The economic impact of cancer diagnosis to individuals and their families: a systematic review

Background The effect of a cancer diagnosis is wide-ranging with the potential to affect income, employment and risk of poverty. The aim of this systematic review is to identify the economic impact of a cancer diagnosis for patients and their families/caregivers. Methods The search covered peer-reviewed journals using MEDLINE, EMBASE, CINAHL, Cochrane Library, Epistemonikos and PsycINFO databases. Quality appraisal was undertaken using CASP tools. Monetary values were converted to US Dollars/2019 using a purchasing power parities (PPP) conversion factor. The review included articles up to and including January 2020, written in English language, for patients with cancer aged ≥ 18 years and focused on the costs up to 5 years following a cancer diagnosis. Results The search was run in January 2020 and updated in November 2021. Of the 7973 articles identified, 18 met the inclusion criteria. Studies were undertaken in the USA, Ireland, Canada, Australia, France, UK, Malaysia, Pakistan, China and Sri Lanka. The majority were cohort studies. Twelve reported out-of-pocket costs (range US$16–US$2523/month per patient/caregiver) consisting of medical expenses (e.g. surgery, radiotherapy and chemotherapy) and non-medical expenses (e.g. travel, food and childcare). Fourteen studies reported patient/caregiver loss of income and lost productivity (range 14–57.8%). Conclusions A high percentage of cancer patients and their families/caregivers experience out-of-pocket expenditure, loss of income and lost productivity. Future research is needed to observe the effects of continuing changes to healthcare policies and social protections on the economic burden among cancer patients and their families/caregivers.


Introduction
The economic impact of cancer on individuals and their families/caregivers is a global phenomenon. While technological advances in cancer detection and treatment have improved survival rates [1,2], they are associated with high costs to healthcare systems and patients and their families/caregivers [3][4][5]. Differences in healthcare systems (e.g. publicly/ privately funded) and social support schemes (e.g. whether there is an unemployment compensation during sick leave) can affect the type and amount of economic impact [6]. It is well known that cancer patients and their families/caregivers can experience economic burden, even within a universal healthcare system [7,8].
Research on out-of-pocket (OOP) costs has been conducted in high and low/middle-income economies. A recent review [9] found that in high-income countries with publicly funded healthcare systems, cancer patients and their caregivers faced OOP costs that range from US$15 to US$400 monthly in Canada and US$58 to US$438 monthly in Australia. Altice et al. [10] report that in the USA, patients receiving oncology treatment can experience OOP costs ranging between US$316 and US$741 per month, and these costs were more than 20% of their annual income.
Additionally, data from several studies suggest that cancer patients are often not able to maintain full-time employment, having to reduce working hours or to cease work in some cases [10][11][12][13]. Losing a job after a cancer diagnosis can lead to both short-term (e.g. paying for bills or food) and long-term (e.g. losing a house) economic impact [14,15].
Previous reviews have examined the economic impact of a cancer diagnosis. A recent review found that in publicly funded healthcare, cancer patients and their families/caregivers experience OOP costs ranging from $17 to $506 per month and income loss ranged from 17.6 to 67.3% [16]. In the USA, which has private healthcare systems, cancer survivors were found to incur financial hardship including OOP costs, income loss and lost working days [10]. Almost half of cancer survivors reported financial distress [10]. Another recent review focussed on OOP costs, reporting that cancer patients and caregivers in the USA spent a higher proportion of their income on OOP costs than that seen in high development index countries with publicly funded healthcare [9]. The same review reported that cancer patients and caregivers spent a higher proportion of their income on OOP costs in low-and middle-income countries than in high-income countries [9]. Other reviews focus on the impact on employment status, reporting that both reduced income and change in employment status (e.g. reduction in work hours and retirement) have been associated with a cancer diagnosis [12,14].
While these previous reviews have considered the impact on patients and family/caregivers costs, they have not explored the types of costs and the key cost drivers [9,10,12,14,16]. This review aims to expand the evidence base by identifying the key cost drivers following a cancer diagnosis and the impact on individuals diagnosed with cancer and their families/caregivers. The review focusses on the 5 years following diagnosis given that previous evidence has shown that while costs are high immediately following diagnosis, they are also likely to be considerable 1-5 years after diagnosis [17]. Figure 1 illustrates the hypothesis with regard to the economic impact on cancer diagnosis.

