Who are the vulnerable lung cancer patients at risk for not receiving first-line curative or palliative treatment?

Abstract

Background

To identify non-small-cell lung cancer (NSCLC) patients in need of comprehensive support, we examined the association between patient and disease-related factors of vulnerability related to not receiving guideline-recommended treatment.

Material and methods

We identified 14,597 non-small-cell lung cancer (NSCLC) patients with performance status <3 during 2013–2018 in the Danish Lung Cancer Registry. Multivariate logistic regression models were used to estimate Odds Ratios (ORs) and 95% confidence intervals (CIs) for receiving guideline-recommended treatment according to stage, comorbidities, age, performance status, long distance to hospital, cohabitation status, education and alcohol abuse.

Results

21% of stage I-IIIA NSCLC patients did not receive curative treatment while 10% with stage IIIB-IV did not receive any oncological therapy. Factors associated with reduced likelihood of receiving curative treatment included: advanced stage (OR = 0.45; 95% CI = 0.42–0.49), somatic comorbidity (OR = 0.72; 95% CI = 0.63–0.83), age ≥ 80 years (OR = 0.59; 95% CI = 0.55–0.64), performance status = 2 (OR = 0.33; 95% CI = 0.28–0.39) and living alone (OR = 0.79; 95% CI = 0.69–0.90). Results were similar for stage IIIB-IV NSCLC patients, although a statistically significant association was also seen for long distances to the hospital (OR = 0.71; 95% CI = 0.58–0.86).

Conclusions

Several factors are associated with not receiving guideline-recommended NSCLC treatment with age, performance status, comorbidity and stage being most predictive of no treatment receipt. Efforts should be made to develop support for vulnerable lung cancer patients to improve adherence to optimal first-line therapy.

Keywords:

• Non-small cell lung cancer
• curative treatment
• palliative treatment
• vulnerability factors

Background

Lung cancer survival is strongly associated with low socioeconomic position (SEP) [1–4]. Social disparities in lung cancer emerge throughout the entire cancer trajectory from the diagnostic work-up into treatment, follow-up and rehabilitation [4], and access to recommended treatment may contribute to poorer prognostic outcomes among patients with low SEP [1]. Several studies have found that lung cancer patients with low SEP (defined by education and/or income) or who live alone are less likely to receive first-line treatment compared with patients of high SEP and patients who live with a partner, regardless of stage and health care system [1,2,4–7]. The underlying factors that drive both treatment decisions and treatment adherence among patients with low SEP are complex and may include clustering effects of adverse health behavior, poor physical condition and limited psychosocial resources [8,9]. Bringing these adverse factors into a highly complex and specialized healthcare system with a focus on efficiency may further inhibit the patient’s ability to adhere to treatment.

To help identify the vulnerable lung cancer patients in ned of comprehensive support, we examined patient and disease-related factors that may contribute to disparities in receipt of guideline-recommend treatment. Clinical decision-making for lung cancer treatment is a collaboration between the treating physician and the patient. Clinical vulnerability factors such as high age, comorbidity, poor performance status and advanced stage have been shown to influence treatment decisions [10–15] likely due to considerations of an increased risk for treatment complications related to poor performance status and high comorbidity burden [14,15]. Relatively few studies have examined the association between patient vulnerability factors such as mental frailty [16,17], alcohol abuse [18] and long distance to the hospital and received treatment [11,19–22]. Finally, no previous study has examined the impact of an increasing number of vulnerability factors on receipt of treatment nor which factors are the strongest predictors in relation to treatment initiation.

The aim of this population-based study was therefore to investigate the association between patient and disease-related factors of vulnerability and the probability of not receiving guideline-recommended treatment among patients diagnosed with non-small-cell lung cancer (NSCLC) during 2013–2018.

Material and methods

We identified a cohort of lung cancer patients within the Danish Lung Cancer Registry (DLCR). The DLCR was established in 2001 and includes detailed and validated clinical as well as first-line curative and palliative treatment information on more than 90% of patients diagnosed since 2003 [23]. We included patients diagnosed since 2013 to have similar treatment recommendations during the study period.

