Estimating the indirect costs associated with adenocarcinoma or squamous cell carcinoma of the oesophagus in Switzerland: evidence from a cross-sectional survey

Abstract

Objectives

Our study investigated the impact of oesophageal cancer and its treatments on work productivity, leisure time, household chores and informal care in Switzerland. We assessed indirect costs for patients treated at different stages of adenocarcinoma or squamous cell carcinoma of the oesophagus.

Design and setting

A retrospective, non-interventional survey was conducted among patients diagnosed with early or advanced oesophageal cancer in Switzerland. Between January 2023 and October 2023, 126 patients participated from seven clinical centres across German, French and Italian language regions.

Outcome measure

Self-reported lost time over a 4-week period, which was monetised to estimate indirect costs using respondents’ approximate earnings derived from national statistics.

Results

Of the 126 patients, 24 (19.0%) were disease-free at the time of the study, while 15.1% of patients were in stage I, 13.5% in stage II, 27.0% in stage III and 25.4% in stage IV. Most patients, across age groups and disease stages, reported no impact on their leisure time (62.2%), household chores (70.3%) or informal care needs (78.4%) in the last 4 weeks. For those affected, the mean loss of time was 9.5 hours per week for leisure (n=28) and 13.3 hours for household chores (n=21). Additionally, patients received an average of 11.8 hours of informal care per week (n=16). Among the patients who were employed at the beginning of the 4-week recall period (n=25), 57.1% reduced the degree of employment and/or missed work due to the disease and its treatments during that time, while 46.7% experienced presentism (decreased productivity at work). The estimated mean indirect costs over a 4-week period was (Swiss Francs) CHF2005 (€1874) per patient, with loss of work productivity being the largest contributor.

Conclusion

Oesophageal cancer in Switzerland affects patients’ professional and personal lives, resulting in lost time and informal caregiving, leading to societal costs.

Correspondence to Dr Sandro T Stoffel; [email protected]

STRENGTHS AND LIMITATIONS OF THIS STUDY

• The study sample consisted of a diverse group of patients diagnosed with early or advanced oesophageal cancer from the French-, German- and Italian-speaking regions of Switzerland.

• To assess lost time and estimate indirect costs, we adapted the Productivity Cost Questionnaire (iPCQ), a simple and validated tool.

• We estimated indirect costs for individual disease stages.

• We were unable to independently verify the self-reported lost times.

Introduction

The economic burden of oesophageal cancer is substantial and is composed of direct medical and non-medical costs, which cover the use of resources inside and outside the healthcare system, and indirect costs, resulting from the loss of resources and opportunities.¹ Indirect costs are the monetary losses associated with the lost times of patients and caregivers and missed opportunities resulting from disease morbidity and mortality.² These costs stem from reduced productivity in paid work, diminished ability to perform household chores and lost leisure time.^1–4 Productivity losses can occur due to premature mortality when patients die during their working age.³ Previous studies using the human capital approach (HCA) have shown that these costs can be substantial.³ In the case of oesophageal cancer, a study estimated it to be (Euros) €175 770 per death in Europe.⁵ Productivity losses can also arise from temporary absence from work due to illness or treatment (sickness absence), reduction of degree of employment or permanent discontinuation in the form of early retirement from work due to permanent incapacity/disability. Presentism, that is, presence at the workplace without being able to provide a normal level of productivity, is another possible, although difficult to measure, component of lost productivity.⁶ The indirect costs of cancer are not just limited to the costs of patients losing productive work in the labour market. Informal care provided by family caregivers/next of kin and patients’ decreased leisure time and time spent on household chores are also elements of indirect costs.^{4 7} As these latter costs are usually not reflected in economic transactions, it is difficult to estimate them through secondary data.⁴ Typically, patient surveys are used to obtain data on the time patients spend in treatment, not working due to short- or long-term illness,⁸ as well as impact of time spent on non-labour market activities, such as household chores and leisure. Care received from informal caregivers also plays a role.⁴

While previous international studies have demonstrated that oesophageal cancer can significantly affect patients’ quality of life and result in substantial indirect costs,^9–12 there remains limited knowledge about these indirect costs in Switzerland.^{1 3} So far, only three studies have estimated the indirect costs of cancer in Switzerland.^13–15

One international study used data from the literature and estimated the indirect cost of oesophageal cancer in Switzerland for 2018 to be (Swiss Francs) CHF57 million.¹³ The other studies estimated the costs for all cancers, regardless of the cancer stage, type and demographic variables of the patient, and did not account for lost leisure time and household chores.^{14 15} Their estimated costs range from CHF2.2 billion to CHF5.8 billion per year, depending on whether only productivity losses due to mortality and morbidity are included or whether also costs for productivity losses arising from informal care are included.^{13 14} The large range suggests that there is a lot of uncertainty around these estimates.

Registry data shows that on average 600 new cases of oesophageal cancers (all subtypes included) are diagnosed in Switzerland per year, predominantly in men (n=450, 75.0%).^15–18 The 5-year prevalence (2015–2020) is estimated to be 865 individuals,¹⁹ and the median age at diagnosis is 70 years for men and 72 years for women.²⁰

The present study intended to reduce the information gap on the societal burden of oesophageal cancer in Switzerland. We specifically aimed to estimate the lost times, productivity losses and indirect costs over 4 weeks associated with local disease (stage I), locoregional disease (stage II or stage III) and metastatic disease (stage IV) among Swiss patients with adenocarcinoma or squamous cell carcinoma of the oesophagus.

