Background: The elderly are a vulnerable group for health problems as they age. Safe care is essential when caring for the elderly at home because adverse events can happen there, such as falls, pressure ulcers, injuries, or dangers caused by medication errors. Safety culture for the elderly at home needs to be developed so that they can avoid unexpected events. This study aimed to develop the model for safety culture among the elderly called Aging Safe from the Risk (ASRi) to improve the safety culture in elderly care at home.

Methods: This exploratory sequential mixed methods study will consist of three stages. Stage 1, a qualitative descriptive study, will be conducted to explore the perceptions of families, nurses, and school administrators for the elderly regarding the safety of the elderly at home. Data will be collected by interviews and focus group discussions. Stage II will involve developing The ASRi model using the modified Delphi method. The list of elderly safety culture indicators will be assessed by experts and then further developed into a model. In stage III, a model test will be conducted with a non-equivalent control group using a quasi-experimental design. The investigators will provide interventions in the form of education and training to families regarding the safety culture for the elderly at home.

Findings: There is a crucial need for studies assessing the safety culture in elderly care by empowering families at home to improve safety for the elderly. The results of this study will help fill the knowledge gap in this research and can aid in developing public health policy and programs for the elderly.

.

This study aims to obtain a model of elderly safety culture by involving families in self-care at home to increase perceptions of safety culture and reduce adverse events at home. The specific aims are:

1. Exploring the perceptions of families, healthcare providers, and elderly school administrators regarding the safety of the elderly at home;
2. Develop the Ageing Safe from the Risk (ASRi) model of safety culture; and
3. Conducting trials of applying the ASRi model to improve the perception of the elderly safety culture and reduce adverse events at home.

This protocol have been approval from the Medical and Health Research Ethics Committee of the Faculty of Medicine, Public Health and Nursing, Universitas Gadjah Mada-Dr. Sardjito General Hospital with the number KE/FK/0632/EC/2021 on the date in 10 June 2021.

Study design This exploratory sequential mixed methods will include three stages, and the quantitative design will be the most dominant in this research. Setting This study will be conducted in Bandung, the capital city of West Java, Indonesia. Bandung has 30 subdistricts, and there are 4 subdistricts that have elderly schools. Elderly schools in Bandung have 275 members and are managed by Elderly Friends. The elderly school model was established in 2019 developed by Indonesia Ramah Lansia (IRL) West Java. Activities at the Bandung Elderly School include education about elderly health concerns and home visits by Elderly Friends to monitor the health of the elderly at home.

Study stages and data collection Stage I (Qualitative Study) In the first stage, the descriptive exploratory will be used to explore the perceptions of families, the elderly, healthcare providers, and school administrators regarding the safety of the elderly at home.

Data collection will be conducted with in-depth interviews with the elderly and families and focus group discussions (FGD) among the healthcare providers and elderly school managers In-depth interviews will be conducted to obtain complete information about perceptions of families and elderly about their safety, especially in four dimensions: the personal dimensions including family values/beliefs about the safety of the elderly, understanding, attitudes and family motivations towards the safety of the elderly; behavioral dimensions including communication patterns, cooperation, family stress and burden; and environmental dimensions including family efforts in creating a safe environment for the elderly. The interview guides (shown in Appendix 1) which contain open-ended questions and probes to explore the perceptions about safety in elderly care will be tested in pilot study. To increase credibility, researchers will use data triangulation by also observing the conditions of the elderly home. Furthermore, will also conduct interviews about the hopes and needs of families related to the safety of the elderly at home. Data collection also will be done among the elderly to obtain information about the experiences related to safety and security and their expectations regarding their safety at home.

Interviews will be conducted face-to-face at participants' homes. Each in-depth interview and FGD will take between 60-90 minutes. In-depth interviews will be terminated when there is no additional information needed anymore (data saturation). In this pandemic period, the invesitagors will apply health protocols when conducting interviews, by using masks, maintaining safe distance, and ensuring the conditions of researchers and participants are in good health. The FGDs will be conducted in two groups: with healthcare providers and with administrators of elderly schools in four sub-district areas. FGDs will be held by the agreements of the participants, whether offline or online.

Before the interviews, participants will provide their sociodemographic and related information such as age, education, occupation and race. Information about this research will be explained to all participants before they are required to sign an informed consent form. All interviews will be conducted in Indonesian or Sundanese language and audio-recorded. All data in Sundanese language will be translated into Indonesian.