Search strategy
The databases searched were MEDLINE, PsycINFO, CINAHL, EMBASE and Cochrane Library. Our review consists of three topics: (1) cancer, (2) cost and (3) patient/ caregiver. Boolean operators and keywords were used with Medical Subject Headings (MeSH) when available (see Appendix). All search results were imported into EndNote X9 software, which was used to remove duplicates.
MeSH terms and keywords: Papers were excluded if they were editorials, commentaries, discussion or reviews, position papers and abstracts; focusing on the costs relating to pre-diagnosis; including costs accrued to sectors or systems rather than the individual and their family/caregivers; papers in which only total costs were reported (i.e. no breakdown of the components of the costs); and including participants under the age of 18 and childhood cancer survivors (see Table 1).

Screening
Initially, two reviewers screened the titles and abstracts. In the next stage, the full texts of the remaining studies were evaluated by two reviewers according to the pre-specified inclusion and exclusion criteria. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [18] (see Fig. 2).

Data extraction
One reviewer extracted data from eligible papers into a bespoke data extraction form, which was checked for accuracy by a second reviewer. The extracted data included authors, year, country, objectives, cancer type, sample description (i.e. sample size, gender and age), study design, patients and/or caregivers perspective, time since diagnosis, costs timeframe, key findings, OOP cost per month (US$2019) (see Table 2).

Data synthesis and quality
All studies were assessed using the Critical Appraisal Skill Programme (CASP) [19] tool by one reviewer, and checked by a second reviewer (see Table 3). Due to heterogeneity in the included studies, it was not appropriate to use metaanalysis techniques. Instead, a narrative synthesis was performed to synthesize study characteristics and key findings.
In the display of findings in this systematic review, the local currencies in the included studies are converted to US Dollars/2019 when possible and appropriate for OOP and travel costs. To enable comparisons, costs were divided by purchasing power parity (PPP) exchange rate from the World Bank [20] to convert all non-USD costs to USD costs and were transformed to reflect monthly expenditure (annual OOP costs were divided by 12 to obtain a monthly estimate) (see Table 2).

Out-of-pocket costs
There was a great deal of heterogeneity in the way that OOP costs were reported across the studies. The two main categories of OOP costs were medical (e.g. surgery, chemotherapy and medications) and non-medical (e.g. travel for treatments and childcare) expenses (see Fig. 1). In the studies that reported both OOP medical and non-medical costs, the highest average monthly cost ($2,523) was observed in China [37], and the lowest average cost per month ($16) was in Canada [29]. In terms of those studies reporting OOP medical costs separately, the highest average cost per month ($281) was observed in the USA [33], and the lowest monthly cost ($12) was in the UK [32]. The main cost drivers for medical costs were n = 9 treatment and medications [21, 22, 24-26, 29, 31-33]. Of the studies which reported OOP non-medical costs, the highest average cost per month ($66) was in the USA [33], and the lowest average cost per month ($26) was in Pakistan [31]. The type of cancer in both studies was breast cancer. The main cost drivers for non-medical costs were n = 9 travel for treatment [21-24, 29,[31][32][33]37] and n = 6 homecare (e.g. cleaning and gardening) [22, 24-26, 31, 33]. Travel costs related to cancer were either reported within OOP expenses [26,31,32,37] or calculated as separate costs. Mahmood et al. [31] revealed that the average travel cost came to $ 297 per month. Ceilleachair et al. [24] found an average monthly total cost of $166.25. The lowest average cost was observed in the UK, at $8 per month [32].
To best understand the cost burden from OOP, it is useful to assess the proportion of income spent on OOP costs for cancer care, a study [29] found that OOP costs represented a median of 2.3% of a family's annual income. Another study [33] reported that the percentage of OOP costs was about 31% of the monthly income.
OOP costs may disproportionally also affect different groups in society that vary in terms of their socioeconomic, demographic and clinical characteristics. Socioeconomic characteristics associated with OOP cost burden included lower incomes [30,33], diagnosed with cancer while working [24,29], high level of education [29,31]. Demographic characteristics linked to high OOP costs included younger age, [24,25,32] and being of an ethnic minority [33]. Regarding clinical characteristics, being at an advanced cancer stage [24] and receiving adjuvant treatment [23,24,29,33] were associated with OOP cost burden.
In the studies reporting the absolute loss of income, one study [25] reported that loss of income declined over time with a median loss of $5078 (0-6 months) to a median of $1553 (13-18 months). Another study [35] found that there was a 21% loss of mean income among breast cancer survivors within 1 year of diagnosis, from $1404 per month to $1110 per month.
Of the studies that reported a percentage change in income or in employment status, Lauzier et al. [28] revealed that an average 27% of the annual wage of cancer patients was lost over the first 12 months following a diagnosis. Another study showed that cancer patients reported an average 21% loss of their mean income within 1 year of a diagnosis [35]. Also, Vayr et al. [36] stated that 57.8% of cancer patients reported that they were not in work a year post diagnosis, while 42.2% were working and among them, 83.3% returned to work after taking a sick leave. Two studies reported that self-employment was associated with negative economic consequences as a result of not being able to benefit from the social security system [28,34].
The individual characteristics that are either associated with a high percentage loss of income or which affect the employment status of patients and their families/caregivers included low educational status [28,34,35], being of an ethnic minority [35], having an advanced cancer stage [35] and receiving chemotherapy [28].