Patient and disease-related vulnerability factors

We retrieved information on clinical vulnerability factors as well as epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) status from the DLCR and defined vulnerability as being diagnosed with high disease stage (IIIB–IV), above 80 years of age, and having a performance status of 2. According to national guidelines in Denmark, stage IIIB is included as high-stage and treatment is defined as palliative. Performance status is based on a clinical evaluation recorded in the DLCR at the date of treatment decision. Lung cancers were categorized as low-stage (I–IIIA) and high-stage (IIIB–IVB) according to the TNM classification to match guidelines for curative and palliative treatment, respectively. First-line treatment guidelines were defined according to both TNM stage and performance status as shown in Table 1. In post-hoc sensitivity analyses, we categorized lung cancers in three groups 1) stage I–II, 2) stage IIIA-IIIB and 3) stage IIIC-IV to examine whether vulnerability factors would differ for stage I-II patients receiving low-toxic treatments with short duration compared with stage IIIA-IIIB patients receiving high-risk toxic treatments with a longer duration.

Table 1. First-line treatment of non-small cell lung cancer according to clinical TNM stage and performance status.

TNM stage	Stage	Performance status	First-line treatment
T1-2N0M0	I-IIA	0–2	Surgery / Stereotactic radiation / radiotherapy + chemotherapy / Ablation
T3-4N0M0	IIB-IIIA	0–1	Surgery / Chemotherapy / Immunotherapy / Radiotherapy + chemotherapy
T3-4N0M0	IIB-IIIA	2	Surgery / Chemotherapy / Radiotherapy + chemotherapy
T0-3N1M0	IA-IIIA	0–1	Surgery + Chemotherapy / Immunotherapy / Radiotherapy + chemotherapy
T0-3N1M0	IA-IIIA	2	Radiotherapy + chemotherapy / Chemotherapy
T0-3N2-3M0	IIIA-IIIC	0–2^a	Radiotherapy + chemotherapy / Chemotherapy / Immunotherapy
T4N1-3M0	IIIA-IIIC	0–2^a	Chemotherapy / Radiotherapy + chemotherapy / Immunotherapy
TxMxM1	IVA-IVB	0–2^a	Chemotherapy / Radiotherapy + chemotherapy / Immunotherapy / surgery

^aExcept no immunotherapy is given to patients with performance status 2.

The choice of treatment regimen depends on the individual patient’s performance score, comorbidities, and mutation status. If diagnosed with oncogene addicted non-small cell lung cancer, patients are treated with targeted therapy.

By linkage to the Danish National Patient Register [24] we obtained information on hospitalization with somatic, mental and alcohol-related comorbidities diagnosed from five years to three months prior to the lung cancer diagnosis using the unique personal identification number assigned to all Danish residents. The Danish National Patient Register includes detailed information on dates of hospitalizations since 1977 and diagnoses are coded according to the International Classification of Diseases (ICD) version 10 since 1994 [24]. Scores from the updated weighted Charlson Comorbidity Index were used to classify lung cancer patients with (≥ 1) or without (=0) somatic comorbidity taking into account the number and seriousness of 19 comorbid chronic conditions such as heart disease, diabetes and (non-lung) cancer [25].

Information on the distance to the hospital, cohabitation status and education was retrieved through linkage to the nationwide registers at Statistics Denmark [26]. Distance to the hospital (short; long) was calculated using hospitals and patients’ place of residence and long distance was defined as more than 200 kilometers to the hospital or more than two hours of transport. Cohabitation status was defined as living alone or with a partner while the highest attained education was dichotomized into short or medium (mandatory school up to 7 or 9, secondary school and vocational education up to 12 years) and high (>12 years).

First-line curative or palliative treatment

We used the DLCR to retrieve information on receipt of first-line curative or palliative treatment (yes; no), respectively, which was the treatment given to the patient.