Methods

General approach

The study collected patient-reported data on lost time through a cross-sectional survey to explore the direct experiences of individuals affected by oesophageal cancer.^{21 22} We linked the responses regarding lost productivity, leisure time and informal care with clinical data provided by the enrolling clinical centres and then estimated indirect costs over a 4-week period for individual disease stages by multiplying the total time by appropriate estimates of the responder’s earnings.^{22 23}

Study population

The study population consisted of male and female patients aged 18 years or older, residing in Switzerland. These patients had previously sought consultation or treatment at one of the participating clinical centres. Additionally, they had been diagnosed with adenocarcinoma or squamous cell carcinoma of the oesophagus, or adenocarcinoma of the gastro-oesophageal junction, at least 3 months prior to their inclusion in this study. The reason for this minimum time since diagnosis was to ensure that the patients had enough time to experience the consequences of the diagnosis in their daily life. As one of our goals was to compare the monthly indirect costs between patients who did receive and those who did not receive any treatment modality for stage I or stage II–III oesophageal cancer in addition to tumour resection, patients, initially diagnosed in these stages, must have undergone a resection (surgical or endoscopic). Additionally, some of them may have received any form of neoadjuvant and or adjuvant therapy. Patients with an initial diagnosis of metastatic disease (stage IV) were required to have received at least one line of systemic therapy, consistent with clinical guidelines for the management of advanced oesophageal cancer²⁴ or best supportive care.

Recruitment and study design

Patients were recruited between January 2023 and October 2023 through seven participating clinical centres, five in the German-speaking part of Switzerland and one each in the French-speaking and Italian-speaking parts. Three centres were public cantonal hospitals, the other two centres were a university and private hospital, respectively. Healthcare professionals in the clinics identified all eligible study participants from their medical records. Each eligible patient was given a unique, anonymous identification number (ID) by the treating centre. Subsequently, the centres sent these eligible patients a letter containing a study information sheet, the questionnaire with the participant’s ID and a prestamped return envelope. The information sheet explained the purpose of the study, highlighted the voluntary nature of participation and emphasised the anonymity of responses and data transfer from the clinical centre to the study team. All participants were required to provide explicit consent by ticking checkboxes on the informed consent form. Centres that chose to distribute study invitations and questionnaires during routine medical visits were instructed to provide patients with an envelope containing the invitation letter, questionnaire and return envelope, asking them to complete it at home. The return envelope was addressed to the study team, which then communicated the study participant IDs to the clinical centres. Reminder letters with a replacement questionnaire and return envelope were sent to non-responders 1 month later. Participants were allowed to complete the survey only once. We ensured this by maintaining a participant list, checking off participants’ IDs on survey submission, and distributing surveys in a controlled environment for close monitoring. For each consenting study participant, the clinical centres shared information including participant ID, birth year and sex, clinical characteristics (including date of diagnosis, stage at diagnosis, current stage, cancer type and histologic type) and type of treatment (such as resection (surgical or endoscopic), neoadjuvant therapy (radiochemotherapy or chemotherapy), radiochemotherapy (as primary treatment), adjuvant or definitive chemotherapy, adjuvant or definitive immunotherapy and best supportive therapy) with the study team. The clinical data were merged with the questionnaire responses using the participants’ IDs. No monetary incentive was offered to participants for completing the questionnaire. To evaluate non-response bias, six of the seven clinical centres shared aggregated data on the age category, sex and cancer stage at diagnosis for all eligible patients. We then inferred the characteristics of non-participants based on the overall sample and the participants.

Patient survey

The survey was based on the Productivity Cost Questionnaire (iPCQ), which was developed by the Institute for Medical Technology Assessment (iMTA), Erasmus University, Rotterdam.^{8 25} The iPCQ is a standardised instrument for measuring health-related productivity losses and consists of 18 questions. Nine questions assess demographic information and respondents’ work status (eg, number of hours of paid work per week and number of working days per week) and productivity loss in the last 4 weeks.²⁵ For the current study, we adapted the existing questions to the study setting and added additional questions to assess presentism (ie, days at work with reduced productivity) due to oesophageal cancer, loss of leisure time, household chores and informal care provided by the next of kin below and above the age of retirement. Overall, the questionnaire captured reduced degree of employment, missed days at work, presentism (where a 50% reduction in regular output from the worker was assumed for such days), informal care provided by all individuals as well as only by working-age individuals, lost leisure time and time spent doing household chores. The modified iPCQ was reviewed by the original developers at iMTA to ensure its integrity.

Patients who were diagnosed at least 6 months before they received the questionnaire were asked to answer additional questions about the impact of the disease and treatment on their lost time in the first 3 months after the diagnosis, covering the same elements as described above.

The survey was offered in French, German and Italian.

Patient and public involvement

We engaged cancer patients in formulating the research question and in the development and testing of the questionnaire. Furthermore, the readability and accuracy of the questionnaire were cross-checked using a think-aloud methodology with four cancer patients who had prior experience as patient advisors on clinical studies, but who were not involved in the development of this particular study. The pilot results confirmed that the questions were easy to understand and that the survey was user-friendly and effective in collecting reliable data.

Indirect costs

For ease of comparison with international studies, cost results are reported in both € and CHF, using an average exchange rate of €1.0=CHF1.07 over the period 2018–2023. We used HCA, which implies a societal perspective and considers as lost all the hours that the patient and informal caregiver were not able to work because of the cancer and its treatments.¹³ To calculate the indirect costs of lost productivity and leisure time over a 4-week period, we multiplied the reported times lost by the approximate earnings of the respondent, which were estimated from national statistics.²² Similarly, for the cost of informal care, we used the proxy good method, which calculates the monetary value of an informal carer’s time, based on the wages a professional carer or suitable substitute (such as, eg, a cleaner) would earn for providing the same care.^{22 23} The method allows to assign monetary amounts to services for which no money is exchanged. Specifically, we used average salaries reported by the Swiss Federal Statistical Office (SFSO) for the year 2020, stratified by age and gender.²⁶ The average monthly salaries provided by the SFSO correspond to a 100% position. To break them down to an hourly level, we divided them by 173 (since the ‘standardised monthly wage from the SFSO represents a full-time equivalent based on 40 hours per week and 4 1/3 weeks per month). For patients who were employed at the time of the survey, we valued their work time at the same rate as the average salary of a worker in Switzerland of matching age and gender. For patients who were not working due to oesophageal cancer, we valued their lost work time based on the average 4-week salary of a Swiss worker of the same age and gender. We adjusted this value based on the average patient’s part-time work, estimated at 93% of a full-time equivalent. Lost leisure and household chores time were valued based on a Dutch study, which estimated this value at €16 per hour in 2014.²⁷ We converted this to a Swiss value of CHF27 (€25.2) per hour in 2022, adjusting for the consumer price index in the Netherlands²⁸ and accounting for purchasing power parities between the Netherlands and Switzerland.²⁹ For the informal care, time was valued at CHF38 (€35.5) per hour, which approximates the rate for home care (Spitex), that is, the professional household and support services.^{30 31}

