Contributed equally to this work with: Shigeyoshi Maki, Kuniyoshi Nagai, Shoko Ando, Koji Tamakoshi Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Writing – original draft * E-mail: s-maki@sugiyama-u.ac.jp Current address: Nursing Course, School of Medicine, Gifu University, Gifu, Japan Affiliation Department of Nursing, School of Nursing, Sugiyama Jogakuen University, Nagoya, Aichi, Japan
Contributed equally to this work with: Shigeyoshi Maki, Kuniyoshi Nagai, Shoko Ando, Koji Tamakoshi Roles Conceptualization, Methodology, Supervision, Writing – review & editing Affiliation Department of Nursing, School of Nursing, Nagoya University of Arts and Sciences, Nagoya, Aichi, Japan ⨯
Contributed equally to this work with: Shigeyoshi Maki, Kuniyoshi Nagai, Shoko Ando, Koji Tamakoshi Roles Conceptualization, Formal analysis, Methodology, Project administration, Supervision, Validation, Writing – review & editing Affiliation Department of Nursing, Nagoya University Graduate School of Medicine (Health Sciences), Nagoya, Aichi, Japan ⨯
Contributed equally to this work with: Shigeyoshi Maki, Kuniyoshi Nagai, Shoko Ando, Koji Tamakoshi Roles Conceptualization, Formal analysis, Methodology, Supervision, Validation, Writing – review & editing Affiliation Department of Nursing, Nagoya University Graduate School of Medicine (Health Sciences), Nagoya, Aichi, Japan ⨯
Schizophrenia is a disorder characterized by psychotic relapses. Globally, about 15%–30% of patients with schizophrenia discharged from inpatient psychiatric admissions are readmitted within 90 days due to exacerbation of symptoms that leads to self-harm, harm to others, or self-neglect. The purpose of this study was to investigate the structure and predictors of in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia. A new questionnaire was developed to assess the extent to which respondents delivered in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia. This study adopted a cross-sectional research design. The survey was conducted with the new questionnaires. The participants were registered nurses working in psychiatric wards. Item analyses and exploratory factor analyses were performed using the new questionnaires to investigate the structure of in-hospital nursing care leading to reduction in early readmission. Stepwise regression analyses were conducted to examine the factors predicting in-hospital nursing care leading to reduction in early readmission. Data were collected from 724 registered nurses in Japan. In-hospital nursing care leading to reduction in early readmission was found to consist of five factors: promoting cognitive functioning and self-care, identifying reasons for readmission, establishing cooperative systems within the community, sharing goals about community life, and creating restful spaces. In-hospital nursing care leading to reduction in early readmission was predicted by the following variables: the score on the nursing excellence scale in clinical practice, the score on therapeutic hold, and the participation of community care providers in pre-discharge conferences. Japanese psychiatric nurses provide nursing care based on these five factors leading to reduction in early readmission. Such nursing care would be facilitated by not only nurses’ excellence but also nurses’ environmental factors, especially the therapeutic climate of the ward and the participation of community care providers in pre-discharge conferences.
Citation: Maki S, Nagai K, Ando S, Tamakoshi K (2021) Structure and predictors of in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia in Japan: A cross-sectional study. PLoS ONE 16(4): e0250771. https://doi.org/10.1371/journal.pone.0250771
Editor: Andrea Gruneir, University of Alberta, CANADA
Received: September 19, 2020; Accepted: April 13, 2021; Published: April 30, 2021
Copyright: © 2021 Maki et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting Information files.
Funding: SM was funded by Japan Society for the Promotion of Science KAKENHI Grant# JP17K17523. URL: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-17K17523/ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Schizophrenia is a disorder characterized by psychotic relapses [1]. Many people with schizophrenia suffer recurrent symptoms of psychosis (i.e., hallucinations, delusions, and disorganized speech) and chronic cognitive deficits (i.e., impaired executive function, memory, and speed of mental processing) [2]. Various indicators are used for the relapse of schizophrenia. Readmission rate is widely used as an indicator of relapse of schizophrenia [3, 4]. Some patients with schizophrenia have been readmitted to hospitals because their condition worsens so that they cannot live in communities [2]. Readmissions lead to rising costs of mental health care [3–5]. Furthermore, readmissions put a strain on patients, affect their prognosis, and lead to a breakdown of their family function [3, 6].
