INTRODUCTION
Most people in the developed world who are aged 25–65 years (“working age”) are busy building a career, buying or renting a home, establishing and maintaining relationships with a partner, having a family, and often also caring for ageing family members. Working age people frequently “juggle” these responsibilities with other contributions to society through social and organizational networks (eg, schools, churches).
A small proportion of people at this stage in their lives will be faced with a premature death, which has potential not only to diminish their capacity to manage their responsibilities but also to threaten family well-being and financial security. Prevalence data show that 23,600 UK parents with dependent children died in 2015 (Childhood Bereavement Network, 2017), whereas 4,400 Australian parents with dependent children died in 2005 (Financial Services Council, 2005). In the United States in 2008, 1.5 million children lived in a single-parent household because of the death of one parent (Owens, Reference Owens2008). Although many of these deaths are likely to be due to unexpected causes (ie, accidental, suicide, or homicide) and cannot be prepared for, an unknown proportion will be due to an expected premature death.
Previous studies have shown that, regardless of age, life-limiting illnesses are associated with increased needs and significant burden because of interrupted employment and ongoing financial expenditure, which span the whole trajectory of terminal diseases, from diagnosis and treatment (Harrison et al., Reference Harrison, Young and Price2009), through to palliative/end-of-life care (Gardiner et al., Reference Gardiner, Brereton and Frey2014; Hanratty et al., Reference Hanratty, Holland and Jacoby2007; McPherson et al., Reference McPherson, Wilson and Murray2007; Ventura et al., Reference Ventura, Burney and Brooker2014), and into bereavement (Stephen et al., Reference Stephen, Macduff and Petrie2015). Early stage cancer diagnosis and/or survivorship have been noted to have a detrimental effects on work, finances, family roles, and social aspects, especially for those who are younger age or have dependent children or mortgages to pay (Amir et al., Reference Amir, Wilson and Hennings2012; Catt et al., Reference Catt, Starkings and Shilling2016). However, little is known about the challenges of working age people facing an expected death from both malignant and nonmalignant disease.
The current study set out to describe research on the effects that facing an expected premature death has on the employment, financial, and lifestyle of working age people and their families. “Lifestyle” is used here as a broad term for the way individuals and families (or households) live and cope with their physical, psychological, social, and economic environments on a day-to-day basis (Business Dictionary, 2017).
METHODS
A systematic review was conducted and reported in accordance with the guidelines for Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) (Moher et al., Reference Moher, Liberati and Tetzlaff2009).
Eligibility criteria
Eligible peer-reviewed English language articles were collected from the year 2000 onward reporting empirical data from an Organisation for Economic Cooperation and Development (OECD) country on: workplace participation, employment, financial, and/or household and lifestyle concerns of working age adults (25–65 years of age) living with an advanced terminal illness (malignant and/or nonmalignant). Age criteria were designed to accommodate a range of reporting styles and could be satisfied in which ≥50% of the sample were aged between 25 and 65; the sample mean age was within 25–65; or the sample median age was between 25 and 65 and the upper and lower interquartile age included 25–65 years. Studies reporting on families experiencing unexpected deaths such as suicide, motor vehicle accidents, or acute medical events were excluded. Only studies reporting issues and/or concerns from a patient's perspective or studies with mixed cohorts where patients consisted 50% or more of participants were included.
Information sources
Four electronic literature databases (MEDLINE, CINAHL, PsycINFO, PubMed) and CareSearch (the Australian palliative care knowledge network) were searched between 19 and 29 July 2016 from their earliest records. Reference lists of included articles were searched manually for additional relevant studies.
Search
Databases were searched via medical subject headings (MeSH) and keywords relating to palliative care as defined by CareSearch (Palliative Care Search, 2016a, 2016b), together with terms for working age and domains of interest (“needs,” “employment,” “finance,” “end of life,” “bereavement”). MeSH descriptor data were used to define the terms “adult” and “middle aged” and corresponding age ranges (Kastner et al., Reference Kastner, Wilczynski and Walker-Dilks2006; US National Library of Medicine, 2016). This review adopted MeSH age definitions, with “young adult” (those aged between 19 and 24) being excluded and “adult” (25–44 years) and “middle age” (45–65 years) included (US National Library of Medicine, 2016). The MEDLINE search strategy is given in Supplementary Table 1.
Study selection
Search results were imported into EndNote version X7 software, and eligibility criteria applied independently by two reviewers (SK, TL). Titles and abstracts of papers were reviewed and full text articles retrieved where in doubt.
