Abstract
Background
Migrants are often diagnosed with cancer at more advanced stages than the background population, resulting in poorer prognosis. The route to cancer diagnosis is associated with the prognosis, but little is known about the diagnostic pathways for migrants. This study aimed to describe the routes to diagnosis (RtD) for migrants in Denmark and compare these with the RtD for patients of Danish origin.
Methods
We conducted a nationwide cohort study based on Danish registry data, including patients aged ≥ 18 years who were diagnosed with incident cancer between 2014 and 2018 (n = 154,647, including 8066 migrants). Using the UN M49 Standard, we categorised the study population into 15 groups based on country or region of origin (defined according to parental birth country or citizenship), including 14 migrant groups and a reference group of Danish origin. Each patient was assigned to one of six specified RtD. Using multinomial logistic regression models, we assessed associations between migrant groups and RtD.
Results
A total of 44.5% of all cancer patients were diagnosed through a primary care cancer patient pathway (CPP), with variations across migrant groups. Compared with patients of Danish origin, a higher risk of diagnosis via an unplanned hospital admission relative to a CPP from primary care was seen for migrants from the former Yugoslavia (relative risk ratio (RRR) 1.30, 95% CI 1.06–1.60), Germany (RRR 1.34, 95% CI 1.12–1.59), Western Asia (RRR 1.61, 95% CI 1.28–2.02), Türkiye (RRR 1.67, 95% CI 1.29–2.15) and Africa (2.02, 95% CI 1.58–2.60).
Conclusions
Some migrant groups experienced higher diagnosis rates via unplanned hospital admission compared with patients of Danish origin, which could impact their prognosis. The results underline the importance of strengthening cancer diagnostic pathways for migrant populations.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12885-026-15845-2.
Keywords: General Practice, Transients and Migrants, Neoplasms, Early Detection of Cancer, Cohort Studies, Denmark
Background
Migration is increasing worldwide, and the World Health Organization estimates that 281 million people currently live as international migrants [1], i.e. individuals residing in a different country than their usual country of residence [2]. The act of migration itself can be considered a social determinant of health [3], as it shapes the access to resources and opportunities in host countries. In this context, migrants often encounter several barriers to accessing healthcare [4]. Barriers that can affect cancer diagnosis include low awareness of cancer symptoms [5–8], language barriers [6, 7, 9], lower screening participation [10, 11] and difficulties accessing primary care services [5, 7, 9, 12]. Moreover, most studies show that migrants and patients with ethnic minority backgrounds have longer diagnostic intervals [13, 14] and are diagnosed at more advanced cancer stages [15–18] compared with the host populations, which may significantly impact the cancer prognosis.
A key predictor of cancer prognosis is tumour stage at diagnosis [19–21]. However, the route to diagnosis is also associated with the prognosis [22–24], with patients diagnosed following an unplanned hospital admission (often referred to as ‘emergency presentation’ in international literature) having the worst prognosis [23–25]. This association is only partly explainable by the level of tumour stage at diagnosis [25]. In Denmark, approximately 45% of all cancers are diagnosed through a fast-track referral to a cancer patient pathway (CPP) from primary care [24]. The one-year overall mortality for patients diagnosed through this route is 16% compared with 53% for patients diagnosed following an unplanned hospital admission [24].
It remains unknown if patients with a migrant background have different routes to diagnosis (RtD) than patients of Danish origin. Hence, we need more knowledge about the diagnostic routes for migrants with cancer in Denmark. We hypothesised that the RtD would vary across migrant groups and that migrants would be less likely to be diagnosed through a CPP initiated in primary care compared with patients of Danish origin. Thus, this study aimed to describe the RtD for migrants in Denmark and compare these with the RtD for patients of Danish origin.
Methods
Design
This nationwide cohort study was based on routinely collected Danish registry data and reported in accordance with the STROBE guidelines for cohort studies [26].
