Out-of-pocket costs for patients diagnosed with high-grade glioma and their carers

Abstract

Background

This study aimed to describe the out-of-pocket costs incurred by patients diagnosed with high-grade glioma (HGG) and their carers in the standard care arm of the Care-IS trial in the 6 to 8 months following their diagnosis.

Methods

Carers completed monthly cost surveys detailing the out-of-pocket costs incurred by patients and carers over a 6-month period. Seventy carers reported out-of-pocket costs at baseline (within 2 months following patient diagnosis), and a maximum of 50% of participants reported costs in any subsequent month. Costs were adjusted to 2023 AUD and reported as medians with an interquartile range. Demographic factors were assessed to determine if any were significantly associated with being in the first or fourth quartile of total out-of-pocket costs at baseline.

Results

Median monthly costs for patient-carer dyads were highest at baseline ($535[IQR:$170–$930]), and 2 months post-recruitment ($314 [IQR:$150–$772]). The largest contributors to patient-carer costs were patient health service use and patient medications. Patient and carer health service use and medication costs varied over time. The median health service use and medication out-of-pocket costs for patients and carers were mostly below $100 per month; however, there was a large variance in the upper 75th percentile for these cost categories. No factors were significantly associated with higher baseline out-of-pocket costs.

Conclusions

A HGG diagnosis has a significant and sustained financial impact on people who are diagnosed and their carers. Patients experience significant additional costs relating to their diagnosis and travel to receive care, and their carers also continue to experience sustained costs whilst managing the additional tasks associated with informal caregiving.

Key Points.

• Patients with high-grade glioma and their carers face significant out-of-pocket costs, especially for health services and medications, peaking in the first 2 months post-diagnosis.

• Monthly median costs varied widely, with the highest at baseline ($535) and showing large variability in the upper-cost ranges, though no demographic factors were linked to higher costs.

• The financial burden on both patients and carers is sustained throughout the disease trajectory.

Importance of the Study.

This is one of the first studies to describe the costs of a high-grade glioma diagnosis to patients and their carers, and it demonstrates the highly variable and sustained costs that patients and carers experience following diagnosis. This information is essential for understanding the financial burden outside of direct healthcare-related costs that are captured in administrative datasets.

A diagnosis of high-grade glioma (HGG, grade III or IV glioma) is distressing for patients, their carers, and family members.¹ In addition to practical and emotional considerations,² a cancer diagnosis can have a marked financial impact. This comprises reduced ability of patients and carers to work,^3–5 and out-of-pocket costs associated with treatment.^6,7 Many health-care systems in developed countries are designed to ensure those experiencing a life-threatening illness can access and afford the treatment they need. Accessibility includes ensuring costs for care are not prohibitive. Australia’s health-care system uses several mechanisms to ensure health-care is affordable.⁸ These include a public health insurance scheme, Medicare, that subsidizes many hospital services; medical services; and tests, imaging, and scans. The Medicare Benefits Schedule (MBS) and Pharmaceutical Benefits Schedule (PBS) list the services and medications respectively subsidized by the Australian government, and the price that the government agrees to pay for each service. When service providers offer services and medications at the price the government has agreed to pay for them, it is collectively referred to as “bulk billing.”⁹ When services and medications cost more than what is listed, the additional cost that the patient needs to pay is referred to as a “gap” or “out-of-pocket” payment. To try and improve the financial accessibility of services and medications for people with frequent interactions with the health system, safety nets are put in place to increase the amount subsidized by the government once a threshold amount has been reached.⁸ In addition to Medicare, Australians may purchase private health insurance to cover some medical and allied health services that are not listed on the MBS and attend private hospitals and health services at a subsidized cost.¹⁰ In 2023, 55% of Australians had private health insurance.¹¹ For people with HGG, most blood tests, scans, and treatments are covered by Medicare; however, there is limited research describing the out-of-pocket costs they may experience in relation to their treatment and care in Australia.

