Report on fraying resilience among the Ontario Registered Practical Nurse Workforce in long- term care homes during COVID- 19

Aim: Registered Practical Nurses (RPNs) are frontline healthcare providers in Ontario long- term care (LTC) homes. Throughout COVID- 19, RPNs working in LTC homes experienced prolonged lockdowns, challenging working conditions, and inadequate resource allocation. This study aimed to describe the personal and professional resilience of RPNs working in LTC during the COVID- 19 pandemic. Design: An open cross- sectional online survey containing the Connor– Davidson Resilience Scale, Resilience at Work Scale®, and Resilience at Work Team Scale®. Methods: The survey was distributed by the RPN Association of Ontario (WeRPN) to approximately 5000 registered members working in Ontario LTC homes. Results: A total of 434 respondents participated in the survey (completion rate = 88.0%). Study respondents scored low on measures of resilience and reported extreme levels of job (54.5%) and personal (37.8%) stress. Resources to support self-care and work- life balance, build capacity for team- based care practice(s) are needed.

for LTC nurses and services is outpacing the investment and organization in infrastructure, and policy (Bell, 2021). Further, the high rate of turnover in nurses, in a profession that relies heavily on skill, knowledge and experience acquisition, is a major source of inefficiency (David & Brachet, 2011). Relative to other health sectors, the COVID-19 pandemic disproportionately impacted Canadian LTC homes, residents, families, and all staff working in them, including unregulated staff (Hsu et al., 2020). A total of 72% of all COVID-19-related deaths in Ontario occurred in LTC homes while only 54% of the health care providers working in LTC agreed that COVID-19 recommendations were a feasible strategy for managing the pandemic (Siu et al., 2020). As a result, the public became starkly aware of the social inequities in the LTC sector (Siu et al., 2020) and the indispensable nature of the approximately 5000 RPNs (Lankshear & Rush, 2018) providing nursing care to older adults living in Ontario LTC homes. The pandemic added significant strain to RPNs working in LTC who historically had experienced high levels of burnout, turnover and working in an environment that is inadequately staffed (White et al., 2021).

| BACKG ROU N D
With ever-rising reports of stress and burnout in the nursing profession the concept of resilience has emerged as an essential attribute for nurses' wellbeing, gaining attention in both research and clinical practice (Cooper et al., 2020). Resilience is key for health professionals to allow them to successfully, and continuously, navigate complex and stressful work environments (Huey & Palaganas, 2020).
Low resilience in the nursing workforce has been found to cause increased health costs, low staff retention and poorer patient outcomes (Mealer et al., 2014;Potter et al., 2013;Rushton et al., 2015).
Resilience is not exclusively an individual trait and is largely impacted by the quality of a person's social and physical ecology (Ungar, 2011).
For the purposes of conceptualizing resilience in this research, we used the ecological model of resiliency proposed by Ungar (2018), in which resilience is understood to be a "sequence of systemic interdependent interactions through which actors (whether persons, organisms, or ecosystems) secure the resources required for sustainability in stressed environments" (p. 2). Personal resilience is conceptualized as "a process by which people 'bounce back' from adversity, frustration and misfortune using the psychological and biological strengths humans employ to cope with challenges and threats" (Newman, 2003, p. 42). Similarly, professional resilience addresses the capacity of individuals to thrive in demanding workplace situations, exemplified by attitude and willingness to act in responding to difficult situations (Be You, 2020).

| Long-term care
LTC presents a stressful work environment with increasing medical complexities, structural deficiencies and resources, and insufficient staffing levels (Siu et al., 2020). According to the Long-Term Care Staffing Study (2020), the healthcare sector ranks second highest for injuries resulting in time lost in Ontario, and people working in LTC are among the most at risk for physical injury in the healthcare sector (Ministry of Long-Term Care, 2020). As of October 2020, nearly three quarters of Canada's COVID-19-related deaths had occurred in LTC (Siu et al., 2020). Evidence suggests that mortality risk in older adults in Ontario is concentrated in LTC, and this risk has increased sharply over the course of the pandemic (Fisman et al., 2020). Researchers from the COVID-19 Ontario Remodelling Group advise that early identification of risk requires a focus on testing and provision of personal protective equipment to staff in LTC and restructuring the LTC workforce to prevent the spread of COVID-19 (Fisman et al., 2020). Ontario rates of mortality in LTC are greater than that in other provinces, such as British Columbia, where researchers have suggested there was greater preparedness compared with Ontario: there was better coordination between LTC, public health, and hospitals; greater funding of LTC; more care hours for residents; fewer shared rooms; more non-profit facility ownership; and more comprehensive inspections . Jackson et al. (2007) views resilience as a quality that is necessary to succeed in nursing and is "favourable to build… as a strategy for assisting nurses to survive and thrive" (p. 7). A review of healthcare worker resilience during the COVID-19 pandemic (Baskin & Bartlett, 2021) suggests that building resilience in nurses and other healthcare workers can serve as a protective factor against negative outcomes related to the job, including burnout, anxiety, and depression, and can improve patient outcomes. The integrative review examined 191 studies that assessed resilience during COVID-19. Results demonstrated that resilience scores of nurses in some countries (i.e., The United States; Petzel, 2021) suggest a decrease in nurse resilience, and nurses in other countries (i.e., China; Lyu et al., 2020) suggest an increase, when compared with pre-pandemic levels.