Sample size

Due to the descriptive nature of the study, no formal sample size calculation was performed beforehand. Based on feasibility considerations, we aimed for a sample of at least 120 oesophageal cancer patients, which is similar to previous studies that focused on indirect costs in individual cancer types.^{32 33}

Statistical analysis

The primary outcome was the monetisation over a 4-week period of the patients’ productivity losses due to morbidity, their reduced leisure time, the time lost on household chores due to morbidity and informal care provided by caregivers. To estimate the indirect costs, we aggregated the time lost in each cost category, monetized it and summed up the costs across the categories. Due to the risk of double counting resulting from asking patients two separate questions—whether the cancer and its treatment influenced their work in the last 4 weeks and whether they missed days at work because of the cancer—we adjusted the answers to the second question for some patients. Specifically, we converted the responses to the second question to a zero value for six participants. Each of these participants had reported an 80% or more reduction in their work time in the previous question and, in addition, each reported 17 or more missed work days in the second question. Furthermore, qualitative responses to questions that were originally intended to elicit a quantitative response were treated as missing values. For example, 11 participants indicated that cancer and treatment impacted their work, leisure time, household chores or informal care needs. However, they did not provide specific numeric values for hours or days lost in any cost category. Consequently, the estimated indirect costs for these patients were zero. We conducted a sensitivity analysis excluding these participants and reported the results in the supplementary files.

Descriptive statistics were employed to summarise the reported impact of oesophageal cancer and its treatment in terms of patients’ and informal caregivers’ lost time and estimated indirect costs, both overall and by disease stage. The study sample’s characteristics were described using counts (percentages) for categorical variables. We assessed the risk of non-response bias by comparing the age, gender and cancer stage at diagnosis among those who participated and those who did not.³⁴ For the univariate analyses, we used χ² tests of independence and Fisher’s exact tests for the categorical variables and Mann-Whitney U tests and Kruskal-Wallis tests for continuous variables. To assess the association of patient, disease and treatment characteristics with indirect costs, we used the generalised linear model (GLM) regressions with a log link function and gamma distribution.³⁵

Associations that were statistically significant in the unadjusted regression model (ie, p<0.05) were subsequently included in an adjusted regression model to assess the robustness of these associations. We excluded patients with missing values on a variable-by-variable basis and did not impute any missing data. Responses to questions intended to elicit a quantitative response but instead resulted in a qualitative response were treated as missing data in the quantitative analyses. Data were analysed using Stata/IC V.16.0 (StataCorp LP, College Station, Texas).

Results

Study sample

A total of 254 questionnaires were distributed among eligible cancer patients. Of these, 126 were completed and returned to the study team, resulting in a response rate of 49.6%. A comparison of age, gender and cancer stage at diagnosis between participants and non-participants did not indicate selection effects (see online supplemental table S1).

Table 1 and online supplemental table S2 summarise the characteristics of the study participants. Looking at the overall sample, which also includes patients currently disease-free, the majority were male (77.8%, n=98) and aged 64 years or older (73.8%, n=93). The mean age of participants was 70.2 years with an SD of 9.5. In terms of education, 33.3% had completed compulsory school, 42.9% (n=54) had upper secondary school education and 23.8% (n=30) had tertiary education. Most participants were initially diagnosed with stage III (42.1%) or II (22.2%). All patients were diagnosed between 2016 and 2023, and the majority received their diagnoses between 6 months and 1 year before the survey (34.1%, n=43) or more than 2 years prior to the survey (27.8%, n=35). The characteristics of the subsample of patients in stages I to IV (ie, not disease-free) closely resembled those of the overall sample. Most of these patients had upper secondary school education (42.1%, n=43), were male (79.4%, n=81) and were aged 64 or older (72.6%, n=74), with a mean age of 70.1 years (SD 9.1). The characteristics of the study patients and their treatments, broken down into the individual disease stages, are shown in online supplemental tables S3 and S4 in the supplementary file.

Table 1

Demographic and clinical characteristics of patients with oesophageal cancer

Variable	Overall sample (n=126)		Sample with active disease only (n=102)
	N	(%)	N	(%)
Age (mean (SD))	70.2	(9.5)	70.1	(9.1)
Age category
Below 60	18	14.3	14	13.7
Between 60 and 64	15	11.9	14	13.7
Older than 64	93	73.8	74	72.6
Sex
Male	98	77.8	81	79.4
Female	28	22.2	21	20.6
Education
Compulsory school	42	33.3	37	36.3
Upper secondary school education	54	42.9	43	42.1
Tertiary education	30	23.8	22	21.6
Employment status
Employee	21	16.7	17	16.7
Self-employed	3	2.4	3	2.9
Unable to work	8	6.4	8	7.8
Retired	88	69.8	70	68.6
Home/house	3	2.4	3	2.9
Other	3	2.4	1	1.0
Language
German	88	69.9	83	81.4
French	27	21.4	10	9.8
Italian	11	8.7	9	8.8
Cancer type
Gastro-oesophageal junction (GEJ)	65	51.6	54	52.9
Oesophageal cancer (E)	61	48.4	48	47.1
Histologic type
Squamous cell carcinoma	33	26.2	23	22.6
Adenocarcinoma	93	73.8	79	77.4
Cancer stage at initial diagnosis
Stage I	23	18.2	12	11.7
Stage II	28	22.2	22	21.6
Stage III	53	42.1	46	45.1
Stage IV	22	17.5	22	21.6
Current cancer stage at the time of the survey
Disease-free	24	19.0	–
Stage I	19	15.1	19	18.6
Stage II	17	13.5	17	16.7
Stage III	34	27.0	34	33.3
Stage IV	32	25.4	32	31.4
Time since initial diagnosis
Between 3 and 6 months ago	16	12.7	12	11.7
Between 6 months and 1 year ago	43	34.1	37	36.3
Between 1 and 2 years ago	32	25.4	26	25.5
More than 2 years ago	35	27.8	27	26.5

Note: active disease is defined as cancer still being present (stages I–IV) at the time of the study.