Setting
The population of Denmark comprises approximately six million inhabitants, and 12% are migrants [27]. The Danish healthcare system is funded by tax revenues, which provide free and equal access to most medical services, e.g. general practice and public hospitals [28]. Approximately 99% of the population is listed with a specific general practitioner (GP) [28]. The GP serves as a gatekeeper to secondary care and can refer patients for further diagnostic investigation, including CPPs, which are pre-specified diagnostic routes aimed at ensuring fast and standardised cancer workup. CPPs were implemented in Denmark between 2008 and 2010 and now cover approximately 40 different types of cancer [29]. National screening programmes exist for breast, colorectal and cervical cancer.
Study population
We included all patients registered with an incident malignant cancer diagnosis in the Danish Cancer Registry (DCR) [30] from 1 January 2014 to 31 December 2018 (n = 159,189), excluding non-melanoma skin cancer, thus including: C00-C43 & C45-C97 according to the International Classification of Diseases, 10th revision (ICD-10). Following the exclusion of 4542 patients, the final study population comprised 154,647 patients with cancer, including 8066 migrants (Fig. 1).
Fig. 1.
Flowchart of excluded patients. Male patients with breast cancer (n = 169) were excluded due to small numbers and no screening of this population. For patients registered with cancer on their death certificate only, no other information about a cancer diagnosis could be found in the registers (n = 793), and these patients were excluded due to the limited number of cases. Descendants of migrants (n = 353) were excluded due to low numbers. The low number is likely due to low age and, thereby, low cancer risk
Data collection
The data were collected from national Danish registers, including national clinical cancer databases. All data was linked at the personal level using the unique 10-digit personal registration number assigned to each resident in Denmark at birth or upon immigration [28].
Definition of migrants and country of origin
Data on migrant status (i.e. Danish origin or migrant) and country of origin were obtained from Statistics Denmark [31]. In line with Statistics Denmark, we defined migrants as individuals born outside Denmark, for whom neither parent meets the criteria of being both born in Denmark and holding Danish citizenship at the time of study inclusion. The country of origin was defined as the birth country or citizenship, if known, of the mother (first choice), the father (second choice) or the individual (third choice if no parental information was available).
There is no universally accepted definition of the term migrant, but it is often used to refer to a person who moves from one place to another, regardless of the reasons for moving, their legal status or the length of their stay [2]. In this context, a migrant can also be a refugee, a person whose status is defined by international law [2]. We acknowledge that the terms migrant and refugee are distinct and designate individuals in non-comparable situations. However, in this article, the term migrant will encompass individuals from various backgrounds and with different reasons for moving, such as to work, study or seek asylum under international law.
In our study population, migrants originated from a total of 157 countries. We categorised the study population into 15 groups based on the regions and subregions from the United Nations (UN) M49 standard country or area codes for statistical use [32] and the individual countries from which most migrants in our cancer cohort originated. This resulted in a reference group with patients of Danish origin (Denmark) and the following 14 migrant groups: Norway, Sweden, Poland, Germany, Türkiye, former Yugoslavia (referring to the Socialist Federal Republic of Yugoslavia and its successor states following the dissolution of the federation in the early 1990s), Eastern Europe (except Poland), Northern Europe (except Denmark, Norway and Sweden), Southwestern Europe (except the former Yugoslavia), Western Asia (except Türkiye), Southern Asia, Asia – others, Africa and Others.
Definition of routes to diagnosis
We based our definition of RtD on an algorithm developed in 2021 by Danckert et al. [24], which was formed in a Danish setting. The approach was conceptually similar to the English RtD algorithm first reported in 2012, although it was suitably modified to capture Danish data sources and diagnostic care organisation [24, 33–35]. The algorithm links individual-level healthcare interactions across multiple registers for the three months preceding a cancer diagnosis. Information on cancer diagnoses and diagnosis dates was based on data from the DCR. Information on hospital contacts, inpatient and outpatient visits and CPPs initiated from primary and secondary care was based on data from the Danish National Patient Registry (DNPR) [36], which covers both the public and private sectors. The Danish Breast Cancer Cooperative Group Database [37], the Danish Colorectal Cancer Group Database [38] and the Danish Quality Database for Cervical Screening [39] provided data on screen-detected breast, colorectal and cervical cancers, respectively. As originally, the algorithm assigned each patient to one of eight distinct RtD [24].