In 2008, Access Economics estimated that despite being an uncommon cancer in Australia,¹² brain cancer was one of the most expensive cancers from a societal perspective with a total economic burden of just over $2.9 million (2023 AUD).¹³ The majority of this cost (83%) was due to the burden of disease, indicative of the high cost of years of life lost with this disease.¹⁴

Even without considering the economic loss resulting from people with HGG dying prematurely, there are substantial costs to the government, private health insurers, and patients and their families affected by cancer. The availability of linked datasets in Australia has provided an opportunity to comprehensively examine costs from a health service perspective. For example, Goldsbury et al.¹⁵ reported excess costs for 7624 people with a cancer diagnosis between 2006 and 2010 from the 45 and Up Study conducted in New South Wales (NSW), Australia. Records from the survey were linked with hospital and emergency department, primary care and outpatient, and pharmaceutical databases to calculate mean excess costs relative to 3 matched controls with no cancer diagnosis. A mean excess cost of $56 835 (2023 AUD) for the year preceding, and first and second year post-diagnosis was reported.¹⁵ Total costs in Australia for those living with cancer in 2013, who had been diagnosed between 2009 to 2013 were also estimated at $8 billion (2023 AUD), although findings from the 45 and Up study population are not generalizable to the Australian population.¹⁶

Other Australian researchers have used secondary data sources to estimate cancer costs.^17–21 Doran et al.²² in their analysis of the costs of skin cancer in NSW have broadened their perspective to consider the human capital losses associated with premature mortality. By contrast, earlier studies such as that by Kang et al.²³ assessing the financial impact of a lung cancer diagnosis have collected primary data, increasing the level of detail possible, but potentially at the cost of comprehensiveness. While the use of secondary data sources makes the above analyses more comprehensive than may be feasible from primary data collection, this approach generally omits the out-of-pocket costs to patients and carers.

To report on out-of-pocket costs experienced by patients, Bates et al.²⁴ used a weighting approach. These authors estimated that the mean out-of-pocket costs during the first 12 months post-diagnosis for people with eye, brain, and central nervous system cancers was $1789 (2023 AUD). This was the highest of their estimates and had the greatest variation, with a standard deviation of $2743. This variability is consistent with a recent report by Deloitte Access Economics on the cost of breast cancer; with the range for the 25th to 75th percentile in this group’s survey $1836 to $20 914 (2023 AUD). Gordon et al.^6,7 also focused on the issue of out-of-pocket costs in recent studies. Each of the above studies used administrative data, tracking out-of-pocket costs via co-payments reported alongside governmental health expenditures. Newton et al.²⁵ and Gordon et al.²⁶ did likewise, albeit with a focus on costs for people with cancer from rural regions in Western Australia (WA) and Queensland respectively, using a primary data collection study design.

None of these studies considered HGG specifically and to our knowledge there is no published assessment in Australia to date of the out-of-pocket costs for HGG patients and their carers. There are some recent international data available on the costs of HGG from the United States,^27,28 France,²⁹ Belgium,³⁰ Spain³¹ and Japan.³² With only one exception,³¹ of these studies focused on direct health-care costs. The study from Spain considered indirect costs, adopting a human capital approach to estimate the monetary value of time away from work (the full text of this paper is available only in Spanish).³¹

There is limited Australian data reporting on cancer costs experienced by people with cancer and their carers generally and for HGG specifically. The Care-IS trial was a phase III randomized controlled trial (RCT) testing a carer support intervention.³³ The data collection for this trial included the collection of costs (a costs and consequences analysis is being separately prepared). The data collected relating to out-of-pocket costs in the context of standard care are valuable given the gap in the literature, particularly for HGG. The aim of this study was to analyze the out-of-pocket costs incurred by participants of the Care-IS trial in the control arm (usual care) and the patients they cared for.

Methods

The primary objectives of this study were to (1) report on costs incurred by participants in the control (usual care) arm of the CARE-IS trial at baseline, and at each month up to 6 months after recruitment; (2) identify timepoints at which costs reported by participants peaked, and the major contributing cost categories; (3) compare participants at the lowest and highest quartile, to see if there are any demographic or treatment-related characteristics associated with the difference in their costs.

Trial methods have been published elsewhere,³³ and are described briefly below. Ethics approval was gained from participating sites (NSW: HREC 16/105; SJOG: 671; SCGH: 2013-172; Curtin University: HR 17/2013). Trial registration number: Australian and New Zealand Clinical Trials Registration (ACTRN) 1261200114787.

Recruitment

The Care-IS trial was an unblinded, phase III RCT. People receiving care for HGG were recruited from oncology providers in WA and NSW from February 2014 to September 2020. Both public (N = 6) and private (N = 1) oncology providers were included. Participants were identified consecutively by their treating medical oncologist, radiation oncologist, neurosurgeon, or neuro-oncology cancer nurse coordinator. After a brief discussion of the trial, carers were referred to a research assistant who provided further information and obtained written, informed consent to participate. Eligible participants were primary carers of patients aged ≥18 years with HGG undergoing current treatment with chemotherapy, radiotherapy (alone or in combination with one another), within 2 months of diagnosis. Carers were then randomized to the intervention or usual care. Full medical care and support provided through “usual care” was available to carers in both the intervention and control arm and there was no attention control. Recruitment and retention for participants completing cost surveys within the trial are detailed in Figure 1.