Data collection
Data were extracted by three reviewers using a Microsoft Excel (2010) electronic pro forma under the following headings: author, year, aim, setting, demographics, age, health condition, study design, outcomes, results, and conclusions.
Risk of bias across studies
The methodological quality of included studies was assessed independently by two reviewers (SK and TL) using the Standard Quality Assessment Criteria for Evaluating Primary Research Papers from a Variety of Fields (Kmet et al., Reference Kmet, Lee and Cook2004). This tool enabled appraisal of the varying study designs.
The level of evidence of quantitative studies was determined by two authors independently (SK, TL) using the Australia National Health and Medical Research Council evidence hierarchy (National Health and Medical Research Council, 1999), and a qualitative studies evidence guide developed for the Australian Palliative Residential Aged Care Project Guidelines (Australian Department of Health and Ageing, 2006).
Studies were not excluded based on their quality assessment scores. Studies were graded to be of higher quality if their combined score was >75% and lower quality if their combined score was <75%, which was the cut-off point for inclusion (Kmet et al., Reference Kmet, Lee and Cook2004).
Synthesis of results
A narrative synthesis using methods described by Popay et al. (Reference Popay, Roberts and Sowden2006) including tabulation and textual summary was used to accommodate the variety of included research designs.
RESULTS
Study selection
The search identified 3,726 references, of which 1,636 were duplicates, leaving 2,091 titles and abstracts to be screened. A total of 1,874 records did not meet the inclusion criteria, leaving 216 records for full-text review. Of those, a further 189 records were excluded, leaving 27 studies for final analysis (Figure 1).
Fig. 1. Preferred Reporting Items for Systematic Reviews and Meta-Analysis flow diagram of included and excluded studies. EOL, end of life.
Study characteristics
These studies used a variety of methods within both quantitative (n = 15) (Boland et al., Reference Boland, Boland and Ezaydi2014; Forbes et al., Reference Forbes, While and Taylor2007; Geyer et al., Reference Geyer, Norozi and Buchhorn2009; Halman et al., Reference Halman, Carusone and Stranks2014; Janda et al., Reference Janda, Steginga and Dunn2008; Koopman et al., Reference Koopman, Benbow and Vandervoort2006; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Mack et al., Reference Mack, Cronin and Fasciano2015; McCarthy et al., Reference McCarthy, Phillips and Zhong2000; O'Connor et al., Reference O'Connor, Guilfoyle and Breen2007; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009; Reed et al., Reference Reed, Simmonds and Haviland2012; Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015) and qualitative (n = 12) (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Galushko et al., Reference Galushko, Golla and Strupp2014; Halkett et al., Reference Halkett, Lobb and Oldham2010; Houldin & Lewis, Reference Houldin and Lewis2006; Janda et al., Reference Janda, Eakin and Bailey2006; Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a, Reference Kennedy and Lloyd-Williams2009b; Murray et al., Reference Murray, Kendall and Boyd2004; Osse et al., Reference Osse, Vernooij-Dassen and Schade2002; Raynes et al., Reference Raynes, Leach and Rawlings2000; Turner et al., Reference Turner, Clavarino and Yates2007; Wollin et al., Reference Wollin, Yates and Kristjanson2006) approaches (Table 1) and originated from several OECD countries, including Australia (n = 9) (Halkett et al., Reference Halkett, Lobb and Oldham2010; Janda et al., Reference Janda, Eakin and Bailey2006, Reference Janda, Steginga and Dunn2008; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; O'Connor et al., Reference O'Connor, Guilfoyle and Breen2007; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009; Turner et al., Reference Turner, Clavarino and Yates2007; Wollin et al., Reference Wollin, Yates and Kristjanson2006), United Kingdom (n = 7) (Boland et al., Reference Boland, Boland and Ezaydi2014; Forbes et al., Reference Forbes, While and Taylor2007; Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a, Reference Kennedy and Lloyd-Williams2009b; Murray et al., Reference Murray, Kendall and Boyd2004; Raynes et al., Reference Raynes, Leach and Rawlings2000; Reed et al., Reference Reed, Simmonds and Haviland2012), United States (n = 4) (Houldin & Lewis, Reference Houldin and Lewis2006; Mack et al., Reference Mack, Cronin and Fasciano2015; McCarthy et al., Reference McCarthy, Phillips and Zhong2000; Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015), Germany (n = 3) (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Galushko et al., Reference Galushko, Golla and Strupp2014; Geyer et al., Reference Geyer, Norozi and Buchhorn2009), Canada (n = 2) (Halman et al., Reference Halman, Carusone and Stranks2014; Koopman et al., Reference Koopman, Benbow and