For this study, we excluded patients registered with cancer only on their death certificate in the DCR because of limited numbers. Due to small groups and similar prognosis [24], we combined the outpatient and planned hospital admission pathways, resulting in six RtD assigned in the following order: (1) Screening: patients diagnosed with breast, colorectal or cervical cancer through a national screening programme; for patients with cervical cancer, a positive cervix-cytological test taken up to 90 days before diagnosis indicated diagnosis through screening, (2) CPP from primary care: patients referred to a CPP by a healthcare professional in primary care within 90 days before cancer diagnosis, (3) CPP from secondary care: patients referred to a CPP by a healthcare professional in secondary care within 90 days before cancer diagnosis, (4) Unplanned hospital admission: patients admitted acutely to the hospital within 30 days before cancer diagnosis, with no previous routes before this, (5) Planned hospital admission: patients scheduled for a hospital visit for non-cancer reasons within 30 days before the cancer diagnosis, with no previous routes before this. This could be a planned hospitalisation (inpatient) or a planned visit to the hospital specialist clinic (outpatient), (6) Unknown: patients who could not be assigned to any of the other five RtD.
Screening differs from other diagnostic methods as it is offered to an asymptomatic subgroup of the population in certain age groups for specific cancer types (breast, colorectal and cervical cancers). In this study, we examined the screening route separately from all other routes, as done in a previous study [34]. Hence, routes other than screening are referred to as symptomatic routes.
Defining descriptive variables
From the DCR, we obtained information on age at diagnosis (categorised into four groups: 18–49, 50–59, 60–69 and 70 years or older), region of residence (Capital Region, Region Zealand, Region of Southern Denmark, Central Denmark Region and North Denmark Region) and cancer diagnosis (categorised into 12 main groups based on ICD-10).
From Statistics Denmark, we obtained information about sex, cohabitation status and income. Cohabitation status was classified as married/cohabiting or living alone. Income was defined according to the OECD-modified equivalence scale, averaged over the three years preceding the cancer diagnosis and categorised into low (first quintile), middle (second to fourth quintile) and high (fifth quintile) income, following the recommendations from a recent Danish national guideline [40].
From the DNPR, we obtained information on comorbidity for up to 10 years before the cancer diagnosis using the Charlson Comorbidity Index (CCI) [41], excluding cancer. Scores were categorised into none (CCI score: 0), low (CCI score: 1–2) and high (CCI score: ≥ 3).
Statistical analyses
We provided a descriptive overview of all six RtD by assessing the proportion of each route, including 95% confidence intervals (CI), according to country or region of origin. The screening route was analysed separately from symptomatic routes. For symptomatic routes, multinomial logistic regression models were used to assess the RtD for patients according to migrant group compared with the reference group of Danish origin. The results are presented as relative risk ratios (RRRs) with CPP from primary care as the reference route. We present an unadjusted model and a sex- and age-adjusted model. Adjustments for age were done using restricted cubic splines with four knots according to Harrell’s recommended percentiles [42]. Cluster robust standard errors were used to account for the effect of clustering of patients at practice level. For the screening route, we used a logistic regression model to analyse screening vs. symptomatic routes. This analysis was stratified by cancer type and the specific age groups entitled to screening (23–64 years for cervical cancer, 50–69 years for breast cancer and 50–74 years for colorectal cancer). The analyses of cervical and breast cancer were adjusted for age, while the analysis of colorectal cancer was adjusted for sex and age. From the regression analyses, we estimated the predicted probability of being diagnosed through each RtD, based on marginal means computed with covariates at their observed values. We made a sensitivity analysis to compare the predicted probability of presenting in each RtD between migrant groups and the reference group of Danish origin.
Use of artificial intelligence
No generative artificial intelligence (GAI) models were involved in idea generation, study development, data management, analysis or interpretation of results. To check for spelling and grammar, Grammarly Pro and ChatGPT 4.0 were used, with the authors critically reviewing and revising each suggested alteration.