Cancer caregiving
This review also identified the economic impact on the carers of patients. The types of costs involved in caregiving were OOP costs, cost associated with impact on employment and care time costs. With regard to OOP costs, caregivers face various expenses including travel costs, medication, food and clothes [22,26,29]. In terms of the impact on a caregivers' employment from providing cancer care, an Australian study found that 40% of caregivers needed to be absent from work, 29% experienced income loss as a result of their caregiving and 8.6% had to leave work or close their business [22]. Also, a Canadian study reported that the absenteeism from work for caregivers was 78.5% with a mean wage loss of $1529 after they were compensated for wage loss due to work absence [38]. Regarding care time costs, Hanly et al. [26] found that over the first 12 months, care time costs accounted for 85% of the total costs incurred by caregivers who provided care at any phase of the disease.

Discussion
This review was undertaken to identify the economic impact following a cancer diagnosis for patients and their families/ caregivers and the individual patient characteristics associated with the costs. Among the results, differences in the types of economic impact were found. The main categories of economic impact that contribute to an economic burden to cancer patients and their caregivers were OOP cost and loss of income/lost productivity. The economic impact of a cancer diagnosis sits within the wider context of the structure of the healthcare system (e.g. national health insurance and co-payment systems) and the social welfare system (e.g. short or long-term sick leave and an early or late disability pension). Thus, in this section, we discuss these contextual issues in greater detail to address the heterogeneity in the findings.

Out-of-pocket costs
OOP expenses are the most common costs that cancer patients face [39]. They occur across different health system  Qualitative study models, including where individuals have health insurance coverage [27,33].
Of the 18 studies, fourteen examined OOP costs in countries with universal healthcare [21-26, 28, 29, 32, 34-38]. Comparisons of cancer costs were limited in these studies due to factors, including different cancer types, the stage of disease at diagnosis and whether patients were in active treatment or follow-up care. However, this review has indicated that even in countries that have systems to provide universal healthcare coverage, patients with cancer face OOP cost burden. Moreover, the economic impact on patients and their families/caregivers may still occur in high-income countries with publicly funded healthcare models. This finding is comparable to a recent review that found cancer patients in public healthcare systems experienced increased OOP expenses [39].
Even in countries that provide universal healthcare, individuals who choose to have health insurance can incur additional OOP expenses [40,41]. In Australia -a country with universal health insurance and optional private insurance -cancer patients faced high medical costs and other hidden costs [25]. The influence of the Australian healthcare system on cancer expenses has also been discussed in recent research [42] where cancer patients who had private insurance experienced higher economic burdens than those who relied on government-funded hospitals as a result of high co-payments (e.g. hospital fees).
This review found that in countries with private healthcare systems -USA [33], cancer patients experience higher OOP costs per month than those in low-income countries -Pakistan and Sri Lanka [21,31]. On the other hand, in lowand middle-income countries where patients rely mainly on OOP payments for cancer care, the economic consequences of a cancer diagnosis can be extreme and may form a barrier to accessing cancer care [43,44]. In this review, one study from Pakistan found that medical expenses were the main factor in the total cost of illness for a patient and their family/caregivers, which was, on average, $1262 [31]. This is a significant expense in a country where the average monthly wage is $268.10 [45] and when approximately half of all Pakistanis live in poverty [46].
The OOP costs of cancer found in the studies include expenses beyond the medical costs; these included costs related to home care tasks, such as cleaning and gardening [25,29,31,33], making necessary home modifications for ease of living [23,24] and paying for telephone, electricity and heating bills [23,28,32]. These findings are similar to those from previous research that found that patients with cancer experience additional costs related to household support [47,48] and household bills [49,50]. Transport costs, parking fees, accommodation for overnight stays and meals [23,26,29,31] also created an economic burden for patients and caregivers. Cancer patients may need to travel long √, Yes; X, distances to access care or treatment when health services are located far from a patient's home and or when frequent trips are necessary [23,26,29,31].
In the studies that included OOP costs, individual characteristics were associated with the nature or type of economic impact following a cancer diagnosis such as lower income [29,31,33], diagnosed with cancer while working [24,26], younger age [24,25,32], being of an ethnic minority [33], being at an advanced cancer stage [24] and receiving adjuvant treatment [23,24,29,33]. These findings are in general agreement with those documented in prior reviews [9,51], where individual characteristics, including a lower income, a high level of education, a younger age and an advanced stage of disease were connected with high OOP costs.

