Burnout among care workers in long-term care institutions: a systematic review and meta-analysis protocol

Abstract

Introduction

Care workers play a fundamental role in delivering care services in long-term care institutions. Burnout has been found to have a negative impact on care recipients and organisations providing care. Little is known about the key factors associated with care workers’ burnout. This systematic review aims to explore the prevalence, severity and correlates of burnout among care workers before and during COVID-19 pandemic.

Methods and analysis

A five-stage framework outlined by Whittemore and Knafl will be used. The following databases will be used to identify relevant literature, including Medline (PubMed), EMBASE, Cochrane library, PsycINFO, CINAHL, Scopus and Web of Science. RevMan will be used to assist the meta-analysis. Heterogeneity of the included studies will be tested using the I² test.

Ethics and dissemination

No ethics approval is required as this study only involves secondary data analysis. The findings will be published in peer-reviewed journals and presented at relevant conferences.

PROSPERO registration number

CRD42024499178.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• A systematic search strategy will be developed to explore the prevalence, severity and correlates of burnout among care workers before and during COVID-19 pandemic.

• The review will be guided by the framework developed by Whittemore and Knafl, and the quality of articles will be assessed using Joanna Briggs Institute tools.

• The ecological framework will be used to integrate and analyse the influencing factors associated with burnout.

• Only articles published in English or Chinese will be included in the review.

• The results may be affected by publication bias, as is the case with all systematic reviews.

Introduction

The increasing ageing population poses significant challenges for healthcare infrastructure worldwide.¹ According to a United Nations report, the global population aged 65 years or above is projected to rise from 800 million (10%) in 2022 to 1.5 billion (16%) in 2050.² The ageing trend aggravates the physiological decline of older adults, leading to a concurrent increase in the number of older adults with disabilities predisposed to chronic health conditions and injuries.³,⁵ China has the largest ageing population in the world, and the largest number of older people with disabilities (44 million).⁶ Notably, many older adults with physical and/or cognitive impairments choose to live in long-term care (LTC) institutions (eg, nursing/care home and assisted living facilities), leading to the increased demand for the number of care workers (nursing assistants, nursing aides, care givers and care providers) and the quality of care^{7 8} (estimated number of LTC institutions and care workers is provided in online supplemental file 1).

Burnout, first described by Freudenberger (1974),⁹ is a psychological condition involving a prolonged response to enduring interpersonal stressors. The concept was further elaborated by Maslach (1981),¹⁰ who identified three key domains of burnout: emotional exhaustion, depersonalisation, reduced sense of personal accomplishment and created a measurement tool (Maslach Burnout Inventory) to assess burnout based on the three domains, which still remain widely used.¹¹

Care workers are the largest workforce in LTC institutions, who provide personal care (eg, washing and dressing), administer medication and assist with activities of daily living.¹² Their dedication directly influences the quality of care received by residents in LTC institutions.¹³,¹⁵ Care workers are prone to burnout due to heavy workload, low wages and lack of promotion channels.¹⁶ Burnout was found to be related to job dissatisfaction, high staff turnover in care workers,¹⁷,¹⁹ poor patient outcomes (increased pain and pressure sores), abuse and neglect (unnecessary physical restraint).^{20 21}

The COVID-19 pandemic has had a profound impact on older adults in LTC institutions, who had increased risk of death from COVID-19 pandemic due to their underlying medical conditions.²²,²⁵ Care workers caring for older adults in LTC institutions were under unprecedented stress, as they were particularly vulnerable to COVID-19 infection, and feared for the safety of their own, family members and patients.^{26 27} This complex web of worry and stress undeniably rippled outward, and affected the quality of care they provided.²⁸

A range of correlates of burnout among care workers have been reported, including organisational, workplace, interpersonal and individual factors. Some factors are protective (eg, high self-esteem, strong organisational commitment, access to sufficient resources, empowerment to control work tasks and adequate staffing levels), which can mitigate burnout. Conversely, other factors are risk factors (eg, role conflict, emotional labour and dysfunctional coping mechanisms) that contribute to burnout.²⁹,³⁴ Spirituality, work environment and recipients can also affect burnout.

Most systematic reviews on burnout among healthcare workers focus solely on nurses, particularly those working in high-stress settings such as intensive care units, accident and emergency department and oncology service.³⁵,³⁷ A limited number of systematic reviews have been conducted to understand burnout among care workers in LTC institutions. Of these, one review focused on burnout of staff working in dementia care (leaders, managers and care assistants) and one review described the prevalence and correlates of burnout, but no meta-analysis on contributing factors was conducted.^{19 34} This leaves a knowledge gap regarding the burnout experiences of care workers in LTC institutions.

The aim of this systematic review is to explore the prevalence, severity and correlates of burnout among care workers in LTC institutions before and during COVID-19 pandemic. This paper presents a protocol of this review.

Methods

Study registration

This protocol was registered on PROSPERO (no. CRD42024499178) and will be reported following the Preferred Reporting Items for Systematic Review and Meta­-Analysis (PRISMA) Protocols statement guidelines.

Study design

The systematic review will follow the framework outlined by Whittemore and Knafl,³⁸ encompassing five key stages: (1) problem identification, (2) literature search, (3) data evaluation, (4) data analysis and (5) presentation. The review will be conducted from April to August 2024.