Results
Descriptive data
The study population comprised 48.5% females, and the median age was 68 years. The patient groups varied notably in their characteristics, including sociodemographic variables (Table 1), with migrant groups being on average younger and having lower incomes than patients of Danish origin. The countries that most migrants originated from in the aggregated groups (regions and subregions) are listed in Table 1; e.g., 85% of migrants from Western Asia originated from Iraq, Lebanon, and Syria. The most common RtD was the CPP from primary care, accounting for 44.5% of all cancer cases, whereas unplanned hospital admission was found in 15.6% of cases. The proportion of patients diagnosed through each route varied across migrant groups (Table 2).
Table 1.
Characteristics of included patients (n = 154,647)
| Country or region* of origin | ||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Denmark | Norway | Sweden | Poland | Germany | Türkiye | Former Yugoslavia | Eastern Europe | Northern Europe | SW Europe | Western Asia | Southern Asia | Africa | Asia – others | Others | Total | |||||||||||||||||
| Total, n (%) | 146,581 | (100.0) | 484 | (100.0) | 471 | (100.0) | 464 | (100.0) | 949 | (100.0) | 523 | (100.0) | 779 | (100.0) | 427 | (100.0) | 721 | (100.0) | 580 | (100.0) | 628 | (100.0) | 653 | (100.0) | 474 | (100.0) | 515 | (100.0) | 398 | (100.0) | 154,647 | (100.0) |
| Age, median (IQI) | 68 | (59;76) | 69 | (61;76) | 67 | (61;73) | 59 | (48;68) | 71 | (63;77) | 59 | (50;69) | 61 | (53;69) | 55 | (42;68) | 64 | (53;72) | 66 | (55;73) | 56 | (47;66) | 59 | (50;68) | 59 | (48;68) | 55 | (45;65) | 64 | (49;72) | 68 | (59;75) |
| Age groups, years, n (%) | ||||||||||||||||||||||||||||||||
| 18–49 | 14,569 | (9.9) | 47 | (9.7) | 42 | (8.9) | 128 | (27.6) | 82 | (8.6) | 130 | (24.9) | 144 | (18.5) | 167 | (39.1) | 145 | (20.1) | 107 | (18.4) | 200 | (31.8) | 161 | (24.7) | 126 | (26.6) | 182 | (35.3) | 100 | (25.1) | 16,330 | (10.6) |
| 50–59 | 22,208 | (15.2) | 59 | (12.2) | 60 | (12.7) | 114 | (24.6) | 111 | (11.7) | 136 | (26.0) | 201 | (25.8) | 84 | (19.7) | 119 | (16.5) | 93 | (16.0) | 177 | (28.2) | 172 | (26.3) | 119 | (25.1) | 138 | (26.8) | 58 | (14.6) | 23,849 | (15.4) |
| 60–69 | 42,471 | (29.0) | 141 | (29.1) | 187 | (39.7) | 131 | (28.2) | 223 | (23.5) | 135 | (25.8) | 244 | (31.3) | 89 | (20.8) | 231 | (32.0) | 165 | (28.4) | 136 | (21.7) | 172 | (26.3) | 130 | (27.4) | 109 | (21.2) | 117 | (29.4) | 44,681 | (28.9) |
| 70+ | 67,333 | (45.9) | 237 | (49.0) | 182 | (38.6) | 91 | (19.6) | 533 | (56.2) | 122 | (23.3) | 190 | (24.4) | 87 | (20.4) | 226 | (31.3) | 215 | (37.1) | 115 | (18.3) | 148 | (22.7) | 99 | (20.9) | 86 | (16.7) | 123 | (30.9) | 69,787 | (45.1) |