Loss of income/lost productivity
There are observed variations regarding the impact of a cancer diagnosis on income loss and lost productivity between countries. The extent of support provided by social security systems is likely a factor here; Lauzier et al. [28] documented that in Canada, working women lost on average 27% of their annual income after receiving compensation (i.e. the system of insurance that reimburses workers in cases of illness). Canadian workers with cancer can make use of different forms of compensation, which can include up to half of their salary. Private employer insurance can cover a higher percentage of their salary compared to government employment insurance. They could also use annual paid leave as salary compensation if they are incapable of performing their normal duties [28]. In addition, this review confirms that self-employment is associated with economic difficulties after a cancer diagnosis [28,34]. Self-employed patients reported limited access to government insurance coverage, resulting in much higher losses [28,34]. These results are consistent with findings from previous studies that found a relationship between self-employment and an increased economic burden [52].
The working regulations of return to work following a cancer diagnosis, which is influenced by social security and health insurance systems (such as paid sick leave), may contribute to the variations in employment and economic hardship between countries [14]. In this review, a study from France, where a social support system grants patients on sick leave, found that one year after a cancer diagnosis, 42.2% of patients were working and among them, 83.3% returned to work after taking a sick leave [36]. However, this study did not mention whether patients on sick leave receive their full salary or only a percentage. Also, in a study from Malaysia, 40.6% of cancer survivors returned to work after their diagnosis [35]. It is worth noting that cancer patients in Malaysia who work in the governmental sector can take up to 2 years of paid sick leave, while those working in private sectors receive only 2 months [35]. The findings of this review are supported by a Dutch study, which found that the proportion of cancer survivors who returned to work decreased as a result of a change in policy in 2004 regarding sickness absence compensation that increased permitted sick leave from one year to two years [53].
The economic impact of a cancer diagnosis on patients and their families/caregivers is associated with some individual characteristics, which result in a high percentage loss of income or which affect a patient's employment status. These characteristics included lower level of education [28,34,35], being of an ethnic minority [35] having an advanced cancer stage [35] and receiving chemotherapy [28]. These results are consistent with previous reviews that have shown that those with a low educational level, who are older, who have advanced cancer or are of an ethnic minority are more likely to experience loss of income and a change in their employment status [12,54].

Cancer caregiving
A number of included studies suggest that caregivers face considerable OOP costs, including travel costs, medication, food and clothing [22,26,29], loss of income/lost productivity [22,30,31,37,38] and care time costs, which range from $110 per month [32] to $2641 per month [26]. These high time costs may explain the negative impact caregiving has on employment and are similar to findings of previous research. A US study found that caregivers experience an average OOP cost of $1243 over a 12-month period, which mainly comprised travel costs [55]. Moreover, another review reported that cancer caregivers are more likely to lose a high percentage of income or may stop working [56].

Strengths
This systematic review has several strengths. Multiple databases were used, and extensive keywords were searched to identify articles related to the economic impact of a cancer diagnosis on individuals and their families/caregivers. The review was not limited to a specific setting and targeted studies that included affected individuals other than cancer patients, including their family, caregivers and spouses.

Limitations
This review has some limitations, which should be noted. In the analysis stage, it was challenging to make comparisons between the included studies as they differed across a number of factors. The studies included several countries with different average ages, different health systems and social support systems and they studied mixed stages of disease or cancer types, different time durations following diagnosis and different costs of items within categories. With such variations, it was impossible to make a direct comparison across studies. Additionally, to convert the costs into a common currency, purchasing power parities (PPP) were used to adjust local currencies in the included studies to the equivalent value of the US dollar as of 2019 [20]. However, the valuation of healthcare may be slightly different from PPP values, which were originally used for the net effect of all goods and services. This issue might influence the cost comparisons between different countries due to the limitations of exact estimates.

Conclusion
This review shows that in the case of a cancer diagnosis, a considerable amount of OOP expenditure is incurred by patients and their families/caregivers and this can cause economic hardship. The impact of loss of income/lost productivity vary, which is likely to be an artefact of differences in social security systems. The economic impact on patients and caregivers varies across countries based on the structure of the healthcare system. Less heterogeneity among studies and increased standardisation of measures would make cost comparisons easier. Future research is needed to observe the effects of continuing changes to healthcare policies and social security systems on the economic burden of a cancer diagnosis.