Most patients were initially diagnosed at stages II and III, accounting for 64.3% of the overall sample (66.7% when excluding currently disease-free patients; table 1). The predominant histological type observed was adenocarcinoma, representing 73.8% of the overall cases (77.4% when excluding currently disease-free patients). Online supplemental table S2 shows that resection, either surgical or endoscopic, was a common treatment, performed in 74.6% of all patients (73.5% when excluding currently disease-free patients). Additionally, neoadjuvant therapy was highly prevalent, with 69.8% of all patients receiving it (67.7% when excluding currently disease-free patients).

Adjuvant therapy was given to 36.5% (n=46) of the study sample. Among these, 54.3% (n=25) received immunotherapy and 45.7% (n=21) were treated with chemotherapy. Palliative chemotherapy, palliative immunotherapy and best supportive care were exclusively administered to stage IV cancer patients (online supplemental table S4).

Patient lost time

Table 2 shows the patients’ lost time of work, leisure and household chores and caregiver’s lost time spent on informal care due to patient morbidity per month. In the overall sample, the majority of patients reported no impact of the disease and treatment on productivity (75.4%, n=95), leisure time (63.5%, n=80), household chores (71.4%, n=90) or informal care needs (80.9%, n=102, see online supplemental figure S1). Overall, 51.6% (n=65) indicated that the cancer and treatment did not impact any of these aspects.

Table 2

Descriptive statistics relating to time estimates for oesophageal cancer patients

Variable	Overall sample (n=126)		Sample with active disease only (n=102)
	N/N	(%)	N/N	(%)
In paid work at the time of the study	30/126	(23.8)	25/102	(24.5)
Retired from work in last 28 days due to cancer	7/96	(7.3%)	6/77	(7.8%)
Partially reduced working hours in last 28 days because of cancer	14/30	(46.7)	11/25	(44.0)
Reduction of work in percentage (mean, SD)	69.6%	37.6%	75.0%	35.3%
Missed work days in last 28 days because of cancer	16/30	(53.3)	13/25	(52.0)
Number of days lost in last 28 days (mean, SD)	3.4	7.8	4.0	8.8
Lost productivity at work in last 28 days because of cancer	14/30	(46.7)	12/25	(48.0)
Number of productive days lost in last 28 days (mean, SD)	8.8	8.3	9.6	8.4
Loss of leisure time in last 28 days because of cancer	46/126	(36.5)	41/102	(40.2)
Number of hours lost per week (mean, SD)	15.7	17.3	16.7	17.9
Loss of household chore time in last 28 days because of cancer	36/126	(28.6)	32/102	(31.4)
Number of hours lost per week (mean, SD)	12.9	13.1	13.8	13.8
Received informal care in last 28 days because of cancer	24/126	(19.1)	20/102	(19.6)
Number of hours received per week (mean, SD)	15.8	15.0	18.2	16.1
Number of hours provided per week by informal caregiver below the retirement age (mean, SD)	6.9	12.8	9.3	14.3

Note: active disease is defined as cancer still being present (stages I–IV) at the time of the study. Descriptive statistics of the lost time estimates broken according to disease stage are shown in online supplemental table S5 in the supplementary file. Self-reported lost productivity at work and lost leisure time, household chore and informal care are also displayed in online supplemental figure S2.

Results were similar for the subsample excluding currently disease-free patients. There, 52.0% (n=53) reported no impact. In the overall sample, the mean loss of time for those affected was 15.7 hours per week for leisure (n=46) and 12.9 hours for household chores (n=36). The mean weekly received informal care in those with any was 15.8 hours (n=24).

In terms of impact on work, seven out of 96 (7.3%) retired patients indicated that they retired in the last 28 days due to oesophageal cancer and its treatments. All 30 (100%) study participants who were employed at the time of the study reported an impact of cancer and its treatment on their work in the last 28 days. A majority (53.3%, n=16) stated that they had missed days at work, while 46.7% (n=14) indicated that they had to reduce their working hours or were less productive. On average, these patients reduced their working hours by 69.6% (n=13), missed 3.4 workdays (n=14) or were less productive for 8.8 days (n=10).

Excluding the disease-free, patients showed a similar impact of cancer on leisure time, household chores and informal care needs across disease stages. The impact of cancer and its treatment on patients’ productivity, leisure, household chores and informal care at different disease stages is described in online supplemental table S4 in the supplementary files. There is no indication of statistically significant variation in impact across the disease stages.

Indirect cost

Table 3 specifies the monetary value of patients’ lost productivity, leisure and household chores due to morbidity, and the monetary value of caregiver time spent for informal care, over a 4-week period. Figure 1 reveals a U-shaped distribution of the estimated indirect costs related to oesophageal cancer during this timeframe (see online supplemental figures S3 and S4 for distributions according to age and gender).

Figure 1. Estimated total indirect cost over a 4-week period in CHF (n=126). Note: to enable comparison with international studies, the costs can be converted into Euros using an exchange rate of €1.0=CHF1.07.