Figure 1.

Recruitment and retention of participants completing cost surveys within the Care-IS trial.

Trial Intervention

An initial telephone assessment was used to identify carers’ needs, followed by a home visit to explain and deliver a tailored resource file within 2–4 weeks of the telephone assessment, with monthly telephone follow-up for twelve months. A specialist neuro-oncology clinical nurse delivered each aspect of the interventions.³³ The focus of the intervention was to improve carer preparedness and reduce carer distress. The differences in the trial’s primary and secondary outcomes are reported elsewhere.³⁴

Out-of-Pocket Costs Study

Separate from the intervention, service use, and associated costs were collected for both the intervention and control arms of the study (Supplementary Material 1). This survey was administered to carers to collect patient and carer information on unplanned hospital visits, emergency department attendance, hospice and respite admission, and visits to health-care services including general practitioner (GP), GP nurses, medical oncologists, radiation oncologists, neurosurgeons, other medical specialists, cancer nurse coordinators, physiotherapists, occupational therapists, social workers, welfare workers, counselors, psychologists, support groups, complementary and alternative medicines, outpatient radiation and chemotherapy, pharmacy consultations, alternative health-care practitioners, and other specialists. Data collected included the number of visits, visit length, total cost, whether they received reimbursement from public or private health insurance, and how much they actually paid.

The checklist of services and costs was measured at baseline and then monthly thereafter (via phone, email or post, as per carer preference). The baseline (T0) was within 21 days of recruitment to the main study. This checklist collected information on: patient and carer health professional attendances, patient and carer medications, patient radiology services, patient pathology for which there was a co-payment required, palliative care, hospice and residential care services, transport and parking, and other agency assistance. Questions covered resource use for both patients and their carers, since the Care-IS trial focused on supporting carers. For each question, out-of-pocket costs were requested. Data were entered using a standardized data dictionary (Supplementary Material 2). Costs were interpreted as reported by participants, and they were adjusted to 2023 Australian dollars using mean consumer price indices.¹³ The net costs of the intervention are reported separately (analysis underway).

Statistical Analyses

Health service use was described as the proportion of participants who reported using services at the time point of interest. Costs were reported as per patient per month costs. Due to the highly skewed nature of health cost data, costs are reported as medians with interquartile range (IQR); as well as means with bootstrapped 95% confidence intervals.

Costs were only compared at T0 as few participants completed surveys beyond the initial survey (Table 1). Due to small cell counts, several demographic variables were condensed where considered appropriate. To identify significant predictors of being in the highest or lowest quartile of spending, chi-squared tests were performed. Demographic factors assessed included gender, age, language spoken at home, marital status, country of birth, level of education, employment status, socioeconomic status, and public hospital status. The perceived financial effect of diagnosis, reported at baseline, was also assessed. This is a measure of financial toxicity and was included to assess if the perceived impact of costs was related to actual spending. Socioeconomic status was derived from postcode, using the Socioeconomic Index for Areas—Index for Relative Disadvantage. In this index, higher deciles indicate less disadvantage. Public hospital status is an indicator of whether chemotherapy and radiation therapy were bulk-billed or not; and was classified in conversation with site investigators. As no significant factors were identified (P < .05), multivariate analysis was not performed. All analyses were performed in “IBM SPSS Statistics for Windows,” version 29 (IBM Corp.).

Table 1.