Stage 1: problem identification

Four research questions have been formulated for this review.

1. What is the prevalence of burnout among care workers?

2. What is the severity of burnout among care workers?

3. What are the factors (including risk factors and protective factors) that influence burnout among care workers?

4. Does the level of burnout among care workers differ before and during COVID-19 pandemic (December 2019–May 2023)?

Stage 2: literature search

A thorough search will be conducted to capture relevant literature published in English or Chinese.

Information sources and search strategy

The following databases will be searched: Medline (PubMed), EMBASE, Cochrane library, PsycINFO, CINAHL, Scopus and Web of Science.

Search terms will be incorporated into titles, abstracts and subject headings as relevant. A summary of the search strategy for each database is listed in online supplemental file 2.

Eligibility criteria

Titles and abstracts of articles which directly matched the identified search terms will be filtered using the inclusion and exclusion criteria listed in table 1.

Table 1. Inclusion and exclusion criteria.

Theme	Inclusion criteria	Exclusion criteria
Geographic	All	None
Participants	Care workers who worked in LTC facilities	Care workers who worked at older people’s home
Exposures	Publications measured burnout with MBI	Papers that did not use the MBI
Outcome	The outcome has quantitatively analysed the burnout and/or the relationship between care workers’ burnout and another factor	The outcome has not quantitatively analysed the burnout and/or the relationship between care workers’ burnout and another factor
Research design	Quantitative study designs	Qualitative study designs

LTClong-term careMIBMaslach Burnout Inventory

Stage 3: data evaluation

Two reviewers (YS and LZ) will independently screen the search results. A third reviewer will be consulted should any disagreement occurs. All references will be uploaded onto Covidence, a platform for comprehensive literature management.³⁹ Covidence will automatically identify and remove duplicate entries. First, the studies will be screened based on the titles and abstracts against the inclusion and exclusion criteria. Second, the selected publications will undergo a full-text evaluation using the inclusion/exclusion criteria. Third, eligible publications identified through the screening process will be subjected to data extraction. Relevant data will be systematically extracted by two reviewers (YS and LZ). The extracted data will include information about each publication (authors and year of publication), details of each study (country of study, study design, sample and sample size), key findings relating to burnout and correlates of burnout.

The PRISMA flow diagram will be used to provide a standardised visual representation of the number of articles identified and screened at each stage of the review process.

Quality assessment

Two authors (YS and LZ) will independently evaluate the presence of bias and concerns regarding the applicability of the studies. Relevant Joanna Briggs Institute critical appraisal tools will be used for evaluating the methodological quality of included studies. The tools consist of a number of items according to study designs: 8 items for cross-sectional studies, 9 items for quasi-experimental studies, 10 items for case-control studies, 11 items for cohort studies and 13 items for randomised controlled trials.⁴⁰,⁴² Each question is answered as ‘yes’, ‘no’, ‘unclear’ or ‘not applicable’. If one question is answered as ‘no’ or ‘unclear’, the study in question is considered at high risk of bias and will be excluded from the review.

Stage 4: data analysis

Meta-analysis will be performed using RevMan Software (V.5.3, Cochrane Collaboration, London, UK). Baseline data will be examined for studies with multiple data collection points. We will calculate a 95% CI for the prevalence or mean of burnout where sufficient information is available. Prevalence from each study will be transformed using the Generalised Linear Mixed models before pooling.⁴³ The criterion for conducting meta-analysis is the identification of three or more studies with comparable burnout measures. Hedges Q statistic and I² statistics will be used to assess between-studies heterogeneity. A p value <0.05 for the Hedges Q statistic indicates statistically significant heterogeneity. If the result is I² >50.0%, heterogeneity is considered to be evident, and a random-effects model will be chosen; if the result is I² ≤50.0%, the heterogeneity will be regarded as being within an acceptable range, and a fixed effects model will be selected.

Factors associated with burnout will be categorised using the ecological framework proposed by Bria et al⁴⁴ and modified by Yeatts et al.³⁰ According to the framework, correlates of burnout are categorised into four broad factors: organisational factors, work-related factors, interpersonal factors and personal factors (figure 1). OR (95% CI) will be used to assess whether a factor is a risk factor or protective factor. OR=1 indicates that the factor has no effect on the occurrence of burnout; OR >1 indicates that the factor is a risk factor; OR <1 indicates that the factor is a protective factor.

Figure 1. The ecological framework on burnout among care workers (based on the framework developed by Yeatts et al).³⁰.

Publication bias will be assessed using Egger’s and Harbord’s tests. Variables such as gender, age, country, before and during the COVID-19 pandemic may be used for subgroup analyses. By omitting one study at a time, a sensitivity analysis will be conducted to evaluate the stability of the combined results. A leave-one-out sensitivity analysis will be performed to determine the influence of each study on the overall effect. A p value <0.05 will be considered statistically significant.

Stage 5: presentation

The key findings will be organised in a table. The asymmetrical shape of the funnel plot and p value <0.05 for Egger’s test will be used to present potential publication bias. Forest plots will be employed to graphically depict the variability in burnout prevalence across studies. Subsequently, the findings will be synthesised and presented in an article for journal publication.

Patient and public involvement

No patients and public were involved.

Ethics and dissemination

Ethical approval is not applicable for this review. The findings will be disseminated via a relevant health professional journal and national and international conferences.