| Sex, n (%) | ||||||||||||||||||||||||||||||||
| Female | 70,861 | (48.3) | 313 | (64.7) | 287 | (60.9) | 282 | (60.8) | 448 | (47.2) | 230 | (44.0) | 392 | (50.3) | 288 | (67.4) | 352 | (48.8) | 253 | (43.6) | 285 | (45.4) | 322 | (49.3) | 176 | (37.1) | 376 | (73.0) | 189 | (47.5) | 75,054 | (48.5) |
| Male | 75,720 | (51.7) | 171 | (35.3) | 184 | (39.1) | 182 | (39.2) | 501 | (52.8) | 293 | (56.0) | 387 | (49.7) | 139 | (32.6) | 369 | (51.2) | 327 | (56.4) | 343 | (54.6) | 331 | (50.7) | 298 | (62.9) | 139 | (27.0) | 209 | (52.5) | 79,593 | (51.5) |
| Marital status, n (%) | ||||||||||||||||||||||||||||||||
| Married/cohabiting | 84,262 | (57.5) | 256 | (52.9) | 264 | (56.1) | 261 | (56.2) | 540 | (56.9) | 406 | (77.6) | 501 | (64.3) | 239 | (56.0) | 394 | (54.6) | 340 | (58.6) | 413 | (65.8) | 432 | (66.2) | 278 | (58.6) | 327 | (63.5) | 250 | (62.8) | 89,163 | (57.7) |
| Living alone | 62,319 | (42.5) | 228 | (47.1) | 207 | (43.9) | 203 | (43.8) | 409 | (43.1) | 117 | (22.4) | 278 | (35.7) | 188 | (44.0) | 327 | (45.4) | 240 | (41.4) | 215 | (34.2) | 221 | (33.8) | 196 | (41.4) | 188 | (36.5) | 148 | (37.2) | 65,484 | (42.3) |
| Income, n (%) | ||||||||||||||||||||||||||||||||
| Low | 27,582 | (18.8) | 104 | (21.5) | 89 | (18.9) | 150 | (32.3) | 243 | (25.6) | 322 | (61.6) | 354 | (45.4) | 186 | (43.6) | 206 | (28.6) | 172 | (29.7) | 410 | (65.3) | 344 | (52.7) | 264 | (55.7) | 170 | (33.0) | 92 | (23.1) | 30,688 | (19.8) |
| Middle | 89,018 | (60.7) | 273 | (56.4) | 259 | (55.0) | 260 | (56.0) | 550 | (58.0) | 181 | (34.6) | 391 | (50.2) | 179 | (41.9) | 368 | (51.0) | 288 | (49.7) | 198 | (31.5) | 266 | (40.7) | 171 | (36.1) | 273 | (53.0) | 201 | (50.5) | 92,876 | (60.1) |
| High | 29,981 | (20.5) | 107 | (22.1) | 123 | (26.1) | 54 | (11.6) | 156 | (16.4) | 20 | (3.8) | 34 | (4.4) | 62 | (14.5) | 147 | (20.4) | 120 | (20.7) | 20 | (3.2) | 43 | (6.6) | 39 | (8.2) | 72 | (14.0) | 105 | (26.4) | 31,083 | (20.1) |
| Region of residence, n (%) | ||||||||||||||||||||||||||||||||
| North Denmark | 16,200 | (11.0) | 58 | (12.0) | 37 | (7.9) | 35 | (7.5) | 66 | (7.0) | 11 | (2.1) | 38 | (4.9) | 33 | (7.7) | 52 | (7.2) | 29 | (5.0) | 25 | (4.0) | 30 | (4.6) | 21 | (4.4) | 49 | (9.5) | 24 | (6.0) | 16,708 | (10.8) |
| Central Denmark | 32,671 | (22.3) | 83 | (17.1) | 47 | (10.0) | 70 | (15.1) | 153 | (16.1) | 68 | (13.0) | 118 | (15.1) | 95 | (22.2) | 111 | (15.4) | 88 | (15.2) | 140 | (22.3) | 104 | (15.9) | 90 | (19.0) | 80 | (15.5) | 50 | (12.6) | 33,968 | (22.0) |
| Southern Denmark | 33,959 | (23.2) | 77 | (15.9) | 51 | (10.8) | 102 | (22.0) | 368 | (38.8) | 58 | (11.1) | 207 | (26.6) | 61 | (14.3) | 133 | (18.4) | 99 | (17.1) | 122 | (19.4) | 99 | (15.2) | 64 | (13.5) | 96 | (18.6) | 50 | (12.6) | 35,546 | (23.0) |