| Resilience in nurses
Further, evidence from a cross-sectional study of 185 frontline nurses during the COVID-19 pandemic suggests a relationship between frontline nurses' psychosocial status, satisfaction with life and resilience (Zakeri et al., 2021). In this study, nurses worked in intensive care units, the general ward, or other related medical departments in Iran. Non-resilience, as measured by a mean score of 59.87 on the CD-RISC, was significantly associated with higher rates of psychological disorders. These findings implicate resilience as a factor related to nurses' mental health and suggest that it should be considered when supporting nurses during a crisis such as COVID-19.

| The current study
Conceptualizing resilience in RPNs as influenced by individual, professional, and workplace factors is useful in assessing the professional and personal social, emotional, psychological, physical, and organizational/ workplace effects of the COVID-19 pandemic. Therefore, the purpose of this study was to explore how RPNs in LTC were managing stress, working conditions, and building self-care networks to identify the components of personal, professional, and organizational resilience in times of the COVID-19 pandemic. RPNs in Ontario earn a diploma in Practical Nursing by taking a program of four semesters over two years in a college program leading to a diploma in Practical Nursing (RNAO, 2022). The COVID-19 pandemic presents a unique opportunity to study our current gap in knowledge about the resilience of nurses working in LTC. Understanding the existing state of resilience for RPNs in LTC homes and identifying areas most challenging for RPNs will support the development of practice resources, recommendations for practice guidelines, inform institutional and governmental action plans, and influence policy change. Identifying and developing supports for unmet needs in sustaining resilience is critical to maintaining and engaging this workforce in LTC (Clark et al., 2020). of COVID-19 occurrence in their workplace(s), potential changes to living situation and household income, and work location, duties, and responsibilities during the pandemic. Respondents were also asked to rate their current physical and mental health status when compared with before the pandemic, and the levels of job and personal stress they experienced since January 2020. The survey was seven pages in total and the number of questions per page ranged from 1 to 42. A total of 121 questions were presented to respondents. Respondents were given the option to navigate backwards in the survey, to skip questions, or not give a response to a question.

| Survey design and development
Adaptive questioning was not used. No time cut-off for the completion of the survey was allocated.