Participant Characteristics for Patients and Their Caregivers

Demographic characteristics		N (N = 69)*	%*
Gender	Male	14	20
	Female	55	80
Age (years)	Mean (SD)	57	(12)
	Median (IQR)	57	(51–64)
Length of care for patient (months)	Mean (SD)	62	(234)
	Median (IQR)	2	1.5–3
Relationship to relative/friend	Husband or male partner	52	75
	Wife or female partner	13	19
	Daughter / mother	4	5
Marital status	Married/partner	66	96
	Never married	3	4
Country of birth	Australia	43	62
	Other	27	38
Language spoken at home	English	60	87
	Other	6	9
Caring for anyone else in your home	No	53	77
	Yes	12	17
Education	Year 10 or below	10	14
	Year 12	8	12
	Technical/ business college certificate/diploma	19	28
	University degree	21	30
	Higher degree (postgraduate)	10	14
Current employment status	Full-time employed	19	28
	Part-time employed	6	9
	Unemployed	7	10
	Retired	18	26
	Other, please state	19	28
Financial effect of diagnosis	Had no effect on my financial situation	16	23
	Had a slight effect on me financially	25	36
	Had a significant effect on me financially	24	35
	Other	3	4
Socioeconomic status (SEIFA quintiles)	Highest disadvantage	5	7
	High disadvantage	9	13
	Moderate disadvantage	12	17
	Low disadvantage	16	23
	Lowest disadvantage	27	39
Hospital	Public	48	70
	Private	21	30
Completed cost surveys (months from baseline)	Baseline	69	100
	1	35	51
	2	26	38
	3	27	39
	4	23	33
	5	22	32
	6	19	28
Consecutively completed cost surveys (months)	0-1	35	51
	0-2	15	22
	0-3	7	10
	0-4	<6	—
	0-5	<6	—
	0-6	<6	—

^*Unless otherwise specified. SD, Standard deviation. IQR, Interquartile range. SEIFA, Socioeconomic index for areas, higher quintiles, lower disadvantage. Where more than once cell has a count <6, small numbers have been supressed.

Results

There were 90 patient-carer dyads in the control arm, and 98 patient-carer dyads in the intervention arm. Participants in the intervention and control arms were compared, to determine if there were any differences in cost survey completion and attrition. In each arm, participants completed a mean of 4.40 cost surveys (SD = 3.06) in the 6 months after recruitment. There was no statistically significant difference in the number of surveys completed between each arm (t(155) = 0.006, P = .498). There were no significant differences in time from recruitment to patient death between groups.³⁴ All subsequent costs are only for participants in the control arm of the study (N = 90).

Characteristics of the patient-carer dyads in the control arm are surmised below (Table 1). Of the 90 carer-patient dyads in the control arm, 69 completed at least one cost survey. Their demographic characteristics are presented in Table 1. Carers were on average 57 years old (standard deviation (SD) = 12 years); had been caring for a median of 2 months (interquartile range (IQR) = 1.5–3 months). The majority of carers were female (80%), caring for their husband (75%) or wife (19%), were married (96%), born in Australia (62%), spoke English at home (87%), had a technical college or university qualification (58%), were employed full- or part-time (37%), and felt that the diagnosis of their partner/relative had any financial impact (81%). Most participants (70%) attended a public hospital.

Patient-Carer Dyads’ Costs, as Reported By Carers

Whilst we intended to collect cost surveys each month following recruitment, carers had the opportunity to opt-in to complete a survey each month to minimize inconvenience. Sixty-nine participants completed a cost survey at baseline when they entered the study, 35 (48%) completed a survey one month after entering the study, and 26%–35% of participants completed the monthly surveys 2 to 6 months after entering the study (Table 1). For this reason, costs are reported on a monthly basis rather than cumulatively. Low retention was observed over the 12-month follow-up period due to participant withdrawal, patients’ decline in health, and death.³⁴

On entry into the study, the median total monthly costs for 69 patient-carer dyads were $535 (IQR:$170–$930). The highest median costs were for childcare (N = 3, $466, IQR:$89-$37 637) and imaging (N = 7, $420, IQR: $333–$640). Costs that comprised most of the carer-patient spending were for childcare (30%) and patient health service use (46%; Table 2). Following the baseline, total costs reported for carer-patient dyads ranged from a median of $198 (IQR: $93–$293) at 6 months, to $314 (IQR:$150–$772) at 2 months.

Table 2.

