Background
The long-term care (LTC) sector has been the epicenter of COVID-19. In Canada, among all COVID-19 cases during the first wave of the pandemic, one in five belonged to LTC residents and one in 10 to LTC workers [
1]. In this timeframe, LTC residents accounted for 81% of all COVID-19 deaths in the country with some jurisdictional variations (e.g., no cases in New Brunswick vs. 97% of all deaths in Nova Scotia). While Canada’s overall COVID-19 mortality rates were considered lower than rates in most countries in the world, LTC rates were shockingly high— double the rate of other member countries of the Organization for Economic Co-operation and Development (OECD) [
2]. In response to the concerning circumstances in LTC and to ensure the health and safety of staff, residents and their families, the Public Health Agency of Canada (PHAC) mandated several rapid redesign and resource redeployment practices such as strict visitation restrictions, COVID-19 screening, use of personal protective equipment (PPE) and the “one high risk site” staffing policy. Emerging pandemic evidence shows despite important contributions to slowing down the spread of infection, these policies were complex and had some unintended consequences for LTC homes, their users, and care providers [
3,
4].
Context
Examining the nuances of the “one high risk” policy (also known as the Single Site Order (SSO)) and its implementation in LTC requires a broad understanding of the sector’s unique context. LTC homes, also known as nursing homes, continuing care facilities, and residential care homes, provide a wide range of health and personal care services for Canadians with medical or physical needs requiring access to 24-hour nursing care, personal care and other therapeutic and support services [
5]. Across Canada, 2,076 LTC homes deliver services to 198,220 Canadians majority aged 65 or older, representing 3% of Canada’s total older adult population [
6]. These care homes are either publicly (46%) or privately (54%) owned. Privately owned homes may have a private for profit (29%), private not-for-profit (23%) or fully private (2%) structure [
5]. An overwhelming majority of LTC residents live with one or more impairments (physical or cognitive) and therefore have highly complex care needs that are met by nearly 254,000 LTC workers nationally [
7].
LTC workers are relatively evenly distributed across fulltime (39.3%) and parttime (38.9%) employment with a smaller proportion (~ 18%) holding casual employment [
7]. A minute proportion of LTC staff are contracted from employment agencies (~ 4%). Despite these national trends, variations exist across jurisdictions and individual care homes. More consistent, however, is the distribution of LTC workers’ classifications and personal demographics. An overwhelming majority (~ 90%) of LTC workers are direct care providers, including unregulated support workers (also known as care aides [CAs]; 65.5%) followed by regulated nurses (registered nurses [RNs] and licensed or registered practical nurses [LPNs]), physicians, and allied health providers (26.1%). Indirect care workers make up a small proportion of LTC staff (~ 10%) [
7]. LTC workers are adversely impacted at the intersection of age, gender, and race; racialized middle-aged or older immigrant women are over-represented in LTC and earn less than their counterparts working in acute care settings [
8,
9]. These workers also have higher rates of self-rated ‘poor’ physical health, primarily due to the heavy physical nature of LTC work [
10].
The SSO was first implemented on March 26th, 2020, in British Columbia LTC homes, and three weeks later in other Canadian provinces including Ontario and Alberta [
11,
12]. The SSO prevented LTC staff from employment at more than one ‘high risk’ site, defined as LTC, assisted living, or provincial mental health facilities [
13]. Other worksites such as acute care settings were not considered ‘high risk’ and therefore were excluded from this definition. The SSO was not equally applied to all LTC workers. While some worker classifications (e.g., nurses, support workers) were included in the SSO, others were excluded (e.g., nurse practitioners, pharmacists, physicians, and dieticians) from the policy [
14].
In British Columbia, the implementation of the SSO involved care homes’ acquiring their staff’s actual and preferred place/s of employment information for submission to their representative health authority. To determine the staff cohort for each LTC home, health authorities reviewed their information along with staffing requirements and available staff resources for each home. Decisions were informed mostly based on operational factors such as safe staffing and continuity of care; when feasible, staff’s personal circumstances and preferences were considered [
13].
To uphold continuity of care, the SSO recommended that regular fulltime employees who also worked as regular parttime or casual employees at any other ‘high risk’ site to be assigned only to their regular fulltime worksite, resulting in a greater loss of casual and parttime employees for LTC homes [
13]. The SSO also recommended that regular parttime and casual employees who worked at more than one worksite to rank their worksite preference. Consequently, the SSO resulted in a smaller pool of staff at individual care homes as well as decreased work hours and increased wage losses for LTC workers [
12]. To compensate for the loss of staff and work hours, care homes encouraged all employees to maximize their hours regardless of their employment status. Regular parttime and casual employees were encouraged to work regularly scheduled fulltime hours, and regular fulltime employees were requested to work overtime [
13].