Vandervoort2006), and the Netherlands (n = 2) (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002, Reference Osse, Vernooij-Dassen and Schade2005). The majority of studies included cancer populations (n = 18) (Boland et al., Reference Boland, Boland and Ezaydi2014; Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Halkett et al., Reference Halkett, Lobb and Oldham2010; Houldin & Lewis, Reference Houldin and Lewis2006; Janda et al., Reference Janda, Eakin and Bailey2006, Reference Janda, Steginga and Dunn2008; Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a, Reference Kennedy and Lloyd-Williams2009b; Mack et al., Reference Mack, Cronin and Fasciano2015; McCarthy et al., Reference McCarthy, Phillips and Zhong2000; O'Connor et al., Reference O'Connor, Guilfoyle and Breen2007; Osse et al., Reference Osse, Vernooij-Dassen and Schade2002, Reference Osse, Vernooij-Dassen and Schade2005; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009; Raynes et al., Reference Raynes, Leach and Rawlings2000; Reed et al., Reference Reed, Simmonds and Haviland2012; Turner et al., Reference Turner, Clavarino and Yates2007; Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015), whereas a smaller number focused on neurodegenerative disorders (n = 6) (Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Koopman et al., Reference Koopman, Benbow and Vandervoort2006; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Wollin et al., Reference Wollin, Yates and Kristjanson2006). The remainder involved congenital heart disease (Geyer et al., Reference Geyer, Norozi and Buchhorn2009), HIV (Halman et al., Reference Halman, Carusone and Stranks2014), or combined lung cancer/heart failure (Murray et al., Reference Murray, Kendall and Boyd2004) populations. Most studies included only patients (n = 18) (Boland et al., Reference Boland, Boland and Ezaydi2014; Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Geyer et al., Reference Geyer, Norozi and Buchhorn2009; Halkett et al., Reference Halkett, Lobb and Oldham2010; Halman et al., Reference Halman, Carusone and Stranks2014; Houldin & Lewis, Reference Houldin and Lewis2006; Mack et al., Reference Mack, Cronin and Fasciano2015; McCarthy et al., Reference McCarthy, Phillips and Zhong2000; Murray et al., Reference Murray, Kendall and Boyd2004; O'Connor et al., Reference O'Connor, Guilfoyle and Breen2007; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009; Raynes et al., Reference Raynes, Leach and Rawlings2000; Reed et al., Reference Reed, Simmonds and Haviland2012; Turner et al., Reference Turner, Clavarino and Yates2007; Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015), whereas nine studies had mixed samples of patients/caregiver (n = 7) (Janda et al., Reference Janda, Eakin and Bailey2006, Reference Janda, Steginga and Dunn2008; Koopman et al., Reference Koopman, Benbow and Vandervoort2006; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Osse et al., Reference Osse, Vernooij-Dassen and Schade2002; Wollin et al., Reference Wollin, Yates and Kristjanson2006), or patients/children (n = 2) (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a, Reference Kennedy and Lloyd-Williams2009b).
Table 1. Summary of included studies reporting the impact on employment, financial, and lifestyle from the patient's perspective (n = 27)
*SD not reported in paper.
$\bar X\; $– population mean; HD – Huntington's disease; MND – motor neuron disease; MS – multiple sclerosis; PD – Parkinson's disease; SPARC – Stroke Prevention in Atrial Fibrillation Risk Tool.
Evidence levels: I – systematic review of all relevant randomized controlled trials (National Health and Medical Research Council, 1999); II – at least one properly designed randomized controlled trial (National Health and Medical Research Council, 1999); III-1 – well-designed pseudo-randomized controlled trials (alternate allocation or some other method) (National Health and Medical Research Council, 1999); III-2 – comparative studies with concurrent controls and allocation not randomized (cohort studies), case control studies, or interrupted time series with a control group (National Health and Medical Research Council, 1999); III-3 – comparative studies with historical control, two or more single-arm studies, or interrupted time series without a parallel control group (National Health and Medical Research Council, 1999); IV – case series, either posttest or pretest and posttest (National Health and Medical Research Council, 1999); QE – qualitative evidence (Australian Department of Health and Ageing, 2006).
Quality assessment: H – higher quality (>75%); L – lower quality (<75%). Scores for quantitative studies were calculated based on scoring of 11 (of 14) items (yes = 2, partial = 1, no = 0); three items did not enter in the final calculation as not available (Kmet et al., Reference Kmet, Lee and Cook2004). Scores for qualitative studies were calculated based on scoring of 10 (of 10) items (yes = 2, partial = 1, no = 0) (Kmet et al., Reference Kmet, Lee and Cook2004).