Table 3

Estimated indirect costs of oesophageal cancer over a 4-week period for the overall sample (n=126)

Item	Overall sample (n=126)			Sample with active disease only (n=102)
	n	Cost in CHF		n	Cost in CHF
		Mean	SD		Mean	SD
Productivity	120	1017	2459	102	1098	2592
Informal care	120	360	1219	96	404	1336
Leisure time	110	461	1223	88	552	1347
Household chores	112	273	830	89	318	910
All of the above		2005	3945	102	2231	4210

Note: to enable comparison with international studies, the costs can be converted into Euros using an exchange rate of €1.0=CHF1.07. Active disease is defined as cancer still being present (stage I–IV) at the time of the study. Online supplemental table S8 in the supplementary files shows the complete case analysis of the estimated indirect costs, which are slightly higher than those reported above.

Most study participants (60.3%, n=76) had zero estimated indirect costs, while 18.3% (n=23) incurred costs exceeding CHF5000 (€4673). The average indirect costs were estimated at CHF2005 (€1874) per patient and at CHF2231 (€2085) when excluding currently disease-free patients (see distribution of indirect costs in figure 2). The most substantial cost item was due to the loss of productivity by the patients (CHF1017 (€950) and CHF1098 (€1026), respectively).

Figure 2. Estimated costs for each cost category over a 4-week period in CHF. Note: to enable comparison with international studies, the costs can be converted into Euros using an exchange rate of €1.0=CHF1.07.

Online supplemental table S6 and figure S5 in the supplementary file provide details on the mean and median indirect costs of oesophageal cancer over a 4-week period, categorised by the current cancer stage. The estimated mean costs were lowest for stage I (CHF1602 (€1497) per patient) and for patients who were disease-free at the time of the study (CHF1040 (€972)). Conversely, the costs were highest for stage IV (CHF2486 (€2323) per patient). However, the results of the Kruskal-Wallis tests did not suggest that these differences were statistically significant (χ²=0.442, p=0.979, df=4). While excluding patients, for whom the costs were zero, reveals a similar pattern (χ²=5.452, p=0.244, df=4). The results of Mann-Whitney U tests did not suggest differences in the median indirect costs between patients in stage I and stage III with and without neoadjuvant therapy in the overall sample (median=CHF0 in both cases, z=−1.556, p=0.121) or the subsample of patients with any indirect costs (median with adjuvant therapy: CHF5291 (€4945) vs median without adjuvant therapy CHF2364 (€2209), z=0.831, p=0.413) to be statistically significant. Indirect costs stratified by adjuvant therapy are shown in online supplemental table S7 and figure S6 in the supplementary file.

Online supplemental table S9 presents the results of the GLM regressions, with the indirect costs of oesophageal cancer over a 4-week period as the dependent variable. Unadjusted regressions indicated that paid employment and having undergone resection or neoadjuvant therapy were associated with higher indirect costs. Conversely, advanced age and having received radiochemotherapy were linked to lower indirect costs. In the subgroup of patients with active disease (ie, cancer still present at the time of the survey), receiving cancer treatment was also associated with reduced indirect costs. However, in the adjusted regression model that incorporated all these variables, only the associations with age and radiochemotherapy remained statistically significant.

The results of additional GLM regression models, with the informal care, lost leisure and household chores time costs due to oesophageal cancer over a 4-week period as the dependent variables, are presented in online supplemental tables S10–S12 in the supplementary file. In the adjusted regressions, higher education was linked to increased costs of informal care. Additionally, a more advanced current cancer stage was associated with higher costs of lost leisure time.

Impact in first 3 months after the diagnosis

A total of 103 patients were diagnosed with oesophageal cancer more than 6 months before receiving the questionnaire. The description of their characteristics is provided in online supplemental table S13 in the supplementary file.

Online supplemental table S14 illustrates that among this specific subgroup, 50.0% of the 14 patients who were employed at the time of diagnosis opted to reduce their working hours, while an additional 57.1% reported missing work days. Moreover, 50.0% of these patients experienced decreased productivity during the initial 3 months postdiagnosis. Furthermore, 56.3% of participants attributed a reduction in leisure time to oesophageal cancer, with 41.8% reporting diminished time for household chores and 34.0% receiving informal care (see online supplemental figure S7).

The impact of cancer on the first 3 months after diagnosis for each initial disease stage is presented in online supplemental table S14. On average, those affected lost 19.6 hours per week for leisure (n=58) and 21.9 hours for household chores (n=43). The mean informal care received per week stood at 18.7 hours (n=35), primarily provided by caregivers below retirement age, 11.6 hours on average. Comparing the impact of cancer within the first 3 months postdiagnosis to the impact in the 28 days prior to the survey’s time point is challenging due to the differing time frames. Nonetheless, a larger proportion of patients lost leisure time and required informal care in the initial 3 months.

Discussion

Our study is one of the first to provide detailed information on the indirect costs of oesophageal cancer in Switzerland. Other studies have estimated the indirect costs of cancer in Switzerland^13–15 but have not focused on oesophageal cancer or have adopted macro-costing approaches, in contrast to our study, which adopted a micro-costing approach through a survey. In line with these studies, we found that oesophageal cancer impacts patients’ lives and for those affected the economic impact is substantial. Specifically, we observed a U-shaped distribution of estimated indirect costs, with most study participants either reporting no indirect costs or incurring rather costs resulting in median indirect costs of zero. We estimated that the mean indirect costs of oesophageal cancer over a 4-week period in Switzerland is CHF2005 (€1874) per patient. This amount is lower than what a previous literature-based study estimated,¹³ but it nearly equals the full Swiss retirement pension (German: AHV), which ranges between CHF1195 (€1117, minimum full pension) and CHF2390 (€2234, maximum full pension) per month.³⁶ The estimates from this previous international study were crude as it considered all digestive cancers and excluded individuals below the age of 50. Additionally, the study imputed missing data for Switzerland using information from Austria, and incidence data were employed to allocate overall costs by cancer type in each country. Our study also differed from these previous studies in that we estimated the indirect costs not only for the overall population but also for individual disease stages and adjuvant treatment. We found that estimated costs were comparable across stages II to IV, and disease-free patients also reported negative impacts of cancer and its treatment on their lives.