Preliminary pandemic evidence has suggested that the SSO (and other pandemic-related factors) exacerbated long-standing staffing shortages in LTC homes. For example, a survey study of nearly 4000 support workers from 94 randomly selected and stratified LTC homes in Western Canada found one quarter of respondents worked in more than one LTC home and 15% worked in non LTC sites before the pandemic [
15]. These support workers, on average, worked for 16 h a week at a home other than the primary LTC home where they held a regular position [
15]. Although it has been suggested that employers exploit casual staffing as a strategy for financial savings, staff identified a variety of beneficial reasons for working for multiple employers including greater flexibility in scheduling and increases in pay [
8]. A case study of LTC workers in British Columbia found the SSO compromised staff mental health and quality of care mainly due to exacerbating LTC staffing shortages [
3]. Consistent with this case study, a national survey of over 5000 LTC homes found that 77% reported an increase in the number of overtime hours, 71% reported an increase in absenteeism and 85% reported other staffing challenges such as critical shortages among direct care employees (e.g., nurses and support workers) during the pandemic [
16]. Critical staffing shortages are defined as a shortage of staff, particularly in key roles (such as directors of care, nurses, or support workers) that had an impact on the quality of resident care and employee safety [
16].
While the available literature has identified valuable information regarding the impact of the pandemic and related policies (such as the SSO) on staff and resident outcomes, most research in this area has been based on leaders’ and staff’s perceptions and reports. To our knowledge, these survey findings have not yet been triangulated with other data sources such as administrative data. As such, the aim of this study was to evaluate the impact of the SSO on LTC homes and their staff utilizing a mixed-methods study involving the analysis of routinely collected administrative data and in-depth interviews with leadership and staff.
Methods
Data collection took place between February and April 2021 in partnership with four publicly funded LTC homes across three municipalities (Vancouver, Richmond, Surrey) and two health authorities (Vancouver Coastal [VCH] and Fraser Health [FHA]) in British Columbia, Canada. Table
1 provides an overview of the partner care home characteristics. Quantitative data were used to track potential changes in LTC homes’ staffing practices before versus during the implementation of the SSO, and qualitative data focused on gaining a nuanced understanding of the policy impact from the perspective of LTC leadership and staff.
Table 1
Characteristics of participating LTC + facilities
Municipality | Mission | Vancouver | Richmond | Vancouver |
# staff | ~ 250 | ~ 400 | ~ 280 | ~ 160 |
# residents | ~ 151 | ~ 250 | ~ 250 | ~ 130 |
To ensure an integrated knowledge translation (iKT) approach, a steering committee of 10–15 LTC stakeholders was established. The steering committee was composed of leadership, nursing, support workers and resident and family representatives from the four partner care homes and/or their respective health authority and met regularly to inform the direction of the research and to offer consultation on all aspects of the project. Select members of the steering committee formed a data subcommittee to guide the quantitative component of the project including data identification, operationalization, and extraction. Harmonized ethics approval was obtained from the Research Ethics Boards of the University of British Columbia, the partner care homes and/or their respective health authority [H20-03967].
Quantitative methods
The quantitative component of the study entailed extracting and analyzing retrospective and prospective staffing data from the administrative database of each partner care home or from that of their representative health authority. Aggregated staffing data were extracted at quarterly intervals between April 2019 and March 2021. The period between April 2019 and March 2020 (four quarters) was defined as ‘pre-COVID-19’, and the following period between April 2020 and March 2021 (four quarters) as ‘during COVID-19’. The research team and the data subcommittee worked together to identify staffing indicators (a) most important in relation to the SSO and (b) consistently operationalized across the partner care homes.
Quantitative measures
The most pertinent staffing indicators in relation to the SSO were identified as staff overtime, turnover (voluntary and involuntary), vacancy, sick time, and leave of absence. Among these indicators, overtime, turnover, and vacancy rates were consistently operationalized across at least three of the four partner care homes and were subsequently included in the analysis. Sick time and leave of absence were excluded from this analysis due to inconsistent operationalization across each partner care home.
Table
2 provides care homes’ operational definitions of the staffing indicators. Given that direct care support workers (i.e., CAs) and nurses (i.e., RNs and LPNs) make up the greatest proportion of LTC staff [
7], staffing indicators focused only on these designations and were aggregated across each of the three designations.
