A text-mining analysis of nursing care for overseas visitors in Japan

Background: The provision of health and nursing care for overseas visitors to Japan is becoming more important for nurses and hospitals owing to the sharp increase in visitor numbers. We investigated the attitudes and concerns of Japanese nurses when caring for overseas visitors, focusing on preparedness. Methods: We carried out a cross-sectional text-mining study. The analysis was both qualitative and quantitative with a descriptive aim, including two open-ended questions to nurses. A total of 513 nurses participated, representing 350 hospitals from eight prefectures across Japan. We collected data from nurses enrolled in a first-level management course from August 2011 to February 2012. We analyzed the responses to the open-ended question with Text Mining Studio 5.1, and numeric data with JMP 11.0. Results: Only 32.8% (167) of the nurses affirmed that their hospitals were prepared to care for overseas visitors. Nurses from hospitals that had prepared for overseas visitors were more concerned about culture


Backgrounds
In modern times, many people move across borders of countries as part of the globalization of goods and services. In Japan, the numbers of overseas visitors have increased from 19.74 million in 2015 to over 31 million in 2018, as a result of the successful "Visit Japan Campaign" since 2003 [1]. The Japanese government has set a goal of attracting millions of additional visitors, including medical tourists, by 2020 [2] . In addition to overseas visitors, the number of foreign students and workers has grown sharply during the last decade [3]. Over 2.5 million foreigners were registered as living in Japan [4]; this represents 1.99% of the total population of Japan in 2018.
Caring for overseas visitors is a challenge for Japanese nurses because of differences in language, lifestyle and culture. In this regard, providing equivalent care for these visitors has become a formidable task [5,6]. Currently, visits to healthcare facilities by overseas visitors often increases the anxiety level of healthcare professionals, particularly nurses. Indeed, nurses often have difficulty understanding the attitudes and concerns of overseas visitors, but few studies have documented these concerns. Consequently, when an overseas visitor is admitted to a 5 hospital, the medical staff is often unable to provide appropriate care [7,8]. Nurses are primarily concerned with the language barrier, informed consent, and payment options [8].
Although there are language support systems for communication problems in some Japanese healthcare facilities [9,10], nurses strongly desire better English language skills [8,11]. Examples of better communication tools are mobile terminals of online systems to prepare for a medical examination [10], but this does not compensate for a lack of speaking ability and cultural sensitivities.
Studies carried out in various countries have determined that nurses must adjust their attitudes to provide safe care to overseas visitors in a culturally sensitive manner, for example, as in Canada [12,13], Sweden [14,15], Norway [16], Iran [17], and Korea [18]. In Geneva, Switzerland, however, where approximately 60% of the residents are non-native, hospitals have faced challenges taking care of overseas visitors, especially those patients who do not speak French, because of the difficultly in retaining trained interpreters who contribute to the quality of care and patient safety [6].
Taking care of overseas visitors involves processes to build a trust 6 relationship with patients, and thus nurses need practical preparations to establish such relationships [19]. Indeed, Nonaka & Higuchi (2010) noted the importance of developing relationships between nurses and overseas visitors. When this is achieved, nurses report that they are more satisfied with the care they provide [20,21].
With respect to the care of overseas visitors, some studies have investigated the attitudes and concerns of Japanese nurses, most of whom lack an international background [8,22]. Problems with verbal communication and inadequate preparation were cited as key elements in taking care of those visitors [11]. However, the effectiveness of preparations was not considered. Therefore, we investigated the attitudes and concerns of nurses who care for overseas visitors in Japan, focusing especially on the effectiveness of preparation. We also reassessed nurses' concerns across prefectures and in popular tourist areas.

Eligibility, study design and procedures
We conducted a cross-sectional study from August 2011 to February 2012. Data were collected from nurses via a questionnaire, 7 hereafter referred to as Mari Meter [8], provided during a first-level nursing management course conducted by the Japanese Nursing Association for nurses in eight prefectures. We targeted both staff nurses and nurse managers who are involved in providing direct healthcare at their hospitals, which are located in prefectural capitals and popular tourist sites for overseas visitors in Japan. One or two nurses from each of 592 hospitals participated in the course, and we received the data from the respondents within one month by regular mail.

Questionnaire
The questionnaire contained three parts, including a nine-item demographic section, open-ended questions 1 and 2, and Mari Meter [8].
The demographic information collected included age, gender, profession, educational level completed, current professional position, current ward, number of years as a professional nurse, experience taking care of overseas visitors, and preparedness for taking care of them ( Table 1 language direction in a hospital, and dealing with medical staff. At the top of the Mari Meter section, two sentences read: "Listed below are issues that concern some but not all participants. Please read each statement and group the response that best suits your experience." Responses were assessed using a five-point Likert-scale: not at all concerned; not very concerned; neutral; somewhat concerned; or, very concerned.