| Resilience measures
As a component of the online survey, respondents were asked to complete three resilience scales to assess their personal resilience, personal resilience at work and team-based professional resilience at work; specifically using the Connor-Davidson Resilience Scale (CD-RISC-25; Connor & Davidson, 2003), Resilience at Work Scale® (R@W; Winwood et al., 2013), and the Resilience at Work Team Scale® (TR@W; McEwen & Boyd, 2018), respectively.
The CD-RISC-25 can distinguish resilient people from nonresilient people in clinical and non-clinical groups and can be used in research and clinical situations (Connor & Davidson, 2003). The CD-RISC-25 measures "personal competence, trust in one's instinct and tolerance of negative affect, positive acceptance of change and safe relationships, control, and spiritual influences" (Manzano & Ayala, p. 246). The scale contains 25 items rated on a 4-point Likert scale ranging from "not true at all" (0) to "true nearly all the time" (4). Total CD-RISC-25 score ranges from 0 to 100, with higher scores indicating greater personal resilience and a cut-off ≥80 is used to characterize the presence of personal resilience (Connor & Davidson, 2003). In their original research, 80 was established as a cut-off score from a sample that contained a "community sample, primary care outpatients, general psychiatric outpatients, clinical trial of generalized anxiety disorder, and two clinical trials of PTSD" (Connor & Davidson, 2003, p. 1). More recent work has established a mean score of 73% on the CD-RISC-25 for nurses working in intensive care units in New Zealand (Yu et al., 2019), 71% for nurses working in Iran responding to the COVID-19 pandemic (Afshari et al., 2021), and 52% in an American sample of nurses working in LTC (Lin et al., 2021). Connor and Davidson (2003) have reported the Cronbach's alpha of the CD-RISC-25 scale to be 0.89, with a reliability coefficient of 0.87 reported for this scale through test-retest reliability in a four-week interval (Connor & Davidson, 2003). The scale has been deemed to have sound validity and reliability (Cronbach's alpha = 0.89; Derakhshanrad et al., 2014).
In previous research from New Zealand of nurses working in intensive care units (N = 93), a CD-RISC-25 mean score of 73% was found (SD = 9.6; Yu et al., 2019). The study sample mean age was 33.9 ± 9.6 years old, with 72.0% of the sample between 20-34 years of age. A total of 73% of the sample reported being female. Similarly, data from a sample of hospital nurses (N = 321) in Iran responding to the COVID-19 pandemic demonstrated a group mean score of 71% (SD = 14.1; Afshari et al., 2021). Approximately 60% of this sample was female and the 20-30-year age group was the largest comprising 54% of their sample. In contrast, a sample of American nurses working in LTC and rehabilitation settings (N = 120) demonstrated a group mean score of 52% (SD = 10.42; Lin et al., 2021). A total of 85% of their participants were female, with a mean age of 42.69 years and an unreported standard deviation. to "strongly agree" (6) with two items reverse-scored. Higher total and subscale scores are indicative of higher resilience (possible range from 0-120; Winwood et al., 2013). On the R@W scale, previous research shows a mean standardized score of 70.27 (N = 482, SD = 11.53) among mental health nurses (Delgado et al., 2020).
In this sample, the Living Authentically subscale (i.e., maintaining personal values, use personal strengths, and have good emotional awareness and regulation at work) had the highest mean score at 79.12 (SD = 12.30), and the Maintaining Perspective subscale (i.e., having the capacity to reframe setbacks, maintain a solution-focus, and manage negativity) had the lowest mean score (M = 52.44, SD = 16.93). Similarly, in a different study, across six hospitals in the western United States (N = 48, mean age = 48) a mean score of 4.2 on the 7-point Likert scale on the R@W has been reported (Carpio et al., 2018). The highest scoring subscale was also Living Authentically with a score of 5.3 (SD = 0.4), whereas the lowest was Maintaining Perspective, with a score of 3.1 (SD = 1.0). Therefore, multiple studies have previously shown that the capacity to focus on solutions at work, reframe difficulties and/or manage negative thinking achieve lower scores than domains capturing individuals' capacity for emotional awareness and self-regulation. industry sectors (i.e., state government, private, and not-for-profit).

| Sample and Recruitment
RPNs working, or who had worked, in LTC homes in Ontario since January 2020 during the COVID-19 pandemic were eligible and invited to participate in the study. Nursing students and other categories of nurses (e.g., Registered Nurses) were not eligible.
Respondents were recruited through their professional association, the Registered Practical Nurses Association of Ontario (WeRPN).
WeRPN sent a series of email invitations, that included the online survey link, over a 5-month period, to approximately 5000 potential respondents currently catalogued as working in LTC homes in Ontario (Lankshear & Rush, 2018). Postings for the online survey were also advertised through the WeRPN newsletter, and social media channels (e.g., Facebook, Instagram, Twitter, and LinkedIn).
A reminder email, sent by the WeRPN, was sent 2-weeks after the initial email to encourage participation as recommended by Sammut et al. (2021). No direct contact was made with potential respondents and survey responses were anonymous. The collection of additional system data (e.g., respondent's IP address, cookies and location) was disabled using Qualtrics software, which uses encryption technology and restricted access authorizations to protect all data collected. No other log file analyses were used. The use of nonprobabilistic sampling, due to the physical and fiscal constraints of obtaining province-wide access to individual contact information, prevented the calculation of a participation rate (i.e., we are unable to determine how many eligible people were exposed to our invitation to participate; Couper, 2000;AAPOR, 2010). Informed consent to participate was obtained on the landing page of the online survey.

| Data management and statistical analyses
Survey data were exported from Qualtrics and organized in Excel software. Data analyses were completed using SPSS Version 25 (IBM). It was determined a priori that only questionnaires that were ≥80% complete would be analysed. Descriptive statistics were used to analyse responses. Any missing data from responses that were between 80% and 100% complete was excluded in descriptive statistic calculations. In the absence of normative data for the R@W and TR@W scales, comparison with other data found with nursing populations will be used for comparison.