Table 2
Operational definitions
Overtime rate (in percentage) | The number of productive hours worked by all direct care nursing employees that were callback or non-callback overtime hours, expressed as a percentage. (Note that productive hours include all hours defined as productive by Finance and worked by regular, temporary, full or part-time employees.) | LTC 1, 2, 3, 4 |
Turnover rate (in percentage) | | LTC 1, 2, 4 |
Voluntary turnover rate | The number of direct care nursing employees leaving voluntarily (resignations, retirements) over the total number of employees, expressed as a percentage. | |
Involuntary turnover rate | The number of direct care nursing employees leaving involuntarily (layoffs, disciplinary, administrative) over the total number of employees, expressed as a percentage. | LTC 1, 2, 4 |
Vacancy rate (in percentage) | The number of vacant regular positions for direct care nursing employees over the total number of regular positions, in percentage. | LTC 1, 2, 4 |
Quantitative data analysis
Staffing data were analyzed using descriptive statistics and data visualization methods whereby separate scatterplots per care home for each indicator were generated, descriptively illustrating the change across eight quarters between April 2019 and March 2021. Scatterplots including data from a minimum of three of the four care homes were also created. Each of these plots were supplemented with a two-part linear trendline presenting the overall trends pre- and during pandemic, as well as smoothed data curves for each care home. The two-part linear trendlines were created by separately fitting simple ordinary least-squares regression models to the pre-pandemic datapoints, and the during-pandemic datapoints. Trend slopes were also annotated to the plots to facilitate comparison of trend changes pre and post pandemic onset. Trends were obtained across total direct care nursing staff and specifically by direct care designation (i.e., RN, LPN, CA). Inferential statistics were not used due to the aggregated nature of each datapoint and limited sample size (i.e., only eight time points for each home).
Qualitative methods
Recruitment and sampling
The research team worked with four partner LTC homes to recruit leadership team members and staff for interviews. Purposive sampling [
17] was used to focus on maximizing diversity across job titles and across LTC homes, while selecting individuals that were especially knowledgeable about the SSO. Leadership team members had to be English speaking and staff had to be English or Cantonese speaking.
Leadership team members circulated an email to their staff that invited them to participate in the study. Staff members then contacted the research team to schedule a date and time for their interview and were sent a consent form to read, sign, scan and return to the researchers via email. Participants were not provided with an incentive to partake in the interviews, and interviews took place during staff working hours.
A total of 10 leadership team members and 18 staff participated in the qualitative interviews (n = 28).
Leadership team participants held job titles including CEO, Executive Director, Nurse Manager, CA Manager, Director of Human Resources, and Clinical Operations Supervisor. Out of the 10 leadership team participants, eight identified as female. Staff job titles included RNs, CAs, Laundry Aides, Chefs, and Housekeepers. Out of the 18 staff, one identified as male; 80% were over the age of 40; and length of time working at the facility ranged from 6 months to 37 years. Each care home had at least two staff and two leaders participate in an interview.
Data collection
After obtaining informed consent from participants, interviews were conducted virtually via Zoom at a time most convenient for the participant. Only the interviewees, the interviewer and a note taker were present at the time of the interview. Interviews ranged in length from 45 to 60 min.
Measures
Interview guides were internally developed by the research team with a focus on evaluating the impact that the SSO had on LTC homes and their staff (see Table 3 in the
Appendix for example interview questions).
Qualitative data analysis
We applied a thematic analysis, where data were sifted, charted and sorted in accordance with key issues and themes using five steps: (1) familiarize; (2) generate initial codes; (3) search for themes; (4) review themes; and (5) define and name themes [
18]. The steps are discussed briefly below. Each interview was fully transcribed verbatim. Initially TF read through all the interview field notes and a couple of the transcripts (familiarize). Through team meetings [SS, TF, JSG] a preliminary thematic framework was developed to simplify and focus on specific characteristics of the data (e.g., overtime, staff turnover). The thematic framework consisted of main themes based on key issues and commonalities emerging from the field notes and initially reviewed transcripts (generate initial codes). Using NVivo 12 software, two team members [SS and TF] coded a subset of interviews based on the thematic framework, with discussion among team members, as new codes and sub-themes were identified (search for themes). Full paragraphs were coded so that contextual meaning was not lost. Data were then summarized by charting illustrative quotes that best exemplified the themes (review themes). As part of the interpretive process a series of team meetings were held to discuss the data for common themes and sub-themes (define and name themes).