Monthly Costs (AUD 2023) Reported by Participants From T0 to T6

Time point	Cost category	N (%) who reported cost data	N (%) who reported costs > $0	Median	Percentile 25	Percentile 75	Sum	% of monthly total cost
0	Imaging	7 (10)	7 (10)	368	292	561	2804	3
	Pathology	4 (6)	2 (3)	3	0	25	49	0
	Silver chain services	2 (3)	1 (1)	3	0	5	5	0
	Hospital admissions	6 (9)	1 (1)	0	0	0	53	0
	Patient travel	52 (75)	48 (72)	146	48	283	9930	9
	Care facility admissions	3 (4)	1 (1)	0	0	33	33	0
	Carer hospital admissions	0 (0)	0 (0)					0
	Carer travel	12 (17)	11 (16)	11	4	57	534	0
	Childcare	3 (4)	3 (4)	409	78	33 015	33 501	30
	Patient health service use	54 (78)	34 (51)	39	0	204	51 986	46
	Patient medications	46 (67)	46 (69)	105	42	214	9240	8
	Carer health service use	28 (41)	17 (25)	18	0	75	1805	2
	Carer medications	34 (49)	34 (51)	52	22	95	2106	2
	Total costs	69 (100)	67 (100)	469	149	816	112 047	100
1	Imaging	1 (3)	1 (3)	430	430	430	430	4
	Pathology	1 (3)	0 (0)	0	0	0	0	0
	Silver chain services	1 (3)	0 (0)	0	0	0	0	0
	Hospital admissions	1 (3)	0 (0)	0	0	0	0	0
	Patient travel	25 (71)	25 (78)	77	29	192	3323	31
	Care facility admissions	1 (3)	0 (0)	0	0	0	0	0
	Carer hospital admissions	1 (3)	1 (3)	266	266	266	266	2
	Carer travel	7 (20)	4 (13)	5	0	11	252	2
	Childcare	0 (0)	0 (0)					0
	Patient health service use	22 (63)	12 (38)	22	0	116	1908	18
	Patient medications	23 (66)	23 (72)	76	31	151	2996	28
	Carer health service use	15 (43)	9 (28)	20	0	36	613	6
	Carer medications	15 (43)	15 (47)	48	33	98	968	9
	Total costs	35 (100)	32 (100)	175	55	387	10 756	100
2	Imaging	4 (15)	4 (15)	250	81	395	951	8
	Pathology	0 (0)	0 (0)					0
	Silver chain services	2 (8)	0 (0)	0	0	0	0	0
	Hospital admissions	3 (12)	2 (8)	38	0	1061	1098	9
	Patient travel	22 (85)	21 (81)	35	16	125	2213	19
	Care facility admissions	0 (0)	0 (0)					0
	Carer hospital admissions	0 (0)	0 (0)					0
	Carer travel	4 (15)	4 (15)	35	11	128	277	2
	Childcare	0 (0)	0 (0)					0
	Patient health service use	21 (81)	14 (54)	27	0	274	2923	25
	Patient medications	19 (73)	19 (73)	76	31	215	2556	22
	Carer health service use	11 (42)	8 (31)	43	0	149	1019	9
	Carer medications	10 (38)	9 (35)	40	28	96	611	5
	Total costs	26 (100)	26 (100)	275	132	677	11 649	100
3	Imaging	4 (15)	4 (17)	364	262	413	1350	21
	Pathology	2 (7)	1 (4)	15	0	31	31	0
	Silver chain services	1 (4)	0 (0)	0	0	0	0	0
	Hospital admissions	2 (7)	0 (0)	0	0	0	0	0
	Patient travel	18 (67)	17 (71)	28	15	84	917	14
	Care facility admissions	0 (0)	0 (0)					0
	Carer hospital admissions	0 (0)	0 (0)					0
	Carer travel	4 (15)	4 (17)	6	5	8	27	0
	Childcare	0 (0)	0 (0)					0
	Patient health service use	17 (63)	8 (33)	0	0	84	1548	24
	Patient medications	18 (67)	16 (67)	70	17	176	1777	27
	Carer health service use	10 (37)	4 (17)	0	0	123	498	8
	Carer medications	10 (37)	8 (33)	33	8	58	357	5
	Total costs	27 (100)	24 (100)	147	22	353	6504	100
4	Imaging	1 (5)	1 (5)	27	27	27	27	0
	Pathology	0 (0)	0 (0)					0
	Silver chain services	3 (15)	0 (0)	0	0	0	0	0
	Hospital admissions	4 (20)	2 (10)	11	0	272	544	10
	Patient travel	12 (60)	11 (55)	45	25	120	1205	21
	Care facility admissions	0 (0)	0 (0)					0
	Carer hospital admissions	1 (5)	0 (0)	0	0	0	0	0
	Carer travel	2 (10)	1 (5)	5	0	10	10	0
	Childcare	0 (0)	0 (0)					0
	Patient health service use	16 (80)	9 (45)	11	0	49	939	16
	Patient medications	18 (90)	17 (85)	47	33	192	2410	42
	Carer health service use	7 (35)	4 (20)	30	0	75	231	4
	Carer medications	7 (35)	7 (35)	14	11	57	356	6
	Total costs	20 (100)	20 (100)	215	98	350	5723	100
5	Imaging	2 (10)	2 (10)	164	11	317	328	2
	Pathology	0 (0)	0 (0)					0
	Silver chain services	2 (10)	0 (0)	0	0	0	0	0
	Hospital admissions	3 (14)	1 (5)	0	0	62	62	0
	Patient travel	14 (67)	14 (70)	37	27	64	5984	30
	Care facility admissions	1 (5)	0 (0)	0	0	0	0	0
	Carer hospital admissions	2 (10)	2 (10)	1065	85	2044	2129	11
	Carer travel	4 (19)	3 (15)	19	9	24	64	0
	Childcare	0 (0)	0 (0)					0
	Patient health service use	15 (71)	9 (45)	32	0	74	4380	22
	Patient medications	19 (90)	17 (85)	75	32	246	6093	31
	Carer health service use	7 (33)	5 (25)	53	0	102	442	2
	Carer medications	8 (38)	8 (40)	33	21	51	303	2
	Total costs	21 (100)	20 (100)	206	85	452	19 786	100
6	Imaging	0 (0)	0 (0)					0
	Pathology	0 (0)	0 (0)					0
	Silver chain services	1 (5)	1 (5)	0	0	0	0	0
	Hospital admissions	0 (0)	0 (0)					0
	Patient travel	13 (68)	13 (68)	43	20	52	519	9
	Care facility admissions	1 (5)	0 (0)	0	0	0	0	0
	Carer hospital admissions	0 (0)	0 (0)					0
	Carer travel	4 (21)	3 (16)	8	3	43	92	2
	Childcare	0 (0)	0 (0)					0
	Patient health service use	14 (74)	11 (58)	49	19	102	953	16
	Patient medications	18 (95)	18 (95)	101	39	147	4108	69
	Carer health service use	4 (21)	3 (16)	16	7	69	151	3
	Carer medications	6 (32)	6 (32)	25	15	41	161	3
	Total costs	19 (100)	19 (100)	174	82	257	5985	100