| Capital | 39,164 | (26.7) | 199 | (41.1) | 257 | (54.6) | 196 | (42.2) | 238 | (25.1) | 288 | (55.1) | 332 | (42.6) | 194 | (45.4) | 330 | (45.8) | 304 | (52.4) | 256 | (40.8) | 371 | (56.8) | 271 | (57.2) | 238 | (46.2) | 235 | (59.0) | 42,873 | (27.7) |
| Zealand | 24,587 | (16.8) | 67 | (13.8) | 79 | (16.8) | 61 | (13.1) | 124 | (13.1) | 98 | (18.7) | 84 | (10.8) | 44 | (10.3) | 95 | (13.2) | 60 | (10.3) | 85 | (13.5) | 49 | (7.5) | 28 | (5.9) | 52 | (10.1) | 39 | (9.8) | 25,552 | (16.5) |
| CCI, n (%) | ||||||||||||||||||||||||||||||||
| None | 97,160 | (66.3) | 311 | (64.3) | 307 | (65.2) | 351 | (75.6) | 594 | (62.6) | 297 | (56.8) | 505 | (64.8) | 329 | (77.0) | 515 | (71.4) | 398 | (68.6) | 402 | (64.0) | 398 | (60.9) | 295 | (62.2) | 405 | (78.6) | 294 | (73.9) | 102,561 | (66.3) |
| Low (CCI 1–2) | 38,843 | (26.5) | 135 | (27.9) | 131 | (27.8) | 85 | (18.3) | 276 | (29.1) | 163 | (31.2) | 199 | (25.5) | 75 | (17.6) | 154 | (21.4) | 149 | (25.7) | 170 | (27.1) | 169 | (25.9) | 116 | (24.5) | 90 | (17.5) | 84 | (21.1) | 40,839 | (26.4) |
| High (CCI ≥ 3) | 10,578 | (7.2) | 38 | (7.9) | 33 | (7.0) | 28 | (6.0) | 79 | (8.3) | 63 | (12.0) | 75 | (9.6) | 23 | (5.4) | 52 | (7.2) | 33 | (5.7) | 56 | (8.9) | 86 | (13.2) | 63 | (13.3) | 20 | (3.9) | 20 | (5.0) | 11,247 | (7.3) |
| Diagnostic group, n (%) | ||||||||||||||||||||||||||||||||
| Stomach + oesophagus | 4640 | (3.2) | 11 | (2.3) | 9 | (1.9) | 19 | (4.1) | 36 | (3.8) | 36 | (6.9) | 36 | (4.6) | 14 | (3.3) | 22 | (3.1) | 21 | (3.6) | 20 | (3.2) | 33 | (5.1) | 15 | (3.2) | 21 | (4.1) | 10 | (2.5) | 4943 | (3.2) |
| Colon | 14,361 | (9.8) | 51 | (10.5) | 35 | (7.4) | 38 | (8.2) | 93 | (9.8) | 34 | (6.5) | 55 | (7.1) | 31 | (7.3) | 58 | (8.0) | 50 | (8.6) | 53 | (8.4) | 47 | (7.2) | 39 | (8.2) | 28 | (5.4) | 28 | (7.0) | 15,001 | (9.7) |
| Rectum | 6894 | (4.7) | 23 | (4.8) | 12 | (2.5) | 17 | (3.7) | 43 | (4.5) | 24 | (4.6) | 31 | (4.0) | 18 | (4.2) | 21 | (2.9) | 30 | (5.2) | 20 | (3.2) | 20 | (3.1) | 12 | (2.5) | 23 | (4.5) | 15 | (3.8) | 7203 | (4.7) |
| Liver + pancreas | 5796 | (4.0) | 14 | (2.9) | 26 | (5.5) | 24 | (5.2) | 45 | (4.7) | 24 | (4.6) | 37 | (4.7) | 12 | (2.8) | 31 | (4.3) | 30 | (5.2) | 25 | (4.0) | 39 | (6.0) | 22 | (4.6) | 31 | (6.0) | 17 | (4.3) | 6173 | (4.0) |
| Lung | 18,275 | (12.5) | 53 | (11.0) | 52 | (11.0) | 47 | (10.1) | 143 | (15.1) | 64 | (12.2) | 116 | (14.9) | 37 | (8.7) | 84 | (11.7) | 68 | (11.7) | 57 | (9.1) | 45 | (6.9) | 51 | (10.8) | 52 | (10.1) | 30 | (7.5) | 19,174 | (12.4) |
| Melanoma | 10,692 | (7.3) | 45 | (9.3) | 40 | (8.5) | 30 | (6.5) | 40 | (4.2) | 5 | (1.0) | 32 | (4.1) | 27 | (6.3) | 56 | (7.8) | 44 | (7.6) | 12 | (1.9) | 4 | (0.6) | 6 | (1.3) | 3 | (0.6) | 25 | (6.3) | 11,061 | (7.2) |