Methodological quality of studies
The highest level of level of evidence was generated by level IV studies (National Health and Medical Research Council, 1999) and all qualitative studies met the qualitative evidence criteria (Australian Department of Health and Ageing, 2006). The methodological quality varied across studies and the heterogeneity of the reported data precluded a meta-analysis from being undertaken (Deeks et al., Reference Deeks, Higgins and Altman2008).
Risk of bias across studies
One-third of included studies scored <75% on quality assessment, with quantitative studies tending to score poorly on controlling for confounding, and qualitative studies tending to score poorly on reflexivity. See Table 1 for scores for each study.
Synthesis
Impact on employment
Being diagnosed with cancer when of working age causes major disruptions to employment (Janda et al., Reference Janda, Eakin and Bailey2006, Reference Janda, Steginga and Dunn2008; Osse et al., Reference Osse, Vernooij-Dassen and Schade2002, Reference Osse, Vernooij-Dassen and Schade2005). More than 60% of Australian patients with brain tumors (n = 75) in one study were not working, with more than half unable to work because of their disease (Janda et al., Reference Janda, Steginga and Dunn2008), whereas more than one-quarter (27.6%) of all patients reported moderate to high supportive needs because of changes in their ability to work (Janda et al., Reference Janda, Steginga and Dunn2008). More than half (53%) of younger Dutch cancer patients (<60 years) reported employment difficulties compared with 24% of older working-age cancer patients (≥60 years) (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005), with income problems also reported more often in younger (40%) versus older (6%) patients (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005). Advanced cancer also influences life roles through changes at work, with another Dutch study reporting that a significant loss of income created feelings of incompetence and shame, and contributed to patients feeling socially dependent on their family (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002).
Employment was also disrupted for people living with nonmalignant conditions (Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Geyer et al., Reference Geyer, Norozi and Buchhorn2009; Halman et al., Reference Halman, Carusone and Stranks2014; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Wollin et al., Reference Wollin, Yates and Kristjanson2006). Only a small percentage (2.2%) of Canadian patients with HIV were employed, with the majority (88.7%) relying on a disability or national pension plan for income (Halman et al., Reference Halman, Carusone and Stranks2014). Full-time employment decreased for people in Germany living with congenital heart disease as their disease progresses (Geyer et al., Reference Geyer, Norozi and Buchhorn2009). This decrease was more prevalent for men than women, who have a lower full-time employment regardless of their clinical status (Geyer et al., Reference Geyer, Norozi and Buchhorn2009). People living with multiple sclerosis were found to experience underemployment (Forbes et al., Reference Forbes, While and Taylor2007) or interrupted employment (Galushko et al., Reference Galushko, Golla and Strupp2014), with impacts on their self-image (Wollin et al., Reference Wollin, Yates and Kristjanson2006). An Australian study found that employment was also significantly affected for other progressive life-limiting neurodegenerative disorders (including motor neurone disease, Huntington's disease, Parkinson's disease, and multiple sclerosis), with more than one-third (36.4%, n = 503) having to give up work/retire early and 8.6% reducing their work hours (Kristjanson et al., Reference Kristjanson, Aoun and Yates2006b). Only one-quarter of these patients living with progressive neurodegenerative diseases and/or their carers were in paid employment, with 60% relying on social security as their main source of income (Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a).
Impact on finances
Financial concerns were universally experienced by patients living with advanced cancer (McCarthy et al., Reference McCarthy, Phillips and Zhong2000; Osse et al., Reference Osse, Vernooij-Dassen and Schade2002, Reference Osse, Vernooij-Dassen and Schade2005; Raynes et al., Reference Raynes, Leach and Rawlings2000; Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015) and other nonmalignant diseases (Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Halman et al., Reference Halman, Carusone and Stranks2014; Koopman et al., Reference Koopman, Benbow and Vandervoort2006; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Wollin et al., Reference Wollin, Yates and Kristjanson2006). One study reported that cancer patients aged <60 years had more significant financial problems compared with those ≥60 years of age (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005). UK cancer patients receiving community palliative care identified needing financial assistance to afford to pay for housework assistance (Raynes et al., Reference Raynes, Leach and Rawlings2000). Many lacked the necessary welfare and financial benefit knowledge to access their entitled benefits, causing them to worry about their being burden to their families (Raynes et al., Reference Raynes, Leach and Rawlings2000). A Dutch study reported that palliative cancer patients struggled with financial issues not necessarily related to their daily life, but to future expenses such as funeral costs (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002). A US study reported that both patients and carers sustained significant financial burden during the past 6 months of life (McCarthy et al., Reference McCarthy, Phillips and Zhong2000). Financial hardship in the United States has been found to be a stronger predictor of use of intensive end-of-life care than insurance, sociodemographic factors, or patient preferences (Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015).