Statistical analyses

To characterize the study population, we computed frequencies and percentages of vulnerability factors according to stage at diagnosis. We used logistic regression models to estimate unadjusted and mutually adjusted (to estimate each vulnerability factor independently while taking into account the other factors) Odds Ratios (ORs) and 95% confidence intervals (CIs) for receipt of first-line curative treatment among low-stage patients and for receipt of any first-line treatment among high-stage patients, respectively, according to the number of vulnerability factors (at least 2–5). Logistic regression models were also used in post-hoc sensitivity analyses to estimate mutually adjusted ORs and CIs for receiving treatment according to stage categorized in three groups (stage I-II, stage IIIA-IIIB and stage IIIC-IV) and mutation status (EGFR and ALK).

In models estimating ORs and 95% CI for receiving treatment according to number of vulnerability factors, we computed the Akaike information criterion (AIC) and Area Under receiver operating characteristic Curve (AUC) to evaluate the best model fit with smaller AIC values and highest AUC values representing better model fit. All analyses were performed in Stata (version 16).

Results

The cohort comprised 23,738 lung cancer patients diagnosed during 2013–2018. We excluded patients who were diagnosed with small cell lung cancer (N = 3337), with no information on pathology (N = 1177), with an Eastern Cooperative Oncology Group performance status ≥ 3 (N = 4228) since there are no recommended treatment guidelines for this patient group, or patients with no information on stage at diagnosis (N = 399). The final cohort comprised 14,597 patients diagnosed with non-small-cell lung cancer. The mean age for low-stage (I-IIIA) patients was 70 years (SD 9.3; range 17–93) and 69 years (SD 9.6; range 18–102) for high-stage (IIIB-IVB) patients. The distribution of vulnerability factors among the 14,597 lung cancer patients was similar in patients diagnosed with low and high-stage NSCLC, except for patients diagnosed with somatic comorbidity (59% low-stage patients vs. 48% high-stage patients) and patients having a performance status of 2 (12% of low stage patients and 20% of high stage patients) (Table 2). At least three vulnerability factors were present among 17% of stage I-IIIA patients and 51% of stage IIIB-IVB patients (Table 2).

Table 2. Characteristics and number of vulnerability factors among 14,597 patients diagnosed with non-small cell lung cancer during 2013–2018 according to stage.

Vulnerability factors	Low-stage (I–IIIA)		High-stage (IIIB–IVB)
	Number	%	Number	%
Total number of patients	6555	100	8042	100
Age at diagnosis
<60	770	12	1150	14
60–69	2240	34	3025	38
70–79	2607	40	2832	35
>80	938	14	1035	13
Gender
Women	3323	51	3957	49
Men	3232	49	4085	51
Year of diagnosis
2013–2015	3676	56	4777	59
2016–2018	2879	44	3265	41
Education
Low	3944	60	4889	62
Medium	1969	31	2418	31
High	496	8	558	7
Living alone
Yes	2770	42	3330	41
No	3777	58	4706	59
Long distance to hospital
Yes	1059	16	904	11
No	5496	84	7138	89
Charlson comorbidity index
Yes	354	59	4085	48
No	2701	49	4356	52
Mental comorbidity
Yes	138	2	168	2
No	6417	98	7874	98
Performance status
0	3602	55	3309	41
1	2187	33	3129	39
2	766	12	1604	20
Alcohol consumption
High	130	2	113	1
Low	6425	98	7929	99
Number of vulnerability factors
No vulnerability factors	943	14	0	0
1 vulnerability factor	2461	38	1118	14
2 vulnerability factors	2047	31	2847	35
3 vulnerability factors	861	13	2610	32
4 vulnerability factors	214	3	1147	14
5 vulnerability factors	28	0.4	281	3
Co-occurring vulnerability factors
At least 1 vulnerability factor	5612	86	8042	100
At least 2 vulnerability factors	3151	48	6924	86
At least 3 vulnerability factors	1104	17	4077	51
At least 4 vulnerability factors	243	4	1467	18
At least 5 vulnerability factors	29	0.4	320	4