Stage 2 (Quantitative Study) In the second stage, the investigators will conduct Delphi methods with some modification (30) to find out indicator drafts that can be used to develop the ASRi Model. Modifications from the implementation of Delphi will be in stage 1 when researchers directly provide a list of indicators based on the results of the phase 1 study. In addition, in stage two the panel discussions will be conducted online or offline depending on the willingness of the experts. The Delphi methods modification will be done in two rounds. In the first round of Delphi, the investigators will compile a list of statement items that are indicators of the elderly safety culture at home based on the results of the first stage. Informed consent will be obtained from all participants at the first round. The investigators will contact the experts individually by phone or email to explain the purpose of research and implementation of Delphi. Then, the experts will be asked to provide an assessment of indicators in the safety culture of the elderly at home. The investigators will allocate two weeks for Delphi experts to respond in the first round. Reminders will be sent to participants one week before each round begins in order to maximize their participation. Then, we will process and refine the findings according to the assessment of the Delphi first round.

Next, the Delphi second round will do by conducting expert panel discussions. The investigators will invite experts for discussions regarding the results of the first round of Delphi. Discussions will be conducted both online and offline (hybrid). All indicators resulting from the first round will be discussed in this expert panel. The results of the second round of Delphi will be developed into the ASRi model of safety culture in the elderly.

Stage 3 (Quantitative Study) The third stage, testing the model will be done by conducting a quasi-experimental study with a non-equivalent control group using a pre-posttest design. The intervention "Ageing Safe from the Risk" (ASRi) will consist of about 16 weeks. First, the investigators will conduct training of trainers to nurses and elderly friends. They will be given modules. At this stage, the investigators will also prepare for the control groups and intervention groups. Selection of control and intervention groups based on geographical locations.

Two research assistants, with minimum education of bachelor nursing, will conduct a survey for pre-test measurements and make home visits in each groups. Previously, the investigators will explain about filling out the questionnaire to the research assistants.

For the intervention groups, planning will be done to provide education and training. Families will be given modules on Safety Culture in the Elderly. Materials provided about the safety of the elderly include changes in the elderly, the concept of safety in the elderly, the type of danger to the elderly, how to prevent harm in the elderly, the first handling in case of accidents, and the role of the family in the efforts of elderly safety. The material will be delivered by experts in elderly health and patient safety (professionals from health centers, social services, and academics). In addition to education and training, families will also receive home visits to check and monitoring the condition of the elderly at home related to adverse events and the home environment. Measuring will be done every 2 weeks by using a monitoring book.

In the control groups, activities as usual will be carried out by the elderly friends. After 16 weeks, researchers will evaluate perceptions of the elderly's safety culture at home and adverse events at home.

STUDY POPULATION Participants Stage I (Qualitative study) The qualitative study aims to explore the perceptions of family, elderly, healthcare providers, and administrators of elderly schools about safety in the elderly. This study will conduct individual interviews for family and the elderly, and FGD for healthcare providers and administrators of elderly schools.

This study will include the elderly and their families who are members of the elderly schools in Bandung, healthcare providers from the Public Health Centers, and administrators of elderly schools in Bandung. The sampling technique will use purposive sampling and maximum variation. Inclusion criteria for elderly are: living with family; level of independence are independent, mild and moderate based on Barthel Index [31], no impaired cognitive function, and communication.

Inclusion criteria for the family are: family members who live with the elderly and age at least 18 years old or more. Inclusion criteria for healthcare providers are: minimum education background of diploma with one year of work experience in caring for elderly. Inclusion criteria for elderly school administrators are: manager of The Indonesia Ramah Lansia (IRL) West Java who manages elderly schools in Bandung, and administrators of elderly schools and active in managing elderly schools. There are four elderly schools in Bandung, and each elderly school will involve two administrators.

Stage II (Quantitative study with Delphi method) Purpose of study is to develop the Ageing Safe from the Risk (ASRi) model of safety culture. The subjects in this stage II study are experts in elderly nursing, family nursing, patient safety, and stakeholders related to elderly health programs. Criteria of experts in this study are: geriatric doctors with experience caring for elderly at home; healthcare providers in elderly programs; elderly school managers; managers of IRL West Java; lecturers in family nursing, gerontics and patient safety; families who have experience caring for the elderly, and stakeholders from social services and National Population and Family Planning Board.