| Research reporting checklist
The Checklist for Reporting Results of Internet E-Surveys (CHERRIES) was used in the writing of this manuscript (Eysenbach, 2004; see Appendix 1).

| RE SULTS
A total of 434 RPNs consented to participate in the survey; 51 surveys were <80% complete and were therefore excluded from data analysis. Additionally, one respondent was removed who indicated they were not an RPN. Accordingly, the total number of respondents who consented to participating in the survey was N = 381 (completion rate of survey = 88.0%; see Table 1 duties, locations and responsibilities are presented in Table 2.
Scores for the CD-RISC-25 and R@W individual and team scales were presented both as Likert-scale means and as standardized R@W scores as indicated in the Resilience at Work® Manual (see Table 3; McEwen, 2019aMcEwen, , 2019b With non-family members 11 2.9 Prefer not to answer 9 2.4 Other 9 2.4 Note: In instances where percentages do not sum to 100, not all respondents answered the survey item. Examples of "other" for employment status included things like "Retired during pandemic", "Quit during pandemic", "maternity leave" and "went back to school." Examples of "other for Role/Job Title" included things like "private nurse", "behaviour support manager" or "BSO" and "foot care nurse". a Examples of other include "became RPN during pandemic", "changed health care sectors", "student", "contract ended", "fell and broke my arm", "self-isolation". In instances where percentages do not sum to 100, not all respondents answered the survey item.
having high levels of personal resilience (Connor & Davidson, 2003 Table 4; McEwen, 2019a, 2019b). On this scale, more RPNs were able to develop their capacity to be Connected (i.e., be cooperative and supportive with each other and encourage a sense of belonging), than they were to develop their Self-Care (i.e., promote and deploy good stress management routines, respond to overload, and support work-life balance).
Self-reported current physical and mental health was measured on a 4-point Likert-scale. Respondents were also asked to retrospectively rate their physical and mental health on the same scale. The largest changes were in self-reported physical and mental health were in the categories excellent and fair (see Figure 1). Respondents reported that their physical and mental health before COVID-19 was TA B L E 3 Group data for RPN scores on CD-RISC and R@W Scales.  ); Scores of 0 in the Min-Max column indicate Likert scores of 0 (strongly disagree) converted to percentages (i.e., at least one respondent indicated they strongly disagreed to items in that subscale). a A score below the cut score of 80 as per the CD-RISC-25 to characterize the possession of resilience based on the collated self-reported responses.

CD-RISC-25
b Subscales on which our sample scored lower than Likert-data provided by Carpio et al. (2018).
c Subscale on which our sample scored lower than standardized data provided by Delgado et al. (2021) and is −1.0 to −0.5 SD below the mean on comparison to normative values for Australian workers (McEwen, 2013).

| DISCUSS ION
Due to the increasing complexity of health care needs for older adults living in LTC homes, low staffing levels, and the "invisibility" of health care professionals outside of traditional hospital settings (Hewko et al., 2015), supporting resilience of nurses working in the LTC sector of the health care system is critical (Turner, 2014 Stress and Building Networks subscale scores were lower in our sample compared with previous scores in nurses (Carpio et al., 2018;Delgado et al., 2020 beliefs), and the total R@W scale mean score of 4.7 was higher than the 4.2 mean scale score previously reported (Carpio et al., 2018).
Our sample also scored lower than comparable samples on the It is possible that the data presented in this study are reflective of RPN normative scores, RPN "usual" scores during crisis, or that they are altered by necessity to build their sense of belonging and purpose at work (Finding Your Calling) to compensate for low capacity to reframe setbacks and manage stress (Maintaining Perspective and Managing Stress). On the TR@W scale, our sample scored an average of 4.5/7, which is slightly lower than mean scores reported by McEwen and Boyd (2018). Capacity for Self-Care and Alignment with their team were low on the TR@W when compared with other samples suggesting that, during this time of crisis, these factors may be more challenging to maintain by RPNs working in LTC and requires assistance with convenience and immediacy.
High levels of resilience contribute to the retention of nurses and helps to sustain their psychological health, by offsetting the personal and professional demands of doing the work of nursing, including the fatigue, burnout, stress, post-traumatic stress, anxiety, and depression attendant to this work (Yu et al., 2019). The purpose of this study was to explore how RPNs working in LTC during the COVID-19 pandemic scored on personal and professional resilience assessment measures, and to identify subscales of resilience that resources and supports need to focus on to build a more resilient RPN workforce. A review of healthcare worker resilience during the COVID-19 pandemic (Baskin & Bartlett, 2021) suggests that building resilience in nurses and other healthcare workers can serve as a protective factor against negative outcomes related to the job, including burnout, anxiety, and depression, and can improve patient outcomes. Therefore, identifying and developing supports for the identified unmet needs in sustaining resilience is critical to maintaining and engaging this workforce in LTC (Clark et al., 2020).
Our findings indicate that resources and supports for this workforce should focus on things detailed in the R@W and TR@W scales as