During the study period, 21% of NSCLC patients with stage I-IIIA did not receive curative treatment (16% received palliative treatment and 5% did not receive any treatment) while 10% with stage IIIB-IV did not receive any oncological therapy. Vulnerability factors significantly associated with reduced likelihood of receiving first-line curative treatment among low-stage lung cancer patients included: advanced-stage (OR = 0.45; 95% CI = 0.42–0.49), somatic comorbidity (OR = 0.72; 95% CI = 0.63–0.83), age ≥ 80 years (OR = 0.59; 95% CI = 0.55–0.64), a performance status of 2 (OR = 0.33; 95% CI = 0.28–0.39) and living alone (OR = 0.79; 95% CI = 0.69–0.90)(Table 3). We found no statistically significant associations between mental comorbidity (OR = 0.69; 95% CI = 0.45–1.08) alcohol abuse (OR = 0.79; 95% CI = 0.51–1.24) as well as education (OR = 0.89; 95% CI = 0.78–1.03) and the likelihood of receiving first-line treatment (Table 3). Results were similar for first-line treatment among high-stage patients, but with a more pronounced and statistically significantly reduced likelihood of receiving treatment when having a long distance to the hospital (OR = 0.71; 95% CI = 0.58–0.86) (Table 3). Only small differences were seen for vulnerability factors and the likelihood of receiving treatment between stage I-II and stage IIIA-IIIB patients (Supplemental Table 1). EGFR mutations were associated with a higher likelihood of receiving treatment among stage IIIB-IV patients (OR 1.88; 95% CI 1.29–2. 75), but the association for ALK mutations did not reach statistical significance (OR 1.92; 95% CI 0.77–4.82) (Supplemental Table 2). Including mutation status in the model led to similar results, except that a significant positive association was seen for comorbidity (Supplemental Table 2).

Table 3. Odds ratio for receiving treatment among low-stage and among high-stage non-small cell lung cancer patients according to vulnerability factors.