Discussion
This study has several key findings. Overall, the results from this study indicate that the SSO exacerbated long-standing staffing-related challenges in LTC. The overall rates of overtime, turnover (voluntary and involuntary), and vacancy trended upward before the pandemic, illustrating the staffing crisis in this sector prior to the onset of COVID-19. The qualitative and quantitative data demonstrate that the pandemic and subsequent SSO and related policies resulted in even greater staffing challenges for LTC homes.
The rate of overtime increased more steeply during the pandemic than before. This phenomenon was described in detail by participants in the qualitative interviews, with staff and leadership linking the increase in overtime to the loss of casual staff as a direct result of the implementation of the SSO. Interviewees revealed that increased overtime may have been an indirect product of LTC- and pandemic-specific factors (including the SSO): LTC workers that experienced a deterioration of their mental and physical health needed time away from work to recover, furthering LTC staffing shortages that were resolved through more overtime. This finding is consistent with a recent study that found that pandemic management policies such as the SSO and strict visitation compounded LTC staffing shortages and heavy workloads, which in turn increased mental health symptoms and absenteeism and vice versa [
3]. In this context, and due to the dire shortage of LTC providers nationally [
16], overtime was the only viable solution available to LTC leaders to ensure adequate staffing to meet basic resident needs.
The increase in rate of overtime was not equal across direct care designations: the overtime rate increased at a speedier pace for RNs and CAs, with RN overtime increasing most drastically. This might be because of a more prominent shortage of RNs in the LTC sector and, therefore, an over-utilization of existing RNs to meet staffing needs. Pre-pandemic evidence from the United States used payroll-based staffing data to show half of LTC homes in the country met staffing standards 20% of the time at most, with RNs driving these staffing shortages [
19]. The increasing pattern of overtime is noteworthy and of particular concern: an extensive body of evidence that links nurse overtime to increased burnout, job dissatisfaction, turnover, and patient safety incidents [
20,
21]. Of note is that the declining overtime trend for LPNs was surprising and may be explained by a markedly different staffing pattern that occurred in one of the four care homes as shown in Fig.
1. Related to this finding, while we observed that RNs demonstrated opposite patterns of vacancy rate (increasing) than LPNs and CAs (decreasing) during the pandemic, vacancy rates declined for all three classifications during the pandemic compared to pre-pandemic (in the case of RNs, the vacancy rate increased at a slower rate during the pandemic). This finding makes sense in the context of increasing overtime that was used to cover nurse vacancies. Furthermore, RNs’ pattern of vacancy rate is consistent with the extreme shortage of RNs in LTC.
The quantitative data also pointed to differing outcomes for the three direct care classifications in other staffing data. To illustrate, the decrease in rate of voluntary turnover during the pandemic for CAs was opposite to the increasing pattern observed for regulated nurses during the same time period. This pattern may have been a result of job dissatisfaction and unequal employment opportunities. Job dissatisfaction is one of the most important contributing factors to nurse turnover, [
22] and while the evidence suggests that all three nursing classifications experienced increasingly negative feelings about their job due to working under more precarious conditions (e.g., increased staffing shortages, overtime, workplace discrimination), they did not have the same employment opportunities available to them. RNs and LPNs had a greater number of job prospects during the pandemic (e.g., public health, vaccination clinics, etc.) available to them compared to CAs, enabling these regulated nurses to more easily act on any feelings of job dissatisfaction [
23]. On the contrary, CAs had little choice but to continue working under increasingly severe working conditions.
This finding may also suggest that pandemic-specific factors interacted to further the existing race and gender inequalities most prominently for already disadvantaged LTC employees. Research from the United States found CAs as the most disadvantaged group of LTC workers with respect to education, availability of health insurance, and family and financial resources [
24]. In Canada, one in five women workers are racialized and employed in the lowest paying and most precarious of caring jobs that carry a high risk of exposure to COVID infection and less likely to have access to important protections such as paid sick leave or health benefits; one-third of all women workers work in high-risk jobs and the large majority are employed as CAs [
25].
Finally, involuntary turnover showed a decreasing trend for CAs and RNs but a slightly increasing trend for LPNs during the pandemic. This staffing indicator reflects employees involuntarily leaving the organization due to layoffs, disciplinary or administrative reasons and therefore is a less meaningful staffing indicator than voluntary turnover to evaluate in the context of the SSO.