Text-mining analysis method
The written by Japanese children [24].
In this study, we used Text Mining Studio to summarize and understand the feelings of 513 nurses responsible for caring for overseas visitors at their various hospitals. We analyzed their viewpoints. The textmining analysis counted words only once based on context even if one nurse wrote that word more frequently. We followed these five steps to analyze the open-ended questions. (3) Positive/negative viewpoints from the reputation analysis ( Figure   4): This analysis quantifies the number of negative and positive 10 words. When nurses write a word that has a negative connotation, it counts as a negative [25]. When nurses write a word that suggests a positive opinion, it counts as a positive. This helps determine which words the nurses assign as having a positive or negative connotation.  there were three possible responses, namely "prepare-yes", "prepare-no", and "prepare-unknown". A word is deemed to have a close meaning with another word if it is consistent with the nurses' choice of "prepare-yes", "prepare-no" and "prepareunknown" This question is based on correspondence analysis with text by attribution of preparations for overseas visitors.

Analysis of numeric data
Baseline characteristics for nurses and Mari Meter were summarized using descriptive statistics in JMP 11.0 [27].   6%, 125).

Basic information
Question 1: The response rate was 513 (100%), comprising a total of 7420 words and 1651 sentences. Question 2: The response rate was 513 (100%), comprising a total of 2551 words and 760 sentences. The nurses also mentioned the following words: bring word (tutaeru) 14 times, relate (kankeisuru) 12 times, contact (fureai) 12 times, understanding (rikai) 7 times, and not charge (motomenai) 2 times. These five words emerged from the 7420 total words as having importance     (Figure 4) The nurses mentioned only three words having a negative connotation, namely emotion (kanjo), pain (itami), and feeling, compared with eight words having a positive connotation.

Positive/negative viewpoints from the reputation analysis
Word relationship network and through co-occurrence ( Figure 5) The nurses' concerns were analyzed using word relationship networks and through co-occurrence. The size of each circle or oval in difference (chigai). In Group A, language was related to several other words (very thick lines in Figure 5) and had a large frequency for the words "not understand" (rikaidekinai), "important" (taisetu), "inpatient" (nyuin-kanjya), and "good" (yoi). Many other words also were classified in    Group A. For Group B, communication was related to "cannot do" (dekinai), "can do" (dekiru), "proficiency" (umai), "language" (kotoba), "conversation" (kaiwa) and "self" (honnin). In Group C, the concept people, was related to "exist" (iru) and "interpreting" (tuyaku). In Group D, the word Japanese was related to "can speak" (hanasu-dekiru). In Group E, anxious was related to "transmitted" (tutawaru). In Group F, difference was related to "culture" (bunka) and "trouble" (toraburu). (Figure 6)

Correspondence analysis by preparation
The largest circle shown in Figure 6 was language (kotoba), which was almost the same distance from "prepare-yes", "prepare-no", and "prepare-unknown". The words "communication" (komunikesion) and "problem" (mondai) were at almost similar distances from prepare-yes, and prepare-no in two squares. The response "prepare-yes" was closely related to "difference" (chigai), "culture" (bunka), "life style" (lifu-stail) and "explain" (hyogen). The word "Japanese" (nihongo) was closely related to "prepare-no" as well as to "interpreting" (tuyaku), "anxiety" (sinpai), "no understanding (wakaranai)" and "need (youkyu)". The nurses who answered that they did not know if their hospitals were prepared for non-Japanese patients were close in distance from the word "concern"

Descriptive statistics analysis
The results for each item of Mari Meter are presented in terms of a percentage in a bar graph with descriptive statistics in all nurses ( Figure   7). On average, 80% of nurses responded as very concerned or somewhat concerned. Of the 15 items for concerned, the top 5 terms were: language communication, informed consent, payment of medical expenses, lifestyle difference, and dealing with medical staff. The reliability of the 15 multiitem questions of Mari Meter was confirmed by calculating Cronbach's alpha, which was 0.82.   Results from our Text Mining analyses found that those nurses from hospitals that had preparations for foreign patients were more concerned about the patients' cultural background. On the other hand, nurses whose hospitals had no preparation for foreign patients were concerned about language. In categorical logistic regression, we aim for the analysis to predict the nurses in the "yes" or "no" for preparation at their hospitals to take care of foreigners, as a concerning factor of Mari Meter questionnaire. The following hypotheses were set.