Time points are months from baseline. Costs are adjusted to 2023 AUD using the Australian Bureau of Statistics Consumer Price Index.

Imaging costs comprised the highest median out-of-pocket cost reported by patient-carer dyads, despite few participants having costs greater than $0 (N = 2–7; Table 2). Other out-of-pocket costs that were consistently high each month were for patient medication and health service use (Figure 2). As a proportion of total costs reported by the patient-carer dyads, cost categories that contributed the most to report costs across the 6 months since joining the study were patient health service use (46%) and childcare (29%) at baseline; patient travel (31%) and patient medication 28% at month one; patient medication (22%) and patient health service use (25%) at month 2; patient medication (27%) and health service use (24%) at month 3; patient medication (42%) and patient travel (21%) at month 4; patient medication (31%) and patient travel (30%) at month 5, and patient medication (69%) and health service use (16%) at month 6. A breakdown of services used at each timepoint by patients and carers are presented in Supplementary Tables 1 and 2.

Figure 2.

Of 25th, 50th, and 75th percentile of costs reported over 6 months from T0, for patient-carer dyads.

Note: Costs are 2023 AUD. * 75th percentile = $33015. ** 75th percentile = $1098. *** 75th percentile = $2044.

Figure 3 demonstrates that, whilst variation in patient travel-related costs clearly decreases over time, patient and carer health service use and medication costs varied. The median health service use and medication out-of-pocket costs for patients and carers were mostly below $50 per month; however, there was large variance in the upper 75^th percentile for these cost categories.

Figure 3.

Boxplots depicting monthly carer- reported costs (2023 AUD) for patients and carers.

Predictors of Costs

When comparing patient-carer dyads with the highest and lowest total out-of-pocket costs at baseline, no demographic factors were associated with significant differences (Table 3). Hence, multivariate analyses were not conducted.

Table 3.