Vulnerability factors	Low-stage (I-IIIA)								High-stage (IIIB-IVB)
	Did not receive curative treatment		Received curative treatment		Unadjusted		Adjusted^a		Did not receive treatment		Received treatment		Unadjusted		Adjusted*
	N	%	N	%	OR	95 % CI	OR	95 % CI	N	%	N	%	OR	95 % CI	OR	95 % CI
Stage
I	483	13	3202	87	1	(reference)	1	(reference)
II	322	22	1150	78	0.54	(0.46–0.63)	0.53	(0.45–0.62)
IIIA	548	39	850	61	0.49	(0.45–0.52)	0.45	(0.42–0.49)
IIIB									96	8	1091	92	1	(reference)	1	(reference)
IIIC									55	11	451	89	0.72	(0.51–1.02)	0.75	(0.52–1.08)
IVA									251	16	1351	84	0.47	(0.37–0.61)	0.50	(0.38–0.64)
IVB									739	16	4008	84	0.80	(0.75–0.85)	0.79	(0.74–1.01)
CCI
No	440	16	2261	84	1	(reference)	1	(reference)	730	18	3355	82	1	(reference)	1	(reference)
Yes	913	24	2941	76	0.63	(0.55–0.71)	0.72	(0.63–0.83)	488	11	3868	89	0.58	(0.51–0.66)	0.77	(0.67–0.88)
Mental comorbidity
No	1322	21	5095	79	1	(reference)	1	(reference)	27	16	141	84	1	(reference)	1	(reference)
Yes	31	22	107	78	0.90	(0.60–1.34)	0.69	(0.45–1.08)	1114	14	6760	86	0.86	(0.57–1.31)	0.72	(0.46–1.12)
Age
<80	988	18	4629	82	1	(reference)	1	(reference)	822	12	6185	88	1	(reference)	1	(reference)
>80	365	39	573	61	0.34	(0.29–0.39)	0.59	(0.55–0.64)	319	31	716	69	0.29	(0.26–0.35)	0.59	(0.54–0.64)
Performance status
<2	1012	17	4777	83	1	(reference)	1	(reference)	685	11	5753	89	1	(reference)	1	(reference)
=2	341	45	425	55	0.26	(0.23–0.31)	0.33	(0.28–0.39)	456	28	1148	72	0.30	(0.26–0.34)	0.38	(0.33–0.44)
Living alone
No	690	18	3087	82	1	(reference)	1	(reference)	544	12	4162	88	1	(reference)	1	(reference)
Yes	662	24	2108	76	0.71	(0.63–0.80)	0.79	(0.69–0.90)	596	18	2734	82	0.60	(0.53–0.68)	0.67	(0.58–0.77)
Long distance to hospital
No	1112	20	4384	80	1	(reference)	1	(reference)	978	14	6160	86	1	(reference)	1	(reference)
Yes	241	23	818	77	0.86	(0.74–1.01)	0.90	(0.76–1.07)	163	18	741	82	0.72	(0.60–0.87)	0.71	(0.58–0.86)
Alcohol consumption
Low	1322	21	5103	79	1	(reference)	1	(reference)	17	15	96	85	1	(reference)	1	(reference)
High	31	24	99	76	0.83	(0.55–1.24)	0.79	(0.51–1.24)	1124	14	6805	86	0.93	(0.55–1.57)	0.91	(0.53–1.57)
Education
High	527	21	1938	79	1	(reference)	1	(reference)	442	15	2534	85	1	(reference)	1	(reference)
Low	826	20	3264	80	1.07	(0.95–1.22)	0.89	(0.78–1.03)	699	14	4367	86	1.09	(0.95–1.24)	0.89	(0.78–1.02)

Abbreviations: N: number; OR: odds ratio; CI: confidence interval; CCI: Charlson comorbidity index.

^aMutually adjusted for all variables in the table.

For patients with low stage, we observed a reduced likelihood of receiving first-line curative treatment with increasing number of vulnerability factors: at least 2 factors (OR = 0.57; 95% CI = 0.46–0.69), at least 3 factors (OR = 0.54; 95% CI = 0.48–0.61), at least 4 factors (OR = 0.40; 95% CI = 0.35–0.47) and at least 5 factors (OR = 0.29; 95% CI = 0.22–0.37) (Table 4). Results were similar for receiving first-line treatment for patients with high stage (Table 4). We found that models including at least 3 vulnerability factors had the best model fit for estimating OR for first-line curative treatment among low-stage patients and for estimating OR for first-line treatment among high-stage patients (with AIC = 6577.6 and AUC= 0.577 and AIC = 6377.5 and AUC= 0.61, respectively) (Table 4). Differences in performance between models both for patients with low and high stages were very small. However, for patients with low stage, the model with the lowest AIC included the following vulnerability factors: age, performance score and comorbidity and for patients with high stage, the model with the lowest AIC included: age, performance score and stage (data not shown).

Table 4. Odds ratio for receiving treatment according to number of vulnerability factors for patients diagnosed with low- and high-stage non-small cell lung cancer.

