure 1 could not be realized.
In the comparison of the ideal and present states in view of h1 and h2, significant differences were observed in the perceptions held in both types of facilities about the “medical information needed in the future.” This result supports conventional studies on information sharing between medical facilities [30] and between medical workers [83]. Fewer errors can be realized by sharing information about medical information, such as already performed procedures, and care options, predictions of the patient’s prognosis, and anticipations of “what could possibly go wrong” [31]. Information on the anticipation for care during hospital discharge, which is not emphasized on according to our results, has the potential to improve the effectiveness of caregivers with less medical knowledge than that possessed by the medical workers. The category of information supporting the quality of life showed very significant differences, and many items had significant effects. Without the sufficient sharing of medical information and the information to support the quality of life, older adults experience relocation stress syndrome [18]. The data collected during the questionnaire design for Study 2 indicated that person-centered psychosocial information such as the life history of older adults [13] was required instead of information such as drug abuse [39]. Although it is desirable to prevent the omission of information characterized by many significant differences, as shown in the following section, their contents largely vary based on the condition of the individual patient’; hence, it is difficult to define a document template in detail (as discussed in 7.3 and 7.5).
Conventional studies have argued for the standardization of medical information [7, 8, 15]. However, our results indicate that standardization alone may not address the lack of information during the hospital discharge process. This process requires a wide range of information, from medical to QoL support information, and the need for each piece of information changes from phase to phase.
This result is also related to the variety of information required for elderly care as described in 7.3. Information on major diseases was prepared in the standard document format, whereas information on other diseases, potential future conditions, care, and livelihood support was likely to be inadequate because of the variety of topics and lack of definitions about writing. Elderly people often have multiple chronic conditions in addition to the primary condition that led to the hospitalization, with a variety of symptoms and combinations. Therefore, even if entries for other symptoms are prepared broadly and in detail, there may be cases where it is difficult for the sender to determine the appropriate content to write. There is a possibility that they will become confused about whether the information should be shared and how it should be recorded.
The rejection of hypothesis H cannot be ignored when considering the information needed in transitional care at the time of discharge. Based on Study 1, even though we thought that the information not conveyed to care facilitates information on caregiving support and QoL support, this was not the case. The results indicated that the difficulty of information sharing could be due to structural issues. Almost all the information the nurses perceived was being communicated was not shared with the care workers, and they were unsatisfied with the shared information. We consider that all of the factors mentioned in this discussion influenced this rejection.
The results of Study 2 were cognition scores that result from relative differences in the knowledge and experience of each worker. The current situation does not reflect accurate shared information and should be interpreted cautiously. For the reality of information sharing, it is necessary to analyze the text in the documents, the frequency of information exchanged, and the relation to other recorded data to better understand the situation.
In summary, information discrepancies perceived by hospitals (senders) were less than half as large as those perceived by care facilities (receivers). In other words, hospitals thought that they could share sufficient information relatively easily, whereas the care facilities thought otherwise. In addition, the information shared in the case of the internal medicine patient may be less sufficient compared with the surgical patient case. Considering information-sharing failures, analyzing the structural problems of information-sharing rather than the loss of individual information items is necessary.
7.2 Immaturity of Appropriate Information-Sharing Methods
In Study 2, significant differences were not concentrated in specific items or categories, which suggests that there may be structural problems in the information-sharing process (e.g., lack of standardized procedures and document formats and different information considered necessary by each facility in transition) as reasons for insufficient information sharing. This supported the qualitative result of the non-standardized sharing process shown in Study 1 (4.2.2 and i1-2 in Table 1). In addition to each facility being operated with its own formats and protocols, there was no standardization of care information required by care homes. This was the cause of the structural problem of insufficient or missing information about older adults, and these deficiencies were partially compensated by the efforts of professionals who are responsible for the work on both sides.
It will be quite difficult to study, but r Researchers in the field need to deeply analyze the problem and develop new systems. To develop this technology, it is desirable to have a system that standardizes the shared content based on evidence, similar to the approach taken in conventional studies [68, 69]. However, even medical information, such as information on mental and behavioral conditions and the contact information for follow-up, were often overlooked [38]. Even with standardized medical information, results showed that medication doses were inaccurate in nearly one-fifth of patient files, the fulfillment of care orders was delayed in nearly one-third, and documentation was partially missing in nearly half [31]. Based on our results, the analysis and development are also challenging because information on caregiving support and QoL support does not deal only with information that can be relatively definable, as is the case with medical information. Caregiving support and QoL support information that are challenging to standardize may result in more omissions. In addition, as Gilmore et al. [17] and our study have shown in the following sections, it is necessary to deal with individual and sensitive information; that is, the information-sharing assistant will be required to equip the ambivalent aspect that consists of pre-formatted registries and freely manageable registries.