Financial insecurity was also felt by people living with HIV (Halman et al., Reference Halman, Carusone and Stranks2014), whereas the need for financial assistance for care was significant across a variety of debilitating progressive neurodegenerative diseases (Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Koopman et al., Reference Koopman, Benbow and Vandervoort2006; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Wollin et al., Reference Wollin, Yates and Kristjanson2006). An unintended consequence of having a significant disability was financial insecurity as a result of loss of employment (Wollin et al., Reference Wollin, Yates and Kristjanson2006), costs associated with travelling to/from appointments (Galushko et al., Reference Galushko, Golla and Strupp2014), vehicle adjustment (Galushko et al., Reference Galushko, Golla and Strupp2014), and/or adaptation of their environment and mobility support (Forbes et al., Reference Forbes, While and Taylor2007). An Australian study found that 2/3 of patients and carers made adjustments to their living environments with more than a third (34%–44%) self-funding the required modifications (Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a). Another study found that people with motor neurone disease and Huntington's disease required help with finances and information about financial assistance (Kristjanson et al., Reference Kristjanson, Aoun and Yates2006b), whereas the need for financial security was also identified by people with multiple sclerosis (Koopman et al., Reference Koopman, Benbow and Vandervoort2006).
Impact on lifestyle
Facing an expected premature death during working age also adversely affects cancer patients' coping abilities, family roles, and relationships. Increasing disability and cancer-related symptom burden led to a loss of independence and autonomy (Boland et al., Reference Boland, Boland and Ezaydi2014; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005), loss of control over one's life (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005), and increased dependency (Murray et al., Reference Murray, Kendall and Boyd2004; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005). Cancer patients reported struggling to cope with their uncertain future (Janda et al., Reference Janda, Eakin and Bailey2006, Reference Janda, Steginga and Dunn2008; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005;), frustration associated with managing their interrupted/changed lives (Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009), and with increased care requirements and fear of suffering (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002). They also struggled with feelings of futility and stress (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002), uselessness (Murray et al., Reference Murray, Kendall and Boyd2004), and felt excluded from roles, family, and social life (Murray et al., Reference Murray, Kendall and Boyd2004). They wished not to be a burden to others and to minimize that effect of their illness on their families (Raynes et al., Reference Raynes, Leach and Rawlings2000). Cancer was described as taking over their lives (Murray et al., Reference Murray, Kendall and Boyd2004) and severely affected their usual roles at work, leisure, and study (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002).
Cancer patients' ability to perform everyday tasks was compromised (Janda et al., Reference Janda, Steginga and Dunn2008; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009) across various domains, including driving/ability to catch public transport (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005; Raynes et al., Reference Raynes, Leach and Rawlings2000), shopping (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005), housework (Boland et al., Reference Boland, Boland and Ezaydi2014; Osse et al., Reference Osse, Vernooij-Dassen and Schade2005; Rainbird et al., Reference Rainbird, Perkins and Sanson-Fisher2009), meal preparation (Boland et al., Reference Boland, Boland and Ezaydi2014; Raynes et al., Reference Raynes, Leach and Rawlings2000), and personal care (Osse et al., Reference Osse, Vernooij-Dassen and Schade2005). Severe symptom burden (especially pain) and a significant functional decline during the last 6 months of life was reported to adversely affect the ability of advanced cancer patients to perform personal and household tasks and participate in recreational activities (McCarthy et al., Reference McCarthy, Phillips and Zhong2000). People's ability to take care of practical and family issues (including healing rifts in relationships) was compromised because of cancer treatment side effects (such as fatigue and nausea) (Murray et al., Reference Murray, Kendall and Boyd2004). Social and psychological well-being was also affected (Osse et al., Reference Osse, Vernooij-Dassen and Schade2002). For some cancer patients, spiritual well-being was found to correlate with a more productive life because of a determination to have a sense of control over their environment, meaning, and purpose in life (O'Connor et al., Reference O'Connor, Guilfoyle and Breen2007).
The high level of disability associated with progressive neurodegenerative diseases was found to have a significant impact on patients' independence (Wollin et al., Reference Wollin, Yates and Kristjanson2006) and their ability to manage everyday tasks, including personal care (Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b), child care (Forbes et al., Reference Forbes, While and Taylor2007), housework (Forbes et al., Reference Forbes, While and Taylor2007; Galushko et al., Reference Galushko, Golla and Strupp2014; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b), and transportation (Galushko et al., Reference Galushko, Golla and Strupp2014; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b). Assistance with social activities and support groups (Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b) and access to equipment for daily living (Kristjanson et al., Reference Kristjanson, Aoun and Yates2006b) was equally important. Managing household tasks was found to be easier if patients had assistance from a network of family or friends (Raynes et al., Reference Raynes, Leach and Rawlings2000).