Regression analysis
H 0 : There is no relationship between having or not having preparation to take care of foreigners and Mari Meter. Results showed a relationship between "yes" and "no" to take care of foreign patients with all 15 items of significance; lifestyle difference, emergency care and protection of privacy (p < 0.001) which were more concern the nurses who had preparation to take care of foreigners at their hospitals ( Table 2). The nurses whose hospital made preparations were more concerned about the patients' backgrounds. By contrast, the nurses who did not make preparation to take care of foreign patients had 18 a greater concern about other 12 items significantly (p < 0.001). The nurses who did not make preparations to take care of foreign patients most two concerns were language communication and informed consent.

Discussion
This study has five main findings. First, in all, 92.3% of the nurses had experienced caring for international patients. Second, only one of three nurses responded that their hospitals had made preparations for overseas visitors. Non-preparedness was associated with more anxiety related to taking care of overseas visitors. Third, many of the nurses expressed strong concern about their foreign language skills with respect to caring for overseas visitors. Fourth, the nurses were more inclined to blame themselves for not understanding overseas visitors. Finally, the nurses were concerned about foreign language communication, informed consent, payment of medication expenses, and lifestyle differences.
The majority of the nurses had cared for international patients despite not having any previous nursing-related education pertinent to international cultures. This result raises concerns about the true ability of Japanese nurses to effectively care for overseas visitors. 19 A total of 32.8% of the nurses answered that their hospitals had made preparations to care for overseas visitors. Participants who answered that their hospitals were prepared noted that the preparation had focused on differences in culture and lifestyle. Non-preparedness was associated with anxiety and expression of words having a negative connotation. Most of the nurses were strongly concerned about their foreign language skills with.
Many of the nurses expressed strong concern about their foreign language skills with respect to caring for overseas visitors. These results echo conclusions reached by other investigators [7,11,22]. From a global perspective, our results mirror what has been documented in other nations. Professional interpreters are vital to communicating medical treatment and healthcare processes. Patient-centered communication is key to successful care of foreigner patients, as clearly demonstrated in Sweden [28]. A study in Switzerland found that immigrants were at a disadvantage for psychiatric inpatient care [29]. However, the local populations of most European countries can speak English or more than one language, so at least there is the possibility of a common-language solution. This is not the case, however, for Japan, where there is less of 20 an opportunity to speak English. This manifests as high stress for many nurses who are assigned the responsibility of directly caring for overseas visitors.
The nursing nurses in this study tended to blame themselves for not understanding overseas visitors. When a nurse participant takes care of a foreign visitor but cannot fully communicate with them, they feel ashamed of their lack of communication skills.
In this study, the majority of Japanese nurses were most concerned about foreign language communication, informed consent, payment of medication expenses, and lifestyle differences. This result is consistent with that from previous research [8].
This study has several strengths. First, our responses were numerous and reflected the nursing population of the most densely populated prefectures in Japan. Indeed, our data were collected from over 500 nurses at more than 300 hospitals in eight prefectures across Japan. Second, we used a text-mining method to analyze the responses to two open-ended questions. Hence, our results can be interpreted both visually and objectively. Third, we tested the results from the text mining analyses and the categorical logistic regression analyses to arrive at the 21 same conclusion. In categorical logistic regression, we examined whether the concerns might be different between the nurses who prepared to take care of foreign patients against those who have not. There is a difference between "yes" and "no" to have preparation to take care of foreign patients with all items being significant (p < 0.001). The nurses whose hospital made preparations were more concerned about the patients' backgrounds instead of language communication. The nurses whose hospital had not made preparations were more concerned to care for foreign patients related to the 12 questions asked.
This study also has some limitations. First, we compared groups based on whether their hospitals made preparations for overseas visitors.
We did not conduct an intervention study. Second, this study does not distinguish a specific target population of overseas visitors. For example, the type of care provided to patients might differ for tourists, foreign students, temporary workers, and immigrants. Finally, our survey did not include the amount of time the nurses felt they spent caring for non-Japanese patients, as it might have led to recall biases.

Conclusions
Our results reveal that text mining of written information provided 22 by a nursing cohort could help with understanding the attitudes and concerns of nurses who take care of overseas visitors. Although hospitals in Japan require that a system be in place to facilitate the treatment of overseas visitors, data obtained from our nursing cohort suggest that not all such systems are effective. Future studies may require an intervention study to obtain stronger evidence for the effectiveness of systems currently in place in Japanese hospitals to prepare for the care of patients from various nations.