Since our categories, separating medical information, care support information, and QoL support information, may be related to the gradation of ease of standardization, they can help solve this problem. As previous studies have indicated [7, 8, 15, 84], standardization will be effective for medical information in the current and in the future throughout this process. Caregiving and QoL support information are expected to be in a semi-standardized format. In particular, the design of registries for QoL support information is a combination of the freely manageable style and the semi-standardized style. Unfortunately, our study could not clarify what information should be written at what granularity, and therefore, careful exploratory research is needed in the future.
7.3 Variety of Information Needed for Caregiving and QoL support
As observed in Study 3, the difficulty in sharing information between medical and care facilities stemmed from the variety of information required for elderly care. Medical information is only a part of the total information handled in the care of elderly people in their final places of residence. Medical information, such as the name of the disease and blood pressure, is relatively easy to share because the information-sharing process and its contents can be easily standardized. In contrast, many types of information related to older adults’ lives, including their life history, personality, future life plans, relationships with family members, and goals for care, are required for care. Unlike medical information, the process of enabling the system to handle care information requires elaboration that considers the holistic approach.
Although there are no studies that focus on the variety of elderly care information, (in particular, studies on supporting QoL) those that have been conducted with chronically ill patients and children with medical complexity in care have shown the importance of incorporating information related to daily life. In the self-care of patients with chronic symptoms, incorporating the domains of daily life into medical aspects helped influence care and the overall health status [85]. In the care of children with medical complexity, parents as caregivers often collected a variety of clinical information daily and considered treatment options in the context of past experiences and quality of life [86]. Even when medical information is shared, communicating not only retrospective objective information, such as the patient’s background, physical condition, and procedures already performed but also care options and predictions of the patient’s prognosis and of “what might go wrong” has been linked to a reduction in errors [31]. As shown in these studies, it is important to convey a variety of information, such as daily life and care options, to the successor facility at the time of transfer.
Our study showed that each facility worker perceived that the presence or absence of entries in the document determined whether certain information was shared. A method that covers the entire range of information necessary for elderly care, such as the concept of an electronic clinical pathway in the medical field, is potentially suitable for development using assistive technology. However, as can be seen from Study 3 (especially “variety of information on elderly care,” “dependence of information sharing on the items in the format,” and “difficulty in sharing information supporting quality of life”), consideration is required if trying to specify each entry in detail. As mentioned in the previous section, considering the balance between the number of pre-formatted input items and the burden of inputting them is essential for technical development.
Study 3 showed that collecting information on patients’ QoL support was challenging, especially in university hospitals. However, the burden on subsequent facilities can be reduced if even a portion of the diverse information can be captured and shared from upstream in transitional care during the hospital discharge process. The nursing profession covers care not only for medical treatment but also for general living support, although to a lesser degree than that of care workers. Information on supported living is unlikely to be included in the records because of the nature of hospitals. However, each nurse is likely to obtain all types of information on older adults in the course of their daily work. The ideal technology should make it possible to holistically share information on supported living obtained by nurses during their daily work with the next facility. Therefore, a function that allows medical workers to record information they obtain on elderly patients in the course of their daily duties in the form of memos is required. This can be achieved by extending the mobile technology to care for older adults, which has been developed in recent years [54, 55, 87].
7.4 Unrecordable Information
When sharing information between facilities, it is necessary to provide sensitive personal information (such as excessive verbal and violent behavior, disruptive behavior at night, family relationships, and affiliation with special groups) according to the needs of the successor facility because the facility needs to prepare to accept the newcomer. However, recording such information in writing and sharing it with the next facility is difficult. Owing to its highly sensitive nature, it is difficult for senders to determine whether it can be communicated, even if it is necessary for care. In a similar field, behavioral health, which deals with mental illness and substance use disorders, it was shown that information-sharing technologies are not widely used because of the barrier of persistent concerns about privacy [88].
Unrecordable information is potentially observed in caregiving and QoL support information sharing. This assumption is considered because these types of information contain sensitive content and, therefore, have natures that make it difficult to share content. There are three reasons why it is difficult to share information. First, there is little information available at medical facilities, as shown in the result of Study 3, and the gathered information might show a tendency to be fragmented. With short hospital stays, it is difficult for hospital nurses to grasp the full range of sensitive information that should be shared with the next facility. Therefore, the information to be delivered to the recipient would also be fragmented. If they want to convey information to the next facility, assumption-based information with fragments is the only way to share information in such a scenario. However, such an ambiguous method would be challenging to apply in an actual context, and thus, unrecordable information shares a low potential. Consequently, the respondents in Study 2 would show mismatched communication in interfacility information sharing. The second reason is the difficulty in deciding the extent of information to be shared because of the sensitivity of the information. It is difficult for the sender to decide how much to say about the degree of sensitive information and how much to let the receiver know. The third reason is the difficulty of being sure whether the content should be shared. Senders would be worried about this problem when they need to discuss certain details of the personal history of the elderly person. For example, connecting the direct signs of depression is often challenging. While we may be able to tell them about the depression, it becomes difficult to convey with conviction the background information that needs to be considered.