Stage III (Quantitative study) The purpose of the study is to test the ASRi model in improving perceptions of safety culture and decreasing adverse events in the elderly at home. At this stage of the study, the population is the elderly and family members of the Bandung City Elderly School in 2020. The sample in this study are the elderly and families who are members of elderly schools in Bandung. Purposive sampling will be used to recruit the samples. Inclusion criteria for the elderly are: living with family; have a history of adverse events over the past one year such as falls, pressure ulcers, injuries, and problems due to treatment; can communicate effectively, and have a level of dependence: independent, mild, and moderate. Inclusion criteria for family are: taking care of the elderly directly, can communicate effectively, and have a minimum age of 18 years. Withdrawal criteria: family did not continue the study, elderly were hospitalized, and died during the study.

Sampling methods and sample size Stage I (Qualitative study) The sampling technique will use purposive sampling. The investigators (AHS) will recruit participants for individual interviewing in subdistrict Central Antapani, Sukajadi, Ujung Berung and Cinambo based on the inclusion criteria. First, investigators will contact the Elderly Friends in each elderly school to recruit for the participants. Then, the investigaot will contact the participants and make an appointment. The number of participants suggested in qualitative studies is 4-10 participants for individual interviews. The number of participants will depend on the saturation of data to be agreed upon by the team.

Meanwhile, the participants for FGDs will be recruited based on inclusion criteria. The investigators will divide the two groups, namely the group of healthcare providers and elderly school managers. For the healthcare providers group, investigators will visit four Public Health Centers, then contact the nurses and managers in the elderly programs. Meanwhile, for the elderly school administrators, the investigators will contact them by WhatsApp chat group and make appointments. Each group consists of 4-12 participants.

Stage II (Quantitative study: Delphi methods modified) The sampling technique will use purposive sampling based on inclusion criteria. There are no standards regarding the sample size in Delphi studies. The sample size varies in many studies from 4 to 171 experts. A minimum of 20 experts will be recruited to this Delphi study. Experts will be invited to participate by the research team. They will be invited by email with a formal invitation to become a part of the panel.

Stage III (Quantitative study: Quasi experiment) The sampling technique will use purposive sampling. The sample count will be calculated based on previous studies, with the proportion of adverse events in the control group as 31.43% and the intervention group as 53.33%. Based on the proportions of previous studies, the researchers will calculate the number of samples needed. Researchers will use a sample calculator, Sample Size Version 2.0, from the World Health Organization (WHO) with Zα= 1,96 and Zβ = 0,84 and obtain 62 results for each group. In anticipation of the possibility of respondents dropping out, 10% of the sample number to 62 + 10% = 68 family members. Furthermore, the researchers will divide the control and intervention groups based on geographical locations so they are far apart from each other. The control group and the intervention will not be allowed to have communication. The control group is an elderly school in Ujung Berung and Cinambo sub-district, and the intervention group is in Central Antapani and Sukajadi sub-districts.

Data analysis Stage I (Qualitative) The investigators will analyze the data with content analysis guided by the total safety culture model. The investigators will read each transcript and then determine the coding. Coding will be done by AHS and discussed with the coauthor CE. Codes will be applied to raw data, and then grouped into categories to generate themes. Consensus concerning the codes and themes among at least three investigators (AHS, CE and PS) needs to be reached. Peer debriefing will be used to increase the credibility and the reliability of the data. Confirmability will be enhanced by keeping an audit trail of the documents.

Stage II (Quantitative study with Delphi method) Advance validity tests are conducted through readability tests by experts based on the ratings given by experts. In the first round of Delphi, the experts will be given a list of indicators that shape the dimensions of elderly safety culture at home. The investigators will provide the essential indicators that are developed for the model. Researchers will include a questionnaire to assess them by giving a score of 1-9. A score of 1-3 means it does not matter, 4-6 means less important, and 7-9 means very important. This scoring will be done individually where the researchers will meet the experts and explain the purpose of the study and provide a list of statements of indicators of elderly safety culture. The experts can revise or add to the indicator by filling in the response columns.

After obtaining the results of the first round Delphi, it will be continued with Delphi's second round. Items which do not achieve consensus in Round 1, with items that need to be revised and additional items will be discussed in Round 2.