The most significant cost driver in our study was the loss of the patient’s productivity, responsible for almost half of the overall indirect cost of oesophageal cancer over a 4-week period. All study participants who were employed at the time of the study indicated that they had either reduced work due to the disease, experienced decreased productivity or missed days at work. Interestingly, across various age groups and disease stages, most study patients reported no impact on their leisure time, household chores or informal care needs in the last 4 weeks. However, those who reported being affected experienced a significant negative impact and a need for informal care per week.

The regression analyses indicated that advanced age and having received radiochemotherapy were associated with lower indirect costs, while a more advanced current cancer stage was linked to higher costs of lost leisure time. Previous studies of the indirect cost of cancer in Switzerland found that informal care costs were approximately equivalent to the indirect costs associated with work-related absences and reductions resulting from morbidity (ie, they observed an approximate 1:1 ratio).^{14 15} Our study in oesophageal cancer found a 1:2 ratio for informal care costs relative to indirect costs associated with work-related absences and reductions. The fact that we estimated a lower proportion of informal care costs may in part be related to the fact that we collected data on informal care hours from patients rather than their informal caregivers and selection bias, as mentioned in the limitation section below. It has previously been found that patients may under-report the number of hours they receive informal care for, compared with the number of hours that their informal caregivers report.³⁷

Furthermore, while indirect costs include lost leisure time, household chores and informal care—in addition to lost productivity—studies typically prioritise the latter aspect.^{4 38} This is because the lost opportunities beyond the labour market are usually not directly quantified in monetary terms, necessitating an approximation of the value of the time lost.^{4 39} In our study, we approximated it with the help of a validated patient questionnaire.⁸

Our study results can serve as inputs for health economic evaluation studies in oesophageal cancer in Switzerland from a societal perspective.⁴⁰ Such an evaluation would require stage-specific outcomes, data on the duration that patients spend in different disease stages, and data on disease progression.³⁹ Although our study did not cover the latter information, it could be derived from epidemiologic and clinical studies. Furthermore, combining such data with information on premature mortality could allow the estimation of lifetime indirect costs associated with oesophageal cancer in future research.³⁵ However, a complete picture of indirect cost associated with oesophageal cancer at different disease stages would be of added value to the society. Accurately measuring and addressing these costs could lead to better healthcare policies, more effective resource allocation and improved support for patients and caregivers. While we regard our results based on a 4-week period as informative, we believe that a prospective study could more effectively address the disease’s impact on lost time and productivity, and better estimate long-term indirect costs associated with oesophageal cancer.

We should acknowledge the shift in the distribution of disease stages at diagnosis in our sample, which does not correlate with published epidemiological data.⁴¹ This observation might be explained by a selection bias driven by the willingness and ability of patients to take the survey. Those very sick or weak may be under-represented; such patients diagnosed with stage IV of the disease may be under-represented due to their shorter course of disease prior to death.

Strengths and limitations

The main strength of this study lies in its bottom-up approach and in the adaptation and the use of the questionnaire from the validated iPCQ questionnaire, which is relatively simple and short to complete.⁸ Consequently, most of the study participants answered the quantitative questions correctly. The response rate of 49.6% to our survey was slightly higher than in previous studies on the indirect costs of cancer,³³ and we did not find obvious hints of selection effects. Still, we cannot exclude the presence of non-response bias, limiting the representativeness of the results. We then combined the results of the survey with clinical data to estimate indirect costs not only for the overall population but also for the different disease stages and treatments.

Our study has several limitations which call for follow-up research. First, although our study covers the entire country geographically, selection effects arise from the inclusion criteria. These criteria require patients to be diagnosed at least 3 months before the survey. Additionally, study patients were recruited from major clinical centres, rather than through office-based oncologists or palliative care institutions. Compared with previously published epidemiological data, where stages III and IV were most common at diagnosis,⁴⁰ our sample under-represented patients initially diagnosed with stage IV of the disease. Additionally, the poorer health state of patients diagnosed with stage IV could have contributed to lower willingness and ability to participate in the survey. Moreover, given the imbalances across disease stages, the sample size may not have been sufficiently large to facilitate a comparison of estimated costs across different disease stages. Additionally, although we assessed non-response bias in our study, we only had data from six out of the seven centres and could only consider gender, age and initial disease stage, so we cannot completely rule it out. Furthermore, at the time of the study, patients had spent varying amounts of time in their disease stage. It is possible that costs did not occur consistently over time while a patient remained in a specific state. Although we believe our sampling approach resulted in valid average results, we could not directly control for this variability. Furthermore, the time frame for assessing indirect costs in our study was too short to fully convey their magnitude. A prospective study would be more appropriate to capture long-term indirect costs that we could not address in the current study.

A further limitation of the study was the inability to verify self-reported lost time. There is a possibility that participants may have encountered difficulties recalling the exact amount of time lost in the last 4 weeks; however, we adhered to the timeframe specified by the iPCQ.⁸ Some patients gave qualitative answers to quantitative questions or indicated that they were affected by cancer but did not provide a response to the subsequent quantitative question, resulting in missing data. However, the vast majority of responses that we obtained were appropriate, suggesting that the survey was generally well understood by the participants. Furthermore, due to the absence of an equivalent Swiss study, we valued leisure and household chores time according to a willingness-to-accept valuation study from the Netherlands.²⁵ Another approach that we could have adopted, but would be unlikely to have changed our results, would have been to use slightly lower values for leisure and household chores time by using the minimum wage in Switzerland.³³ Similarly, we used the proxy good instead of the opportunity cost method for informal care. The latter values the caregiver’s lost time in terms of leisure time forgone, setting the unit costs equal to the value of 1 hour of leisure time.²³ This approach would have set the cost per hour to CHF27 instead of CHF38, and would have reduced the estimated cost of informal care by approximately 30%, resulting in CHF256 (€239) for the overall sample and CHF287 (€268) for the subsample of patients with active disease. Finally, although the questions used to assess the lost time in the first 3 months after diagnosis were based on the iPCQ, they have not been validated for this time frame. Consequently, the prolonged recall period may compromise the accuracy and reliability of the data collected.