Several studies examined the factors that contribute to readmission. Schizophrenia is one of the diagnoses of mental disorders associated with psychiatric readmission [7–9]. Factors associated with readmission can be classified based on three criteria: patients’ clinical characteristics, health system characteristics, and characteristics of hospitalizations. Regarding patients’ clinical characteristics, younger age [10, 11], marital status of “unmarried” [11], complications [6, 12], medication nonadherence [13–15], and maladaptive functioning of family systems [16] are associated with increased risk for readmission. Regarding health system characteristics, proportion of experienced psychiatrists at a hospital [17], multiple uses of health service [6], and unplanned discharge [12] are associated with increased risk for readmission. Regarding characteristics of hospitalizations, having a history of previous hospitalizations [6, 11, 12], duration of involuntary admission [11, 12], and total admission duration [6, 11] are associated with increased risk for readmission. Furthermore, a length of stay in a psychiatric ward of more than 28 days plays a protective role in preventing readmission [4].
In particular, readmissions within 90 days of discharge are usually defined as early re-admissions [10]. Between 15%–30% of patients with psychiatric disorders experience early readmission [4, 11, 18, 19]. Early readmissions reflect the degree of continuity between hospital care and community care [10]. Patients take an average of 90 days to establish therapeutic relationships with community care providers after their discharge [20]. Many patients encounter critical situations within 90 days of discharge [21, 22]. Reduction in early readmission could lead to successful transitions of patients from hospital care to community care. In-hospital care that leads to reduction in early readmission is important for achieving successful transitions from hospital care to community care.
Revealing the structure of such in-hospital care will facilitate the development of intervention programs that support this transition. Previous studies have identified the structure of the following six types of in-hospital psychiatric care: care and management of inpatients who exhibit self-cutting behaviors [23]; case management practice [24]; health services to promote mental health [25, 26]; psychiatric services in forensic inpatient care [27]; therapeutic attitudes of professionals working with drug abusers [28]; and rehabilitative care in longer term mental health facilities [29–32]. However, the structure of in-hospital care leading to reduction in early readmission is unclear.
Most psychiatric care in Japan is still provided in inpatient settings [33], despite Japan’s efforts to facilitate the transition from inpatient care to community care [34]. The number of psychiatric beds in Japan is four times higher than the Organization for Economic Cooperation and Development (OECD) average (267 beds vs. 66 beds per 100,000 population) [35]. The average length of stay in a psychiatric hospital in Japan is approximately 300 days, much longer than in other OECD countries [36, 37]. Community mental health services in Japan are not sufficient [37]. Approximately 25% of inpatients experience readmissions within 90 days of discharge (i.e., early readmissions) [38]. In Japan, delivering community mental health services for patients in communities are not enough to decrease early readmission in patients [4]. The context of mental health care in Japan is different from elsewhere in the world in that the length of hospitalization is longer and the more resources are invested in inpatient care rather than community care.
Therefore, it is desirable to identify the predictors of in-hospital nursing care leading to reduction in early readmission in the Japanese mental health care context. Specifically, this information could be used to develop strategies to support the successful transition of patients from hospital care to community care. In-hospital nursing care leading to reduction in early readmission may be predicted by nurses’ environmental factors as well as their individual factors. The risk of early readmission could be predicted by not only community follow-ups but also by length of stay in hospital [4]. Nursing performance is influenced by nurses’ environmental factors as well as personal factors [39]. A positive social climate in hospitals is associated with higher patient treatment motivation, treatment engagement, and patient-nurse therapeutic alliance [40]. A significant relationship between social climate and staff performance and morale has been reported [41]. We hypothesized that in-hospital nursing care leading to reduction in early readmission could also be predicted by the social climate in psychiatric wards as well as by nurses’ individual factors.
The purpose of this study was: (a) to investigate the structure that underlies in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia; and (b) to examine the factors predicting in-hospital nursing care leading to reduction in early readmission, by focusing on both the social climate in psychiatric wards and nurses’ individual factors.
This study adopted a cross-sectional research design and utilized self-administered questionnaires, which participants completed anonymously.
This survey was conducted at hospitals where psychiatric beds accounted for more than 60% of all beds. The hospitals were selected in three steps. First, the hospitals were extracted from the Japan Medical Analysis Platform of the Japan Medical Association (http://jmap.jp/). Of the 1,271 hospitals extracted, 59 were public hospitals, 69 were non-profit corporation hospitals, and 1,143 were private hospitals. Second, the 128 public hospitals and hospitals owned by non-profit corporations were all included in the study, as well as a randomly selected sample of 128 private hospitals, which totaled 256 hospitals. Third, we then mailed request documents for research cooperation to the directors of the 256 hospitals. Written consents for research cooperation were obtained from 40 hospitals (i.e., 19 public hospitals, six hospitals owned by non-profit corporations, and 15 private hospitals).