People with multiple sclerosis reported struggling to cope with changes in their identity because of interrupted work and/or lifestyle (Galushko et al., Reference Galushko, Golla and Strupp2014); family roles (Wollin et al., Reference Wollin, Yates and Kristjanson2006); diminished ability to perform daily tasks such as personal care, housework, and child care (Forbes et al., Reference Forbes, While and Taylor2007; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a; Wollin et al., Reference Wollin, Yates and Kristjanson2006); and need for assistance with service coordination (Galushko et al., Reference Galushko, Golla and Strupp2014) and transportation (Forbes et al., Reference Forbes, While and Taylor2007; Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a). Leisure activities were also affected (Galushko et al., Reference Galushko, Golla and Strupp2014), and household adaptation and rehousing was often required (Forbes et al., Reference Forbes, While and Taylor2007). Similar to cancer, people with multiple sclerosis wanted to be productive, have control over their lives, and require information about the disease progression (Koopman et al., Reference Koopman, Benbow and Vandervoort2006).
Results from another study found that working age Canadians living with HIV were more likely to face homelessness and interrupted living conditions: of 81 patients, 13.6% reported living on the street or in a shelter and 6.1% reported staying with family or friends (Halman et al., Reference Halman, Carusone and Stranks2014).
In a US study, working age women with metastatic breast cancer reported lower social/family well-being, and were less satisfied with their personal experience of care and support from health professionals compared with those >65 years (Reed et al., Reference Reed, Simmonds and Haviland2012). The same study reported that working age women without children (compared with those with) had significantly higher functional well-being in the last year of life (Reed et al., Reference Reed, Simmonds and Haviland2012). Age was reported to also influence decision-making preferences, with younger patients (<40 years) being more concerned about time away from family and side effects of their cancer treatment (Mack et al., Reference Mack, Cronin and Fasciano2015). A UK study reported that 40% of patients (n = 32) worried about the impact their progressive illness had on their family, whereas 31% reported that their illness had adversely affected their sexual life (Boland et al., Reference Boland, Boland and Ezaydi2014).
Working age parents with cancer reported being unprepared for the disease and experiencing challenges dealing with this diagnosis whilst also managing their children's needs (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Houldin & Lewis, Reference Houldin and Lewis2006). Parents reported trying to cope and limit the effects of the illness on their children by creating as normal and stable life as possible (Houldin & Lewis, Reference Houldin and Lewis2006; Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a), maintaining a positive outlook (Houldin & Lewis, Reference Houldin and Lewis2006), accepting help from others (eg, a partner or other family members) (Houldin & Lewis, Reference Houldin and Lewis2006), and discussing all issues relating to the disease with their children, while also concentrating on management of their physical symptoms and distress (Houldin & Lewis, Reference Houldin and Lewis2006).
They identified discussions with their children as one of the most difficult issues, including telling their children about the diagnosis (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a; Turner et al., Reference Turner, Clavarino and Yates2007), dealing with their questions about the progressive illness (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a), and initiating conversations about the parent's expected death (Turner et al., Reference Turner, Clavarino and Yates2007). Cancer was also reported to interrupt their parenting and family life (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Houldin & Lewis, Reference Houldin and Lewis2006) and decrease their social activities with their children (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a). Their ability to perform household tasks (eg, cleaning, cooking) was also affected (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005), resulting in their older children assuming more household and/or shopping responsibilities (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a). Parents required practical coping strategies to deal with these challenges (Turner et al., Reference Turner, Clavarino and Yates2007). They also desired practical help from health professionals on how to navigate the disease journey (Turner et al., Reference Turner, Clavarino and Yates2007) and information about how the disease would affect their children (Elmberger et al., Reference Elmberger, Bolund and Lutzen2005; Turner et al., Reference Turner, Clavarino and Yates2007). Strengthening the relationship with their children and learning to value what was important was seen as a positive aspect of their disease journey (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a).
DISCUSSION
This systematic review has highlighted that facing an expected premature death adversely affects workforce participation, financial security, and lifestyle for working-age people with life-limiting illnesses and their families. Changes in patients' daily routines and lives as their disease progresses and their symptoms become more pronounced alters the family dynamics and redefines the roles of patients, their partners, and children. These lifestyle changes are accompanied by a change in the patient's perception of self, loss of independence, and increased dependence on others. Although the evidence mostly comes from studies on cancer, these adverse impacts appear to be equally prevalent and cover similar domains and issues for people living with nonmalignant progressive conditions, albeit played out over a more extended timeframe.