Conclusion

In conclusion, our study found that in Switzerland, oesophageal cancer impacts the professional and personal activities of patients through ‘lost time’ and generates informal care provision. This generates relevant indirect costs to the society.

We thank PD Dr Sara De Dosso from the Oncology Institute of Southern Switzerland, Ente Ospedaliero Cantonale, Bellinzona for providing input on the study design. We thank Nicola Miglino, Christian Gutschow and Ralph Fritsch for data capture at the University Hospital Zurich.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants, and ethics approval for the study protocol and subject informed consent were sought from the Ethikkommission Nordwest- und Zentralschweiz (EKNZ). This local ethics committee determined that no ethics approval was required, as the study was purely observational and solely focused on collecting time data from patients through a questionnaire. However, a voluntary ethics assessment with a positive result was obtained from EKNZ (project ID: AO_2022-00061) on 30 November 2022. Participants gave informed consent to participate in the study before taking part.

Footnote

Contributors All authors confirm that they have made substantial contributions to the work, including involvement in conceptualisation, data collection, analysis, writing, editing and manuscript revision. Conceptualisation: STS, PS, MS, AB and CO-N; data collection: DK, TK, AS, LD, MP, TK and AW; data curation: AB and STS; formal analysis, AB and STS; interpretation of results: all authors; investigation: STS, PS, AB, AG and CO-N; methodology: STS, PS, AB and CO-N; supervision: MS and CO-N; writing—original draft: STS; writing—review and editing: STS, PS, MS and CO-N; final approval: all authors. STS acted as a guarantor.

Funding This project is funded by Bristol Myers Squibb. The views expressed are those of the author(s) and not necessarily those of the sponsor.

Competing interests STS, AB, PS and MS report research funding (via employment institution) from Bristol Myers Squibb to conduct the study and write this manuscript. CO-N is employed by and has stock ownership in Bristol Myers Squibb. All other authors declare no competing interests.

Patient and public involvement Patients and/or the public were involved in the design, conduct, reporting or dissemination plans of this research. Refer to the Methods section for further details.

Provenance and peer review Not commissioned; externally peer reviewed.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

References

¹ Hofmarcher T, Brådvik G, Svedman C, et al. Comparator report on cancer in Europe 2019–disease burden, costs and access to medicines. Lund: The Swedish Institute for Health Economics (IHE), 2019.

² Yabroff KR, Lund J, Kepka D, et al. Economic Burden of Cancer in the United States: Estimates, Projections, and Future Research. Cancer Epidemiol Biomarkers Prev 2011; 20: 2006–14. doi:10.1158/1055-9965.EPI-11-0650

³ de Oliveira C, Bremner KE, Pataky R, et al. Understanding the costs of cancer care before and after diagnosis for the 21 most common cancers in Ontario: a population-based descriptive study. CMAJ Open 2013; 1: E1–8. doi:10.9778/cmajo.20120013

⁴ Iragorri N, de Oliveira C, Fitzgerald N, et al. The Indirect Cost Burden of Cancer Care in Canada: A Systematic Literature Review. Appl Health Econ Health Policy 2021; 19: 325–41. doi:10.1007/s40258-020-00619-z

⁵ Hanly P, Ortega-Ortega M, Soerjomataram I. Cancer Premature Mortality Costs in Europe in 2020: A Comparison of the Human Capital Approach and the Friction Cost Approach. Curr Oncol 2022; 29: 3552–64. doi:10.3390/curroncol29050287

⁶ Schultz AB, Chen CY, Edington DW. The cost and impact of health conditions on presenteeism to employers: a review of the literature. Pharmacoeconomics 2009; 27: 365–78. doi:10.2165/00019053-200927050-00002

⁷ Andersson A, Levin L-A, Emtinger BG. The economic burden of informal care. Int J Technol Assess Health Care 2002; 18: 46–54.

⁸ Bouwmans C, Krol M, Severens H, et al. The iMTA Productivity Cost Questionnaire: A Standardized Instrument for Measuring and Valuing Health-Related Productivity Losses. V Health 2015; 18: 753–8. doi:10.1016/j.jval.2015.05.009

⁹ Li Y, Xu J, Gu Y, et al. The Disease and Economic Burdens of Esophageal Cancer in China from 2013 to 2030: Dynamic Cohort Modeling Study. JMIR Public Health Surveill 2022; 8: e33191. doi:10.2196/33191

¹⁰ Daroudi R, Nahvijou A, Arab M, et al. The economic burden of esophageal cancer in Iran. Indian J Cancer 2022; 59: 499–506. doi:10.4103/ijc.IJC_1009_19

¹¹ Xiao H, Bertwistle D, Khela K, et al. Patient and caregiver socioeconomic burden of first-line systemic therapy for advanced gastroesophageal adenocarcinoma. Future Oncol 2022; 18: 1199–210. doi:10.2217/fon-2021-1449

¹² Xiao H, Bertwistle D, Khela K, et al. Global retrospective analysis of clinician- and patient-reported clinical characteristics and humanistic burden of patients with gastroesophageal cancers on first-line treatment. BMC Cancer 2023; 23: 186. doi:10.1186/s12885-023-10553-7

¹³ Hofmarcher T, Lindgren P. The cost of cancers of the digestive system in Europe. IHE Report. 2020.

¹⁴ Hofmarcher T, Lindgren P, Wilking N, et al. The cost of cancer in Europe 2018. Eur J Cancer 2020; 129: 41–9. doi:10.1016/j.ejca.2020.01.011

¹⁵ Wieser S, Tomonaga Y, Riguzzi M, et al. Die Kosten Der Nichtübertragbaren Krankheiten in Der Schweiz. Schlussbereicht Im Auftrag Des Bundesamts Für Gesundheit (BAG), Abteilung Nationale Präventionsprogramme. Bern: BAG, 2014.