Nurse managers at the 40 selected hospitals provided information on the number of registered nurses (RNs) working at their hospital who met the inclusion criteria listed below. They reported that 1,995 RNs fulfilled the criteria. Subsequently, request documents for research cooperation to the RNs with self-administered questionnaires were mailed to the nurse managers, who then distributed them to the 1,995 RNs identified. The RNs, who gave written consent for research cooperation, returned the completed questionnaires to the researchers by mail. The inclusion criteria were: 1) full-time RN, and 2) working in a psychiatric ward. The exclusion criteria were: 1) nurse managers, 2) RNs working in outpatient wards, and 3) part-time nurses. Data were collected between February and March 2018.
Sample size estimations were based on the recommendations of Pett, Lackey & Sullivan [42]. They recommended that the minimum number of subjects for an exploratory factor analysis (EFA) was 10 to 15 per initial item. In this study, 430–645 subjects were required. Considering a valid response rate of approximately 30%, 1,433–2,150 questionnaires had to be distributed. The sample of 1,995 utilized in this study can be considered adequate.
The following demographic characteristics of the participants were examined: gender, years of psychiatric experience, advanced practice registered nurse (APRN) status (certified nurse or certified nurse specialist), experience as a psychiatric home-visiting nurse, experience in providing psychiatric outpatient care, experience in somatic care wards, and educational level. Further, the following characteristics of the hospitals or wards in which they were working were asked: hospital establishment, adoption of primary nursing, whether or not pre-discharge conferences were usually held, the participation of patients’ families in the pre-discharge conferences, and the participation of multidisciplinary teams in the pre-discharge conferences. Only participants who answered that “pre-discharge conferences were usually held” responded to the question if the patients’ families attended the pre-discharge conferences. Similarly, only participants who explained that “pre-discharge conferences were usually held” responded to the question of whether the conference member consisted only of hospital staff or consisted of both community and hospital staff.
A 43-item IRERSS was developed in this study. In the IRERSS, respondents were required to recall a patient with schizophrenia who had previously been readmitted within 90 days of discharge, but who could live in a community for more than 90 days after receiving an in-hospital intervention. The respondents answered to what extent they provided the nursing care indicated in the questionnaire during the patient’s in-hospital intervention. To minimize recall bias, a diagram representing the time axis was provided with the questionnaire (see S1 Fig), and the items were expressed in concrete contents. It was easy for the respondents to recall patients who had experienced readmissions within 90 days of discharge, namely, early readmissions, because the Japanese public health insurance program defines hospitalization within 90 days of the discharge as a readmission and after more than 90 days after discharge as a new hospitalization [43]. The nursing care delivered during the above in-hospital intervention would be considered “in-hospital nursing care leading to reduction in early readmission” if patient was not re-hospitalized within 90 days following discharge after receiving the nursing care.
Responses to each item were recorded on a 5-point Likert scale that ranged from 1 (strongly disagree) to 5 (strongly agree). The items of the IRERSS were developed based on Maki et al.’s findings [44]. In their qualitative study, they examined “in-hospital nursing care leading to reduction in early readmission” among patients by conducting interviews with 17 proficient psychiatric nurses. They revealed 38 concepts of nursing practice leading to reduction in early readmission among patients with schizophrenia. Based on the 38 concepts, we created 38 items of in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia. Furthermore, five of the items (Items 39, 40, 41, 42, and 43) were added based on the feedback provided by the five experts (i.e., two researchers in the field of mental health nursing, one psychiatrist, and two experienced nursing researchers). To make the expression of the items clearer and more concise, they were revised based on the opinions of a five-person panel (i.e., two graduate students, one university faculty member, and two psychiatric nurses). Accordingly, the IRERSS assesses the degree of implementation of “in-hospital nursing care leading to reduction in early readmission among patients with schizophrenia.”
EssenCES is a 17-item questionnaire that measures three aspects of the social climate in psychiatric wards (PC: Patients’ cohesion and mutual support, ES: Experienced safety, and TH: Therapeutic hold). Each subscale included five items, and two items were not included in any of the subscales. Each item used a 5-point Likert scale from 0 (not at all) to 4 (very much). Higher scores indicated that respondents perceived the hospital ward climate as more positive. “Patients’ cohesion and mutual support” reflected an essential quality of therapeutic communities and effectively working treatment groups [45]. “Experienced safety” referred to the level of perceived tension and threat of aggression or violence [46]. “Therapeutic hold” indicated the extent to which the unit was perceived as supportive of patients’ therapeutic needs [47]. The questionnaire was originally developed to assess the social and therapeutic atmosphere of forensic psychiatric wards [48]. This questionnaire’s transferability to general psychiatric settings was confirmed [46]. The Japanese version of this scale had good internal consistency and construct validity [49, 50].
Nursing performance was influenced by not only nurses’ individual factors but also environmental factors [39, 51]. This study investigated the relevance between in-hospital nursing care leading to reduction in early readmission and the environmental factors, using EssenCES-JPN.