The investigators will use median, minimum and maximum scores to analyze the data and will be reported to panel members as feedback after each round. The assessment of Delphi Round 1 and Round 2 will indicate if the item scored <7 it is considered not valid and > 7 the item is valid.

Stage III (Quantitative) Data analysis will be conducted with univariate, bivariate and multivariate analyses. The investigators will also perform normality and homogeneity tests before determining parametric or non-parametric statistical tests. Tests of normality and homogeneity are conducted to determine the equivalence between intervention and control groups. Homogeneity tests will be performed by the Levene test. When the p-value is more significant than alpha, the intervention and control groups can be inferred as equivalent or homogeneous. The normality test will use Kolmogorov Smirnov tests.

Data security All individual data will be kept in privacy and all members of team will be assured to data secrecy. The data collected from all work study will be anonymized and kept in accordance with the appropriate data protection guidelines.

Outcomes Stage I (Qualitative) The outcome of qualitative research will describe the themes about the safety in elderly related to the culture of safety in the elderly at home. The emerging themes will be adapted to the theory of Total Safety Culture which consists of three dimensions, including the personal, behavior and environmental dimensions.

Stage II (Quantitative) Based on the results of stage I, and the literature review, the investigators will formulate potential indicators to rate by the experts using the Delphi methods in stage II. The results of stage II, will be stated as indicators of elderly safety cultures. The list of indicators then will be organized into models.

Stage III (Quantitative study) The assessment of the culture of the safety of the elderly at home was developed from the results of a literature study on the instrumented assessment of safety culture. Based on the results of the study, there is no instrument that assesses the culture of elderly safety in the family, so we will make some modifications to the Nursing Home Survey of Patient Safety Culture (NHSOPSC) issued by AHRQ . The NHSOPSC consists of 42 questions plus two outcome questions, using answers in the form of Likert scales ranging from strongly agree, agree, neither agree or disagree, disagree, and strongly disagree. All items with positive statements are given a score of 5 (Strongly agree), 4 (agree), 3 (neither agree or disagree), 2 (disagree) and 1 (strongly disagree). Negative statements are scored as follows: If there is an answer they do not know, then it is considered missing (not answered) and will not be counted. Furthermore, the percentage score of 1-2 will be combined into a negative response, a score of 3 is neutral and a score of 4-5 is a positive response. Each dimension of the safety culture will be calculated on the average of the percentage of each item in that dimension and only calculated for a positive response.

Adverse events will be measured by the type and frequency (no events, 1-2 events, >3 events) experienced by the elderly at home before and after the model trial. Sociodemographic information will be collected in the baseline assessment, including age, sex, educational level, occupational, and relation with elderly.

Data management plans Interviews will be recorded, and all data will be transcribed verbatim and managed using Opencode software 4.03 version. Data in the Sundanese language will be translated into Bahasa Indonesian. Data will be entered into data collection sheets and confidentially stored in ongoing computer database. For quantitative data, th einvestigators will use SPSS release 22 (IBM Corp., Armonk, NY) administered by UGM. For qualitative study, the investigators will report the study by Consolidated Criteria for Reporting Qualitative Studies (COREQ) and the quantitative study will be reported by Strengthening the Reporting of Observational Studies in Epidemiology (STROBE). For mixed methods study, the investigators will reported by Good Reporting of a Mixed Methods Study (GRAMMS) .

Ethical considerations Investigators will inform respondents either verbally and in writing about the objectives, methods, procedures and measurements during the study. They will also be informed about ethical issues such as confidentiality, their right to ask any questions during the study, and their right to withdraw at any time without consequence. To confirm that all respondents have received appropriate information about the study and have agreed to participate, all participants will be asked to sign a written consent form. This research will be conducted using the approval from the Medical and Health Research Ethics Committee of the Faculty of Medicine, Public Health and Nursing, Universitas Gadjah Mada with the number KE/FK/0632/EC/2021.

Inclusion Criteria:

• Inclusion criteria for the elderly are: living with family; have a history of adverse events over the past one year such as falls, pressure ulcers, injuries, and problems due to treatment; can communicate effectively, and have a level of dependence: independent, mild, and moderate.
• Inclusion criteria for family are: taking care of the elderly directly, can communicate effectively, and have a minimum age of 18 years.

Exclusion Criteria:

• family did not continue the study, elderly were hospitalized, and died during the study.