Vulnerability factors	Low-stage lung cancer patients (I-IIIA)					High-stage non-small cell lung cancer patients (IIIB-IVB)
	Did not receive curative treatment	Received curative treatment	OR (95 % CI)	AIC^a	AUC^b	Did not receive treatment	Received treatment	OR (95 % CI)	AIC^a	AUC^b
At least 2 factors
No	129 (14%)	814 (86%)	1 (reference)			94 (8%)	1024 (92%)	1 (reference)
Yes	1224 (22%)	4388 (78%)	0.57 (0.46–0.69)	6643.680	0.5306	1047 (15%)	5877 (85%)	0.51 (0.41–0.64)	6532.091	0.5330
At least 3 factors
No	538 (16%)	2866 (84%)	1 (reference)			347 (9%)	3618 (91%)	1 (reference)
Yes	815 (26%)	2336 (74%)	0.54 (0.48–0.61)	6577.618	0.5767	794 (19%)	3283 (81%)	0.40 (0.35–0.45)	6377.486	0.6101
At least 4 factors
No	968 (18%)	4483 (82%)	1 (reference)			753 (11%)	5822 (89%)	1 (reference)
Yes	3985 (35%)	719 (65%)	0.40 (0.35–0.47)	6530.765	0.5732	388 (26%)	1079 (74%)	0.36 (0.31–0.41)	6378.678	0.5918
At least 5 factors
No	1241 (20%)	5071 (80%)	1 (reference)			1025 (13%)	6697 (87%)	1 (reference)
Yes	112 (46%)	131 (54%)	0.29 (0.22–0.37)	6596.674	0.5288	116 (36%)	204 (64%)	0.27 (0.21–0.34)	6470.320	0.5361

Abbreviations: OR: odds ratio; CI: confidence interval.

^aThe Akaike information criterion (AIC) is a mathematical method for evaluating how well a model fits the data it was generated from. The smaller the AIC value, the better the model fit.

^bArea Under the Curve (AUC) areal under Receiver Operating Characteristics (ROC) curve.

Discussion

To our knowledge, this is the first nationwide cohort study to assess the influence of co-occurring patient and disease-related vulnerability factors on the likelihood of receiving first-line NSCLC treatment. We observed that the vulnerable patients at risk for not receiving first-line curative or palliative treatment were characterized by being diagnosed with a high disease stage, comorbidity burden and performance status, being above the age of 80 years, and living alone and far from the hospital. A similar pattern was seen when taken into account EGFR and ALK mutation status. The likelihood of not receiving first-line treatment increased with increasing number of vulnerability factors with age, performance score, stage and comorbidity being most predictive for not receiving first-line treatment. This study provides updated evidence of clinical and patient-related factors that may influence the likelihood of receiving first-line treatment. Interventions aimed at optimizing receipt of treatment should consider these factors to identify the vulnerable patients at risk of not receiving guideline-recommended treatment.

The assessment of the patient’s ability to comply with the recommended treatment is in principle a collaboration between the treating physician and the patient with mutual consideration of patient and disease-related factors. A key clinical component when evaluating the patient’s ability to tolerate the recommended treatment is performance status [18], which also were among the most predictive factors identified in our study for not receiving treatment. Moreover, in agreement with previous studies [4,12,13,27,28], we observed that patients with comorbidities and high age were less likely to receive guideline-recommended treatment, which could be due to coexisting limited physical functioning and thus a higher risk of toxicities. Some physicians may not endorse guideline-recommended treatment for elderly patients, perhaps because they do not expect the gains in survival to outweigh the potential harms to patient morbidity and quality of life [27]. Other patient-specific factors such as patient willingness could explain why elderly patients might decide with their physicians not they engage in cancer treatment [4]. Previous studies form the United States have shown that the proportion of patients receiving guideline-recommended treatment is lower among patients diagnosed with a more advanced disease stage (range 50–72%) compared with low-stage patients (range 69–76%) [12,27]. However, this pattern was not clear in our study where a higher proportion of patients with advanced disease received any oncological treatment (90%). These differences in distributions of given treatment are likely due to dissimilar categorization of stage groups and definitions of curative and palliative treatment across studies. Even though we observed fewer vulnerability factors among patients diagnosed with low-stage disease compared with patients diagnosed in advanced stages, there were still 21% of patients in the low-stage group who did not receive the recommended curative treatment, which may be related to patients ending treatment due to a high burden of toxicities.