Symptom burden also affects patients' ability to maintain employment and perform everyday tasks. Less obviously, there is evidence that people often end up working longer hours to meet additional financial demands to cover illness-related expenses (Zafar et al., Reference Zafar, Peppercorn and Schra2013) before being forced to reduce their working hours or to cease working altogether because of escalating physical dependence (Halkett et al., Reference Halkett, Lobb and Rogers2015). Thus, for this population, continuing employment is likely to be dependent both on physical well-being and a desire to maintain financial security, which can be greatly affected if employment is interrupted.
In addition to underemployment, working age patients face multifaceted financial burden because of healthcare costs as well as the hidden costs associated with home-based care, all of which are often compounded by limited access or lack of financial assistance. Other studies have shown that financial considerations (such as increased healthcare costs and out-of-pocket expenses) significantly affect their medical treatment, including nonadherence to medication (Neugut et al., Reference Neugut, Subar and Wilde2011; Ubel et al., Reference Ubel, Abernethy and Zafar2013; Zafar et al., Reference Zafar, Peppercorn and Schra2013) and quality of life (Lathan et al., Reference Lathan, Cronin and Tucker-Seeley2016). Our review further supports those findings. For example, when faced with competing financial priorities, patients often opt for care that incurs fewer out-of-pocket expenses to reduce the impact on family finances (Tucker-Seeley et al., Reference Tucker-Seeley, Abel and Uno2015). Better understanding the relationship between financial burden and care decisions is important to reducing unnecessary and medically futile treatment at the end-of-life.
The need for financial assistance for care and support services for working age patients with progressive illnesses occurs in all countries, including those with universal healthcare and welfare systems (Kristjanson et al., Reference Kristjanson, Aoun and Oldham2006a, Reference Kristjanson, Aoun and Yates2006b; Raynes et al., Reference Raynes, Leach and Rawlings2000). Although the financial toxicity of cancer treatment in countries without universal healthcare cover is well documented (Huntington et al., Reference Huntington, Weiss and Vogl2015; Zafar, Reference Zafar2016; Zafar & Abernethy, Reference Zafar and Abernethy2013a, Reference Zafar and Abernethy2013b), out-of-pocket healthcare treatment costs are rising in countries with universal healthcare (Currow & Aranda, Reference Currow and Aranda2016; Gordon et al., Reference Gordon, Walker and Mervin2015), especially with the advent of new-generation cancer treatments, prompting calls for full disclosure of costs for the whole episode of care (Currow & Aranda, Reference Currow and Aranda2016). Patients face an ever-increasing demand on their personal finances because of costs of treatments (including new unsubsidized therapies) (Currow & Aranda, Reference Currow and Aranda2016; Ubel et al., Reference Ubel, Abernethy and Zafar2013), innovations (sometimes with marginal benefits) (Currow & Aranda, Reference Currow and Aranda2016), and/or failure to access less costly alternatives (Currow & Aranda, Reference Currow and Aranda2016; Ubel et al., Reference Ubel, Abernethy and Zafar2013). Health professionals need to provide opportunities for discussions on financial issues at all stages of the patient's illness trajectory (Amir et al., Reference Amir, Wilson and Hennings2012; Ubel et al., Reference Ubel, Abernethy and Zafar2013), especially because many patients avoid initiating these conversations during the early stages of the illness as they focus on their diagnosis, treatment, and recovery (Amir et al., Reference Amir, Wilson and Hennings2012).
The impact of loss of income on the family is a high priority requiring urgent social work input, especially in the context of the patient and family not having access to other income streams. There also needs to be consideration of the longer term financial implications, particularly in relationship to the carer returning to work after bereavement. In addition to affecting spousal health (Currow et al., Reference Currow, Burns and Agar2011; Jassem et al., Reference Jassem, Penrod and Goren2015), a prolonged bereavement can adversely affect the longer term employment of carers and their financial well-being, including retirement savings (Girgis et al., Reference Girgis, Lambert and Johnson2012). Also of concern, the premature death of a parent has been found to have a detrimental effect on the future wealth and employment of their children as adults (Australian Bureau of Statistics, 2010). Better understanding how education and health services can collaborate to ensure that school age children are appropriately supported before and after their parent's death is urgently required. Optimizing their chances of successfully completing their schooling is central to their future financial security and emotional wellbeing.