Demographic Characteristics of Those in the Lowest and Highest Quartile of Total Out-of-Pocket Costs at Entry Into Study

		Quartile 1*		Quartile 4*
		(N = 17)	%	(N = 17)	%	P value**
Gender	Male	5	29	3	18	.419
	Female	12	71	14	82
Marital status	Married/Partner	15	88	16	94	.965
	Never Married	1	6	1	6
Country of birth	Australia	9	53	14	82	.067
	Other	8	47	3	18
Language spoken at home	English	16	94	16	94	.325
	Other	0	0	1	6
Perceived financial effect of diagnosis	Had no effect on my financial situation	6	35	5	29	.661
	Had a slight effect on me financially	7	41	7	41
	Had a significant effect on me financially	3	18	5	29
	Other (please state)	1	6	0	0
Education status	Yr12 or below	3	18	5	29	.607
	Technical/ business college certificate/diploma	5	29	3	18
	University degree or higher	8	47	8	47
Employment status at time of T0	employed (FT/PT)	8	47	5	29	.264
	retired/not employed	6	35	6	35
	other	2	12	6	35
Age***	Mean (SD)	58.5 (11.4)		54.2 (10.8)		.131
Socioeconomic status (SEIFA quintiles)	Highest disadvantage	2	12	2	12	.335
	High disadvantage	1	6	0	0
	Moderate disadvantage	1	6	1	6
	Low disadvantage	6	35	2	12
	Lowest disadvantage	3	18	7	41

SD, Standard deviation. SEIFA, Socioeconomic index for areas, higher quintiles, lower disadvantage. *data presented are counts and percentage, unless otherwise specified. **Data are compared using chi squared tests. *** Compared using t-test.

Discussion

This is one of the first studies to report on out-of-pocket costs from the perspective of patients and their primary carers in the first 6 months following the diagnosis of HGG. Whilst other studies have examined the substantial medical costs for people with HGG and the health system and indirect costs caused by loss of productivity,^27–31 self-reported costs provide important insight into the financial burden experienced by people diagnosed with HGG and their carers from the patient’s perspective. The financial impact of a cancer diagnosis on carers has received limited attention.³⁵ Quantifying the costs they experience following the diagnosis of the care recipient is essential to understand the full costs of caregiving, and help to ensure services and policies are able to meet the needs of patients and their carers.

It is difficult to compare these findings to other Australian studies that reported costs for people receiving cancer care as we were unable to collect data for costs experienced to the end of treatment, which is the time period used by other Australian studies investigating self-reported cancer costs.^25,26 Whilst collecting self-reported cost data from people with terminal diagnoses was possible in this study by obtaining costs from carers, the high loss to follow-up and missing data between months suggest this study may not reflect costs for people with HGG and their carers experienced during periods of high health-care use. The costs reported still provide an important insight into costs experienced by people receiving care for HGG and their carers. Interestingly, we found that the monthly out-of-pocket costs reported for some patients (Figure 3) were higher than the weighted yearly estimate for out-of-pocket costs calculated by Bates et al. for people with brain, eye, or central nervous system cancers.²⁴ These costs were attributed to patient health service use, travel costs, and medication. Whilst few participants reported these high costs, it is concerning that they paid 1.2 to 12.9 times more in one month than Bates et al. estimated they would pay in a year.²⁴ This is particularly concerning in the context of Rodriguez-Acevedo et al.’s findings, who analyzed cancer costs for Australians diagnosed with breast, lung, colorectal, melanoma, or prostate cancer from 2011 to 2015 and found that they increased substantially over time.³⁶

High costs were also observed in other cost categories. At baseline, childcare costs accounted for the greatest portion of overall expenditure (30%), and subsequently dropped off thereafter. These costs were incurred by only 3 participants, indicating the large financial impact a brain cancer diagnosis can have on families with children. Limited research has explored childcare costs for people with cancer. Whilst it has been acknowledged as an indirect cost people may face,³⁷ few studies investigate the childcare needs of people with advanced cancer. Li et al. recently surveyed patients with minor children at home at a large Canadian cancer treatment center.³⁸ They found that balancing childcare with treatment significantly affected participants (N = 64). 40% needed to reschedule and 14% missed at least one appointment due to childcare conflicts or needs. Further research investigating the needs of cancer patients with children, and identifying ways they can be supported to coordinate affordable childcare is needed.