Managing Stress, Staying Healthy, Maintaining Perspective, Building
Networks, Self-care, Alignment, Robust, and Capability rather than things like Interacting Cooperatively and Perseverance.

| Strengths and limitations
To our knowledge, this is the first study describing the individual and  inherent with self-report measures is a limitation of our findings (e.g., recall bias and confirmation bias); however, our data accurately reflect the experiences of the RPNs in our sample. Moreover, our sample was self-selected (i.e., we are unable to determine why or how participants chose whether to complete our survey). It is possible that nurses who perceived themselves to be more stressed were more likely to respond to the invitation to participate because they wanted to share their experience, or that nurses who perceived themselves to be more stressed were less likely to respond to the invitation to participate because they were at maximum capacity already. We acknowledge that our data may not be generalizable to RPNs in all health care sectors.

ACK N O WLE D G E M ENTS
We acknowledge that this research would not be possible without our partnership with the Registered Practical Nurses Association of Ontario (WeRPN).

This research was funded by a SSHRC Partnership Engagement
Grant.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have no conflicts of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request. Randomization of items or questionnaires

E TH I C S S TATEM ENT
To prevent biases items can be randomized or alternated. 11 Adaptive questioning Use adaptive questioning (certain items, or only conditionally displayed based on responses to other items) to reduce number and complexity of the questions. 8

Number of Items
What was the number of questionnaire items per page? The number of items is an important factor for the completion rate.

8
Number of screens (pages) Over how many pages was the questionnaire distributed? The number of items is an important factor for the completion rate. 8

Completeness check
It is technically possible to do consistency or completeness checks before the questionnaire is submitted. Was this done, and if "yes", how (usually JAVAScript)? An alternative is to check for completeness after the questionnaire has been submitted (and highlight mandatory items). If this has been done, it should be reported. All items should provide a non-response option such as "not applicable" or "rather not say", and selection of one response option should be enforced.

8
Review step State whether respondents were able to review and change their answers (eg, through a Back button or a Review step which displays a summary of the responses and asks the respondents if they are correct).

8
Unique site visitor If you provide view rates or participation rates, you need to define how you determined a unique visitor. There are different techniques available, based on IP addresses or cookies or both.

11
View rate (Ratio of unique survey visitors/unique site visitors) Requires counting unique visitors to the first page of the survey, divided by the number of unique site visitors (not page views!). It is not unusual to have view rates of less than 0.1% if the survey is voluntary.

12
Participation rate (Ratio of unique visitors who agreed to participate/ unique first survey page visitors) Count the unique number of people who filled in the first survey page (or agreed to participate, for example by checking a checkbox), divided by visitors who visit the first page of the survey (or the informed consents page, if present). This can also be called "recruitment" rate.

12
Completion rate (Ratio of users who finished the survey/users who agreed to participate) The number of people submitting the last questionnaire page, divided by the number of people who agreed to participate (or submitted the first survey page). This is only relevant if there is a separate "informed consent" page or if the survey goes over several pages. This is a measure for attrition. Note that "completion" can involve leaving questionnaire items blank. This is not a measure for how completely questionnaires were filled in. (If you need a measure for this, use the word "completeness rate".)

12
Cookies used Indicate whether cookies were used to assign a unique user identifier to each client computer. If so, mention the page on which the cookie was set and read, and how long the cookie was valid. Were duplicate entries avoided by preventing users access to the survey twice; or were duplicate database entries having the same user ID eliminated before analysis? In the latter case, which entries were kept for analysis (eg, the first entry or the most recent)?

11
IP check Indicate whether the IP address of the client computer was used to identify potential duplicate entries from the same user. If so, mention the period of time for which no two entries from the same IP address were allowed (eg, 24 h). Were duplicate entries avoided by preventing users with the same IP address access to the survey twice; or were duplicate database entries having the same IP address in a given period of time eliminated before analysis? If the latter, which entries were kept for analysis (eg, the first entry or the most recent)?

11
Log file analysis Indicate whether other techniques to analyse the log file for identification of multiple entries were used. If so, please describe.