Qualitative data added additional context for the patterns we observed in the administrative staffing data: an overall upward slope for overtime and vacancy rate had a detrimental effect on remaining staff. It was explained that employees worked more (through an increase in overtime, working back-to-back shifts, and not taking sick time when necessary) which negatively impacted employee mental health. This finding is consistent with research evidence linking nurse overtime to poor mental health [
21].
Training new staff was highlighted as a particular stressor for original employees that were already overworked before the additional training duties were included in their roles. Feelings of work overload and not taking sick time when needed were discussed as factors that lead to increased musculoskeletal injuries for staff, and a potential factor for decreasing quality of care for patients. These accounts are consistent with previous research that identified the sick time policy during the pandemic as inflexible and as the most inappropriate aspect of the LTC work environments according to LTC workers’ reports. Previous research has also found a link between increased workload and negative nurse and patient outcomes [
26,
27].
Consistent with previous research [
26], the interviews painted a picture of a work environment fraught with stress and exposure to traumatic events. Staff and leadership described a perception of lack of safety in their workplaces, particularly with not having adequate protection against contracting COVID-19 (i.e., PPE, vaccines). Further, the prevention of visitation from family and volunteers meant both an increase in workload for LTC workers to ensure that residents were adequately cared for and exposing staff to seeing residents in isolation and psychological distress. Finally, data from the interviews suggested that staff were exposed to racism and discrimination within their workplace from residents. The experience of racism in any workplace is immensely harmful but could be even more so in the LTC sector when considered with the fact that LTC workers are often individuals with more limited access to financial and social supports [
24]. Additionally, fewer employment opportunities and reduced job mobility, as elucidated above, could mean that certain LTC workers (i.e., CAs) could be in a position of experiencing a deeply harmful work environment without the ability to find employment in a safer situation.
Taken together, the qualitative and quantitative data point to immense challenges experienced by LTC leadership and staff during the pandemic. It is clear that the SSO interacted with many other ongoing challenges that the LTC sector faced be findings (e.g., negatively impacted mental health, reduced perceived safety) reflect the scope of implications this pandemic has brought onto the LTC sector.
Suggestions
Through the completion of this study, it also became clear that consistently operationalized administrative data across LTC homes is sorely lacking. The inability to include numerous key staffing indicators (such as sick leave, leave of absence, etc.) hampered our ability to garner a robust understanding that the SSO and pandemic had on staff and their resident care provision. An investment in consistently captured, defined and publicly available data in this sector is an important and necessary step in garnering a better understanding of how to improve outcomes for LTC residents and staff.
Finally, the results of this study point strongly to the need to address the systemic issues that have plagued the LTC sector, and most specifically, the need to bolster the nursing shortage within LTC. Very few staffing policies could have been implemented without issue in the LTC sector given the long-standing lack of investment in this area and staffing challenges. Effective resident centered care delivery is highly dependent upon adequate and appropriate types of care providers. Thus, to improve resident care, it is imperative that the working conditions in LTC are improved to boost recruitment and retention of nurses, particularly RNs, to this sector. Without addressing the shortage, LTCs will suffer from the same disastrous outcomes when the next health crisis occurs.
Limitations
We acknowledge the challenge we encountered in evaluating the isolated impact of the SSO in the context of a crisis that triggered many changes in the LTC sector, including concurrent adoption of several new policies. The design of this study did not allow for an isolated evaluation of only the SSO, but rather, provided a more global overview of the effect that the onset of the pandemic and implementation of subsequent policies within the first year of the pandemic had on the four participating LTC homes.
Additionally, as described above, the availability of quantitative data proved a limitation in this project. The variability in definitions for nursing staffing data has been well established in the literature [
22] and was a challenge in our study as well. The inconsistent operationalization of quantitative data limited our ability to compare and contrast all staffing indicators across all partner care homes. Further, the data that were available may not have fully represented staff behaviour and outcomes. To illustrate, vacancy rates may not have captured employees taking leave of absences or receiving re-assignments (i.e., their role may have appeared to remain filled even if staff were not actively on staff). Similarly, the voluntary turnover rate would not have captured if a staff member was still technically an LTC employee but not taking any shifts. The infrequent timepoints of the data, partly due to their quarterly nature, were another limitation, hindering our ability to determine the statistical significance of the noted trends. Finally, the findings of this study should be cautiously generalized to other LTC homes, contexts, jurisdictions, and time periods as it is unclear if staff experiences in the partner care homes during this specific period in time (i.e., the beginning of the COVID-19 pandemic) are representative outside of our specific study population and data collection period.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.