Few participants reported having out-of-pocket costs for imaging, and the substantial median out-of-pocket costs they reported are concerning as they represent high costs paid by a small portion of participants. Frequent and repeated imaging including MRI or CT scans are essential for the management of HGG, with guidelines recommending scans occur within 48 hours following surgery, 2–8 weeks post-irradiation, then 2–4 months for 3 years, and 3–6 months thereafter.^39,40 These guidelines do not include additional scans that often occur due to changes in patient symptoms, or additional scans required for treatment planning. In Australia, imaging is typically billed directly to Medicare if it is part of usual care. The available data does not indicate whether these scans were routine, or in addition to usual care; so we are unable to discern if the high costs are due to imaging requested outside of usual care or if these are large gap fees from service providers. Furthermore, the available data does not include if participants were able to comfortably afford these imaging costs, however, it is concerning that a small proportion of participants paid some of the highest median out-of-pocket costs in a given month. Jiang et al.’s American study found >7 MRI scans to be significant contributors to higher medical costs.²⁸ Furthermore, Di Nunno et al.’s recent systematic review of articles investigating the correlation between economic income and survival in patients with glioblastoma found that lower economic income was associated with poorer survival (pooled hazard ratio: 1.09, 95% CI: 1.02–1.17).⁴¹ Growing concerns about financial toxicity in cancer care have led to calls for increased financial transparency between providers and patients; health professionals have indicated a need for resources to aid financial discussions and referrals for support.⁴² Future research investigating out-of-pocket costs needs to include measures to assess the perceived financial impact of these additional costs, such as the Comprehensive Score for Financial Toxicity measure.^43,44 This tool measures financial distress experienced by cancer patients.

Patients had understandably higher costs compared to carers as demonstrated in Figure 3; however, it is important to recognize that carers reported persisting health-related costs that demonstrate their own needs they need to meet whilst juggling the needs of the care recipient. Although the Australian public health-care system has mechanisms in place to safeguard families against high medical costs it is evident that medication and health service use still incur substantial out-of-pocket costs to the people with HGG and their carers. Several medications used in HGG are not covered by the Pharmaceutical Benefits Scheme (PBS). The PBS is part of Australia’s public health system and regulates the costs of approved medications, ensuring Australians pay a subsidized co-payment for approved medications.⁴⁵ Unsubsidized treatment options including bevacizumab (prior to 2019), lomustine, and procarbazine are not PBS listed, and if prescribed patients may have to pay the full cost out-of-pocket.

Interestingly, no demographic factors predicted high out-of-pocket costs at baseline. This is in contrast to studies conducted in other cancer patient populations that found factors such as younger age, being a First Nations person, rurality, household income, employment status, and private health insurance status to be associated with greater out-of-pocket costs.⁴⁶ It is possible that factors that we were not able to account for, such as household income, may have been stronger predictors of out-of-pocket costs. Additionally, our analysis looked at costs for participants who were 2 to 3 months from diagnosis, while other studies compared total costs at the end of treatment.

Limitations

Several factors limit the generalisability of these findings. It is possible that some details in the reported costs may be over- or under-represented as these data were self-reported and not verified with documentation. For each survey, participants were asked to report on costs experienced in the last 4 weeks and they may have forgotten costs to report. Furthermore, in the analysis of the primary trial outcomes it was reported that unmarried, divorced, or widowed carers were more likely to drop out at 2 months, and those who had reduced work hours or stopped working were more likely to drop out at 4, 6 and 12 months—these participants may have been experiencing costs that differed significantly from those that continued to stay in the study.³⁴

The lack of data on household income prevents assessing these out-of-pocket costs as a percentage of annual household income—health expenditure as a proportion of household income is a measure of financial catastrophe.⁴⁷

Conclusion

A HGG diagnosis has a significant and sustained financial impact on people who are diagnosed and their carers. Patients experience significant additional costs relating to their diagnosis and travel to receive care, and their carers also continue to experience sustained costs whilst managing the additional tasks associated with informal caregiving. Work is needed to support people with HGG and their carers manage the financial impact of HGG on the patient-carer dyad.

Supplementary material

Supplementary material is available online at Neuro-Oncology Practice (https://academic.oup.com/nop/).

Funding

This project was funded by a Cancer Australia Priority-driven Collaborative Cancer Research Scheme project grant (APP1105307). Georgia Halkett is currently supported by a Cancer Council of WA Research Fellowship. Jade Newton was a recipient of an Australian Government Research Training Program Scholarship and Cancer Council of Western Australia Top Up Scholarship. Data analysis was supported by funding from the Medical Research Future Fund for the Brain cancer Rehabilitation, Assessment, Intervention of survivor NeedS (BRAINS) project.

Conflict of interest statement

None to report.

Authorship statement

G.H. provided the dataset. J.N. and C.W. prepared the material. R.M. and C.W. provided statistical advice. J.N. conducted the data analysis and prepared the first draft of the manuscript. G.H., A.N., M.O., R.M., C.W., and J.N. commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Patient consent statement

All participants in this research provided written, informed consent prior to taking part in research activities.