Lengthening hospital stays, increasing numbers of patients with age-related complications, and a shortage of nursing staff have been of major concern in medical psychiatry in Japan [1] [2] .

The “Restructuring Plan for Mental Healthcare and Medical Care” was implemented in 2004 by the Ministry of Health, Labor and Welfare to address the problem of prolonged hospitalization by implementing an outreach policy to promote a shift from hospital-based healthcare to community living. The aim was to discharge a total of 150,000 patients in the ten years from 2004 to 2013. However, the number of inpatients who were discharged was approximately 70,000, which did not meet the required goal [3] [4] . There are several reasons for the difficulties in supporting the required number of discharges. Psychiatric diseases are characterized by exacerbations and remissions [5] and are associated with a chronic course [6] . Furthermore, social stigma surrounding mental health negatively affects the quality of social support provided to patients [7] [8] . In particular, patients with schizophrenia, which accounts for the majority of psychiatric diseases, not only have subjective symptoms, such as hallucinations and delusions, but also dysfunction and disability [9] [10] . Social disadvantages are associated with these factors [11] . Considering these circumstances, the decisive factors for living in a community are appropriate medical treatment and care for age-related complications, adequate rehabilitation for discharge during hospitalization, and the readiness of communities to receive discharged patients.

As mentioned previously, direct psychiatric care for patients has become increasingly important [12] . However, a great deal of work by nursing staff, along with the increased amount of care, is a significant concern in psychiatry, as well as in other clinical settings, owing to the increased number of patients with aging-related comorbidities such as bone fractures, cancer, heart disease, and diabetes mellitus [13] [14] . Under these circumstances, countries such as Japan that face a super-aging society and a decline in the working-age population have been advocating for the use of advanced information and communications technology (ICT) to improve the efficiency of medical treatment and care [15] [16] [17] [18] .

The introduction of electronic health records has also been promoted. The introduction ratio of electronic health records to large hospitals for general practice with 400 or more patient beds was 70.1% in 2015. The introduction ratio has increased annually. However, in 2020, the ratio of electronic health records introduction was less than 50% in hospitals with fewer than 200 patient beds [19] [20] . Especially, small hospitals face difficulties owing to the considerable cost of introducing an electronic health record system [21] [22] .

To solve this problem, our investigation team developed the Psychiatric Nursing Assessment Classification System (PsyNACS^©) [23] [24] in 2015, with the aim of computerizing and improving the quality of psychiatric nursing and care. This system is available at a low cost and can be introduced by small hospitals.

Nursing is characterized by the provision of interpersonal services by nurses. These processes and outcomes of nursing practice, the accurate recording and preservation of nursing care are required by the Enforcement Regulations on the Medical Care Act in Japan. Nurses display more stigmatization and dehumanization toward people with a psychiatric disorder than toward people without a psychiatric disorder. These perceptions are associated with poorer patient care and increased structural discrimination [25] . Therefore, if nurses have a strong dehumanization problem with patients with psychiatric disorders, it may be expressed as an inappropriate description in the nursing record.

Electronic health records enable the uniform management of a large volume of records. However, all the content in nurses’ free writing forms is automatically preserved. At present, these medical records can be requested for disclosure by patients or their bereaved families. Therefore, terms and expressions in records must be ethical, professional, and precise.

This study aims to develop the “Monitoring Application for Inappropriate Expressions in Nursing Records” for use with PsyNACS^©, which will enable psychiatric nursing plans to harness the advantages of ICT.

First, enter information such as basic patient information (patient name, disease name, complications, family structure, insurance type, etc.). Next, the PsyNACS^© database was categorized into nine patient assessment data (PAD), with twoto five cluster assessment data (CAD) each. PsyNACS^© is a classification of items assessing healthcare needs within the Japanese psychiatric nursing care environment. This system can also be used in various psychiatric patient care situations in all psychiatric units. Thirty-one CADs comprise the patient assessment data: (PAD1) “psychological symptom and stress,” (PAD2) “information about treatment,” (PAD3) “function of eating, intake and output,” (PAD4) “life and value,” (PAD5) “vital signs and health assessment,” (PAD6) “self-care,” (PAD7) “social support,” (PAD8) “activity, sleeping, and mobility capability,” and (PAD9) “sexual function and sexual behaviour.” Additionally, nurses can use this information to create nursing care plans for each type of PAD.

This logical system detects inappropriate terms and displays recommendations for appropriate terms.

1) Identification of users who enter information.

2) A necessary database for lists of inappropriate expressions.

3) Development of a matching function to recommend proper writing input, and a warning function.

4) A management function for an inappropriate expression list database.

PsyNACS^© has blank spaces for writing freely as shown below (Figure 1).

1) Identification of users entering information:

This system was designed to have a login function; access is not permitted without entering a username on the PsyNACS^© login screen by a nurse who inputs information on the computer. Thus, the person who has updated the patient record can be identified (Figure 2).

2) Creating a database necessary for inappropriate expression list applications:

Sometimes, there are several cases of inappropriate expressions, such as infringement of human rights and unsuitable usage for a medical professional. Consequently, three categories were created: human rights protection (ethics), unsuitable usage for a medical professional, and improper descriptive usage (abbreviations and others).

A database containing a list of inappropriate expressions was created using information from websites and meetings of development team members. Questionnaire surveys were conducted with 16 head nurses working for specialized psychiatric hospitals in the Shikoku district regarding the contents and expressions that had been written in nursing records and used in daily operations. The data were carefully examined and added to the database. Subsequently, a default database was created (Table 1).

3) Development of the matching function to recommend appropriate writing inputs and the warning function:

We decided that the warning function should cover all written expressions in all the free-text entry fields. Consequently, in addition to free text in the nursing plan, the medical history and symptoms fields were covered as check targets. The timing of the warning is as follows: a warning is shown twice at “confirmation” after writing text and at “screen transition” after input. The two warning messages can reinforce user recognition. However, even when users are warned, the system allows them to input their original expressions as necessary, without modification. In some cases, these expressions have been used to describe patient complaints. To properly assess changes in the mental condition of psychiatric patients, patient descriptions must be recorded without modifications.

The warning system is designed to function as follows: in the case of an inappropriate input, a warning message is shown in the display (Figure 3), and when the recommended terms are listed, these terms are shown.

4) Management function for inappropriate expression list database:

Inappropriate items in the medical records, such as abbreviations and proper nouns, varied depending on the hospital department and hospital. From this

viewpoint, this application prepared a function to add information and revise the database by nursing managers and researchers or system administrators in the hospital using PsyNACS^©. Another function was implemented to output the information enlisted for information management and educational use. By logging into the system on the display (Figure 4) with an exclusive ID and password, users can utilize the database customized for an individual hospital.