The presence of sensitive information (i.e., unrecordable information) stems from current operational issues such as triggering lawsuits and conflicting family members. Even if the information is necessary for caregiving, it must be disclosed if it is an official record when requested by the patient or family. Suppose there is information in the documents that the family does not know, such as the existence of a common-law wife. In that case, it could trigger a lawsuit against the facility or a conflict among family members. However, we should not be so focused on risk avoidance that we overlook other serious problems. It is important to emphasize that the same mistakes can be repeated in any facility where an older adult is transferred because of the lack of information sharing.
Another problem is that it is not easy for senders to document sensitive information. From our experience composing the questionnaire in Study 2, there are limited ways to formally share sensitive information necessary for improving caregiving and QoL. The only official method was to record the information in the free-text format in the “Others” column. Again, in this document style, it can be assumed that the diversity and balance of information are difficult to achieve when attempting to convey sensitive information, which is representative of unrecordable information. Further, it is difficult to convey content that is so varied and communicated in so many different ways (e.g., family troubles, sexual harassment, verbal abuse, violence, introductory notes about key persons, emotional distress), such that it relies on the dedication of the sender. Another problem with this method is that the amount and content of information entered depends on the writer’s skill.
It is not easy to find a solution to this problem. Continued exploratory research is needed to find information worth sharing and identify the means to share it for both senders and receivers. Technical support for information sharing is required for elderly care, but that is difficult to record, and this needs to be approached with caution while considering ethical issues. Although a study pointed out the sharing of psychosocial information, no paper has been found that has addressed the importance of unrecordable information in transitional care for older adults. The discovery of the importance was considered one of the empirical contributions to the health information research of this study.
7.5 Structural Factors That Lead to Inadequate Mutual Understanding
From 7.2 to 7.4, we discussed the characteristics of the information needed for elderly care. In this section, we focus on structural factors in interfacility cooperation.
Study 3 showed that in collaboration between facilities of different natures, the sender often did not understand exactly what information the receiver required, and mutual understanding was insufficient. In our findings, the balance of the medical and care information about older adults can be considered along the time shift, as shown in Figure 1. In the beginning, the staff, primarily nurses in hospitals, intensively gather medical information to provide quality treatment to the patient. During the treatment, particularly during an operation, the major concerns for information sharing are the patient’s medical conditions because deterioration can be rapid when a problem occurs. The significance of care information increases with a gradual improvement in the symptoms. In other words, in the process of switching from being a patient in the hospital to becoming a resident in the care facility, the care staff (primarily professional care workers, care managers, and social workers) start collecting holistic information about the person. After residing in the care facility, medical information becomes a small part of the entire information package, and care information becomes the primary material.
In addition to the discrepancies between the shared information and the present condition of the elderly person, the care facilities identified a lack of information on livelihood support and care in transition. Therefore, the information necessary to care for older adults may not be at hand at the start of care. In contrast, it was difficult to collect such information in medical facilities because of the short period of hospitalization. Therefore, there were discrepancies in the information obtained and in the situation of older adults because of the differences in the tasks that each facility intensively coped with. However, inadequate mutual understanding occurred because of the lack of means and opportunities to understand each other’s scenarios and the differences in viewing elderly care. In the literature, the lack of understanding of mutual roles and business in transitional care has been shown to be a challenge [47]. A lack of mutual understanding will likely lead to a lack of consideration, which may be an important issue for the entire research field of care support in health information research.
Peterson’s study [41] was insightful in understanding the inadequate mutual understanding that the recognition of nursing and caregiving was different in hospital and home care nurses. This study indicated that cross-sectional collaboration among nurses was hampered by differences in the definitions of nursing in the “two worlds.” Medical care is rarely provided in Japanese care facilities. Therefore, there may be a gap between nursing in Japanese hospitals and caregiving in care facilities that is equal to or greater than this study’s finding. Their study concluded that cross-sectional collaboration among nurses should be recognized as inter-professional collaboration and not intra-professional collaboration. We would like to emphasize the importance of this point from our results.
Assistive technologies are required to support mutual understanding for solving the problems described in this section. The support includes the following two points: (1) to support the understanding of the differences in the nature of the facilities, and (2) to support the grasping of the extent to which each facility fulfills its role in response to the differences in its nature. The former cannot be satisfied by the traditional approach to standardization [68–70], which is merely to fill in a default form with predetermined information in a working manner. During the input of information, the informant needs to consider the nature and environment of the recipient’s facility and whether the information to be shared is appropriate for the recipient. For the latter, a function that clarifies the areas of responsibility of each facility and visualizes the degree of achievement is required.