In contrast to the majority of previous studies [1,2,5–7], our study did not find a statistically significant association between education and treatment receipt. Still, we know that engaging patients in treatment decisions while balancing considerations of potential benefits and harms can be particularly difficult for patients of low SEP [29], possibly because of limited health literacy. Physicians may be challenged in their communication with patients with low SEP because this patient group may have misconceptions about the disease, potentially resulting in under-treatment and poor treatment adherence [30]. Previous studies have shown that patients with low SEP measured by education and income are less likely to ask questions and be actively involved in the treatment [31,32]. A recent qualitative study among cancer patients (29% lung cancer patients) and health professionals suggested that patients who did not engage actively in consultations did not receive the same degree of nuanced information and opportunities for the best treatment [33]. Thus, communication between patients and healthcare professionals may also partly explain the findings from previous studies [1,2,5–7] that patients with low education are less likely to receive treatment. Supportive strategies may be adapted to communicate treatment information and encourage dialogue with patients with limited health literacy to improve treatment initiation and adherence.

In line with previous studies from Denmark [1,2] and USA [27,34,35], we observed that lung cancer patients who lived alone were less likely to receive first-line treatment compared with patients who were living with a partner. A partner may provide both emotional and practical support to adhere to prescribed treatments and to navigate the healthcare system. Moreover, our study showed that distance to the treating hospital as a structural measure of vulnerability outside the hospital system was associated with less access to treatment for patients diagnosed in a high disease stage. No access to a car was associated with difficulties in reaching a clinic in a study of 406 stage II–IV NSCLC patients in the US, even if at a relatively short distance from the patient’s home [11], suggesting that mode of transport may also influence assess to treatment. In two larger population-based studies of NSCLC patients from the UK, those with more than 55 min distance to the hospital received less surgery compared with those having less than 15 min travel distance [19] and patients living furthest away and in areas with highest levels of deprivation were less likely to receive surgery compared with patients with same travel distance but living in the least deprived areas [19,20]. Overall, these studies support our results suggesting that transportation barriers may be an additional vulnerability factor to consider when addressing disparities in access to lung cancer treatment. Since transport expenses to treatment are covered by the state in Denmark, it may be other than financial barriers that mainly drive our findings such as poor physical functioning or limited mental resources to plan and travel long distances.

The strengths of our study include our large study population of lung cancer patients from a nationwide database with detailed information on clinically assessed performance status and receipt of first-line treatment. This study also has several limitations that should be considered. First, we did not have information on health behavior. However, in an attempt to evaluate the influence of alcohol on receipt of treatment, we identified alcohol-related diseases from discharge codes in the National Patient Registry, which restricted alcohol consumption to only the most severe and very limited events. Second, we only included diagnostic codes on comorbidities that required hospitalization reflecting the most severe cases with complications that are not controlled within primary care. However, patients with less severe comorbidity managed in primary care would be expected to be at higher risk of not receiving recommended treatment than patients without comorbidities at large, thereby biasing our results toward the null.

Conclusions

NSCLC patients at risk for not receiving first-line curative or palliative treatment were characterized by being diagnosed with a high disease stage, comorbidity burden and performance status, above the age of 80 years, living alone and living far from the hospital. An increasing number of vulnerability factors had an inverse association with receipt of first-line treatment with age, performance status, comorbidity and stage being most predictive of no treatment receipt. Clinicians need to consider these factors when determining treatment options for vulnerable lung cancer patients. Interventions to optimize receipt of first-line treatment recommendations for vulnerable lung cancer patients should target the underlying factors that drive both decisions of treatment and treatment adherence to reduce disparities in lung cancer survival.

Ethical approval

In Denmark, approvals from Ethical Committees are not required for register-based studies that do not involve biological samples.

Authors’ contributions

Rikke Langballe, Erik Jakobsen, Maria Iachina, Susanne Oksbjerg Dalton and Pernille Envold Bidstrup contributed to the study conception and design. Data analysis was performed by Maria Iachina. The first draft of the manuscript was written by Rikke Langballe and Pernille Envold Bidstrup and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

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

Additional information

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.