¹⁶ Feller A, Fehr M, Bordoni A, et al. Trends in incidence of oesophageal and gastric cancer according to morphology and anatomical location, in Switzerland 1982-2011. Swiss Med Wkly 2015; 145: w14245. doi:10.4414/smw.2015.14245

¹⁷ Cirillo PFA, Hošek M, Kuehni C, et al. Schweizerischer Krebsbericht 2021: Stand und Entwicklungen. Neuchatel: Bundesamt für Statistik (BFS); Nationale Krebsregistrierungsstelle (NKRS); Kinderkrebsregister (KiKR). 2021.

¹⁸ Schweiz K. Krebs in der Schweiz: wichtige Zahlen, 2020. Available: https://www.krebsliga.ch/ueber-krebs/zahlen-fakten/-dl-/fileadmin/downloads/sheets/zahlen-krebs-in-der-schweiz.pdf

¹⁹ International Agency for Research on Cancer. Cancer Today, Available: https://gco.iarc.fr/today/home

²⁰ Arndt V, Feller A, Hauri D, et al. Swiss Cancer Report 2015-Current situation and developments. 2016.

²¹ Lauzier S, Maunsell E, Drolet M, et al. Validity of information obtained from a method for estimating cancer costs from the perspective of patients and caregivers. Qual Life Res 2010; 19: 177–89. doi:10.1007/s11136-009-9575-y

²² Landfeldt E, Zethraeus N, Lindgren P. Standardized Questionnaire for the Measurement, Valuation, and Estimation of Costs of Informal Care Based on the Opportunity Cost and Proxy Good Method. Appl Health Econ Health Policy 2019; 17: 15–24. doi:10.1007/s40258-018-0418-2

²³ van den Berg B, Brouwer W, van Exel J, et al. Economic valuation of informal care: Lessons from the application of the opportunity costs and proxy good methods. Soc Sci Med 2006; 62: 835–45. doi:10.1016/j.socscimed.2005.06.046

²⁴ Obermannová R, Alsina M, Cervantes A, et al. Oesophageal cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol 2022; 33: 992–1004. doi:10.1016/j.annonc.2022.07.003

²⁵ Bouwmans C, Krol M, Brouwer W, et al. IMTA Productivity Cost Questionnaire (IPCQ). V H 2014; 17: S1098-3015(14)03721-8. doi:10.1016/j.jval.2014.08.1791

²⁶ Swiss Federal Statistical Office. Gross monthly wage (median and quartile range) by age, professional position and gender - Private and public sectors combined, Available: https://www.bfs.admin.ch/asset/en/21224918

²⁷ Verbooy K, Hoefman R, van Exel J, et al. Time Is Money: Investigating the Value of Leisure Time and Unpaid Work. V H 2018; 21: 1428–36. doi:10.1016/j.jval.2018.04.1828

²⁸ World Bank consumer price index, Available: https://data.worldbank.org/indicator/FP.CPI.TOTL?locations=NL

²⁹ OECD Purchasing power parities, Available: https://data.oecd.org/conversion/purchasing-power-parities-ppp.htm

³⁰ Pacheco Barzallo D, Hernandez R, Brach M, et al. The economic value of long-term family caregiving. The situation of caregivers of persons with spinal cord injury in Switzerland. Health Soc Care Community 2022; 30: e2297–307. doi:10.1111/hsc.13668

³¹ Spitex Zurich tariffs, Available: https://www.spitex-zuerich.ch/ueber-uns/tarife

³² Longo CJ, Fitch M, Deber RB, et al. Financial and family burden associated with cancer treatment in Ontario, Canada. Support Care Cancer 2006; 14: 1077–85. doi:10.1007/s00520-006-0088-8

³³ Finney A, Hayes DA, Collard SB. The financial impacts of cancer. methodological appendix. 2013. Available: http://www.bristol.ac.uk/media-library/sites/geography/pfrc/pfrc1211-financial-impacts-methodological-appendix.pdf

³⁴ Vo CQ, Samuelsen PJ, Sommerseth HL, et al. Comparing the sociodemographic characteristics of participants and non-participants in the population-based Tromsø Study. BMC Public Health 2023; 23: 994. doi:10.1186/s12889-023-15928-w

³⁵ Thein H-H, Jembere N, Thavorn K, et al. Estimates and predictors of health care costs of esophageal adenocarcinoma: a population-based cohort study. BMC Cancer 2018; 18:: 694. doi:10.1186/s12885-018-4620-2

³⁶ Kind A. The Swiss pension system. InInternational comparison of pension systems: an investigation from consumers’ viewpoint. Singapore: Springer Nature Singapore, 2022: 201–39.

³⁷ Urwin S, Lau Y-S, Grande G, et al. The extent and predictors of discrepancy between provider and recipient reports of informal caregiving. Soc Sci Med 2021; 277: S0277-9536(21)00222-7. doi:10.1016/j.socscimed.2021.113890

³⁸ Cong Z, Tran O, Nelson J, et al. Productivity Loss and Indirect Costs for Patients Newly Diagnosed with Early- versus Late-Stage Cancer in the USA: A Large-Scale Observational Research Study. Appl Health Econ Health Policy 2022; 20: 845–56. doi:10.1007/s40258-022-00753-w

³⁹ Simoens S. Health economic assessment: a methodological primer. Int J Environ Res Public Health 2009; 6: 2950–66. doi:10.3390/ijerph6122950

⁴⁰ McCabe C. Expanding the Scope of Costs and Benefits for Economic Evaluations in Health: Some Words of Caution. Pharmacoeconomics 2019; 37: 457–60. doi:10.1007/s40273-018-0729-z

⁴¹ Arnold M, Morgan E, Bardot A, et al. International variation in oesophageal and gastric cancer survival 2012-2014: differences by histological subtype and stage at diagnosis (an ICBP SURVMARK-2 population-based study). Gut 2022; 71: 1532–43. doi:10.1136/gutjnl-2021-325266