Findings from our review suggest that patients living with debilitating neurodegenerative diseases, including brain tumors, experience significant physical disability and cognitive impairment (brain tumors), often early in their illness trajectory. Their disability greatly affects their ability to perform household/personal tasks. Living with advanced cancer has a huge impact on people's independence, family interactions, and especially parenting roles. Sometimes, the disruption of normal household routine caused by the disease creates a ripple effect within the family, with children taking up the responsibilities around the house (Kennedy & Lloyd-Williams, Reference Kennedy and Lloyd-Williams2009a). Other studies have shown that this can be particularly evident in homes where there is a lack of domestic help, so household duties become an additional burden for adolescent children (Huizinga et al., Reference Huizinga, van der Graaf and Visser2003). Given the daily struggle with loss of independence and increased reliance on their partners and other family members, patients with life-limiting illnesses may experience additional emotional and psychological burden. Studies have shown that adequate rehabilitation and prehabilitation have positive effect on function of cancer survivors and also may improve their employment rates and reduce early retirement (Silver, Reference Silver2014; Silver & Baima, Reference Silver and Baima2013; Silver et al., Reference Silver, Baima and Newman2013). Thus, providing adequate support services to promote independence for as long as possible as well as timely rehabilitation and prehabilitation should be integral to healthcare provision for working age people living with life-limiting illnesses.
The impact that a life-limiting illness has on families' overall quality-of-life is complex. From the time working age patients are diagnosed with a progressive illness, their family's life changes dramatically and irreversibly. Their day-to-day lives change in terms of work engagement, housework chores, interruptions caused by going to medical appointments, and ongoing treatments. Changing roles and responsibilities in the family unit (patient vs. their partner vs. their children) alter the family dynamics. Many of these observed changes persist into the palliative stages of the patient's illness and into bereavement. Although the severity, timing, and priority placed on these issues may vary between malignant and nonmalignant conditions, comparing and contrasting their trajectories and priorities would be helpful to better understand people's needs and inform health policy and practice.
Having a better understanding of the impact an expected premature death has on working age patients and their families provides an opportunity for specialist services, including palliative care to address the holistic needs of patients and their families. This may be particularly important for families who are navigating the final stages of the illness without adequate resources and support systems and where health professionals' involvement would help mitigate the often hidden, but detrimental, effects of the life-limiting illness on that family's financial and emotional welfare (Payne, Reference Payne2017). At the same time, comprehensively addressing the employment, financial, and lifestyle needs is likely to fall outside the jurisdiction of palliative care providers and healthcare more generally. Cross-sector partnerships may be needed to provide appropriate advice and support.
LIMITATIONS
This review has a number of limitations. Restriction to studies that explicitly stated the age of their participants may have resulted in relevant articles being missed. Also, more than half of the included studies focused on cancer; given that nonmalignant conditions such as cardiovascular and chronic respiratory diseases are the leading causes of premature death for working age people (Australian Institute of Health and Welfare, 2016; Heron, Reference Heron2016; Office for National Statistics, 2013; Organisation for Economic Co-Operation and Development, 2006), there is scope for further studies on how impacts might vary. Finally, none of the studies considered effects on extended social networks or described interventions to address this population's unmet needs.
CONCLUSIONS
Working age patients facing an advanced progressive disease experience significant underemployment, workforce displacement, financial security, and changes to family dynamics. Further research is needed to better understand the financial and employment concerns of working age patients and their carers, especially as they affect their relationships and roles within the family and extended social networks. Further research is also needed to understand the broader economic implications this has on health policy and the provision of health/palliative care. There is also a need for prospective population-based studies to inform understanding of how impacts may change over time (Burns et al., Reference Burns, Abernethy and Leblanc2011). Finally, evaluative studies are needed of interventions developed to support working age people and their families to overcome the challenges identified in this review. Interventions that explore cross-sectoral partnerships within the context of policy are urgently needed to better address these families' needs.
ACKNOWLEDGMENTS
The research team acknowledge the valuable contribution of Ingrid Amgarth-Duff and Layla Edwards, research assistants (IMPACCT – Improving Palliative, Aged and Chronic Care through Clinical Research and Translation, University of Technology Sydney), for their assistance with the screening of references and data extraction.
CONFLICTS OF INTEREST
The authors declare no conflict of interest.
FUNDING
This research was undertaken, in part, with funding support from AMP Services Limited for consultancy work that was designed to provide insights into the characteristics and needs of working-age people faced with an expected premature death and their families and carers.
SUPPLEMENTARY MATERIAL
The supplementary material for this article can be found at https://doi.org/10.1017/S1478951517000979.