Higher scores were indicative of self-reported nursing excellence in clinical practice. The NES consists of 35 items and seven subscales, namely: (1) collecting and using client’s information continuously and efficiently, (2) performing with appropriate knowledge/skills in clinical settings, (3) developing relationships with clients/families through communication, (4) overcoming difficult conditions of clients and/or in the work environment, (5) identifying potential problems for clients and solving them creatively, (6) protecting the personality and human dignity of clients, and (7) being aware of fulfilling a number of roles as a medical and nursing team member and being able to perform them. The NES had good reliability and validity [52, 53]. The NES had similarities with the IRERSS in that nurses could easily assess their own nursing practice. To control for self-assessment bias, the questionnaire instructions explained that the content of the responses to this survey would not affect the evaluation of the participants’ performance. The NES, which measures nurses’ individual nursing practice, was used as one of the predictors of the IRERSS.
All statistical analyses were conducted using IBM SPSS Statistics version 25. Questionnaires that contained missing responses were excluded. Descriptive analyses were performed to examine the demographic characteristics of the participants and the features of the hospitals/wards in which they worked. Participants’ years of psychiatric experience were classified into three categories: less than 5 years, 5–14 years, 15 years or more. The normality of data was evaluated by means of a visual inspection of histograms and QQ-plots. All statistical tests were two-tailed and p < .05 was considered significant.
First, item analyses were performed on the items of the IRERSS. Ceiling and floor effects were examined for each item of the IRERSS. Inter-item and item-total correlations were computed. In the inter-item correlation analyses, items with a correlation < .30 with all items, or items with a correlation >.90 with any item were removed to avoid the risk of multicollinearity [54]. In the item-total correlation analyses, an item was excluded if the correlation between the item score and total score without the respective item was < .30.
Second, an EFA was conducted using maximum likelihood extraction and promax rotation. The Kaiser-Guttman criterion (eigenvalues > 1) was used to determine the number of factors that should be retained. Items with loadings that were lower than (i.e., < .40) onto all factors were excluded. Additionally, items that were strongly loaded (i.e., >.40) onto two or more factors were also excluded. The analysis was repeated until all the items were strongly loaded onto only a single factor. The adequacy of the EFA was evaluated by the Kaiser-Meyer-Olkin (KMO) statistic and Bartlett’s test of sphericity. The internal consistency of the scale was examined by computing Cronbach’s alpha values for the overall scale and each subscale of the IRERSS. The factors identified through the EFA were considered subscales within the overall scale. Each subscale score was calculated by summing the included items.
Correlation and univariate analyses (i.e., unpaired t-tests, and one-way analyses of variance with post hoc analyses) were performed to examine the data obtained. Correlations between the overall scale and each subscale’s score on the IRERSS, overall score on the NES, and each subscale’s score on the EssenCES-JPN were computed. Using unpaired t-tests, differences in the overall scale and each subscale of the IRERSS were compared between groups dichotomized based on each of the following variables: gender, APRN status, experience as a psychiatric home visiting nurse, experience in providing psychiatric outpatient care, experience in somatic care wards, and adoption of primary nursing. Using one-way analyses of variance and post hoc analyses (Bonferroni correction), differences in the overall scale and each subscale of the IRERSS were compared among the three groups classified using each of the following variables: years of psychiatric experience, educational level, hospital establishment, participation of families in pre-discharge conferences, and participation of multidisciplinary teams in pre-discharge conferences.
Forward-backward stepwise multiple regression analyses were conducted to control for confounding variables (inclusion value = .05 and exclusion value = .10). In the stepwise multiple regression analyses, the overall scale, and each subscale scores on the IRERSS were entered as dependent variables. The variables for which the correlation analyses and univariate analyses yielded significant results were entered as independent variables.
Ethical approval to conduct this study was granted by the ethics committee of the Graduate School of Medicine, Nagoya University, Japan (No: 17–155). The participants were informed about the aims of the study and the benefits and risks of participation through printed forms. The participants provided written informed consent. They responded to each questionnaire anonymously, enclosed the completed questionnaires in sealed envelopes, and returned them to the researchers. Hospital directors and nurse managers did not participate in the data collection process. If participants had any questions about the study, they were able to call or email the principal investigator, whose phone number and address were listed on the documents for research cooperation. Permissions to use the EssenCES-JPN and the NES were obtained from the copyright holders.
Data were collected from 823 RNs (response rate = 41.25%). Excluding missing responses, the final sample size was 724 (valid response rate = 36.29%). The demographic characteristics of the participants and the features of the hospitals/wards in which they worked are presented in Table 1.