| Breast | 20,633 | (14.1) | 86 | (17.8) | 99 | (21.0) | 84 | (18.1) | 129 | (13.6) | 76 | (14.5) | 134 | (17.2) | 101 | (23.7) | 112 | (15.5) | 90 | (15.5) | 127 | (20.2) | 134 | (20.5) | 61 | (12.9) | 132 | (25.6) | 81 | (20.4) | 22,079 | (14.3) |
| Female genitals | 7613 | (5.2) | 35 | (7.2) | 28 | (5.9) | 37 | (8.0) | 49 | (5.2) | 28 | (5.4) | 64 | (8.2) | 51 | (11.9) | 43 | (6.0) | 16 | (2.8) | 20 | (3.2) | 42 | (6.4) | 21 | (4.4) | 67 | (13.0) | 22 | (5.5) | 8136 | (5.3) |
| Prostate + other male genitals | 21,549 | (14.7) | 45 | (9.3) | 64 | (13.6) | 50 | (10.8) | 126 | (13.3) | 54 | (10.3) | 86 | (11.0) | 19 | (4.4) | 108 | (15.0) | 74 | (12.8) | 72 | (11.5) | 75 | (11.5) | 78 | (16.5) | 19 | (3.7) | 59 | (14.8) | 22,478 | (14.5) |
| Urinary system | 7,673 | (5.2) | 22 | (4.5) | 22 | (4.7) | 21 | (4.5) | 45 | (4.7) | 38 | (7.3) | 41 | (5.3) | 13 | (3.0) | 30 | (4.2) | 25 | (4.3) | 42 | (6.7) | 45 | (6.9) | 27 | (5.7) | 10 | (1.9) | 12 | (3.0) | 8066 | (5.2) |
| Haematolo-gical cancer | 11,199 | (7.6) | 40 | (8.3) | 31 | (6.6) | 31 | (6.7) | 90 | (9.5) | 62 | (11.9) | 50 | (6.4) | 48 | (11.2) | 59 | (8.2) | 52 | (9.0) | 72 | (11.5) | 80 | (12.3) | 58 | (12.2) | 45 | (8.7) | 50 | (12.6) | 11,967 | (7.7) |
| Other | 17,256 | (11.8) | 59 | (12.2) | 53 | (11.3) | 66 | (14.2) | 110 | (11.6) | 78 | (14.9) | 97 | (12.5) | 56 | (13.1) | 97 | (13.5) | 80 | (13.8) | 108 | (17.2) | 89 | (13.6) | 84 | (17.7) | 84 | (16.3) | 49 | (12.3) | 18,366 | (11.9) |
Abbreviations: SW Europe = Southwestern Europe, IQI = Interquartile Interval, CPP = cancer patient pathway, CCI = Charlson Comorbidity Index
*Regions: 85% of patients from Western Asia originated from Iraq, Lebanon and Syria; 77% of patients from Southern Asia originated from Iran, Afghanistan and Pakistan; 27% of patients from Eastern Europe originated from Romania; 49% of patients from Northern Europe originated from the United Kingdom; 58% of patients from Southwestern Europe originated from France, the Netherlands and Italy; 41% of patients from Africa originated from Morocco and Somalia; 67% of patients from ‘Asia - others’ originated from Thailand, Vietnam and the Philippines; 40% of patients from the ‘Other’ group originated from the USA
Table 2.
Proportions of patients presenting in each route to diagnosis according to country or region of origin (n = 154,647), screening included*
| Route to diagnosis | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Screening | CPP from primary care |
CPP from secondary care |
Unplanned hospital admission |
Planned hospital admission |
Unknown | |||||||
| Country or region of origin |
n | (%) | n | (%) | n | (%) | n | (%) | n | (%) | n | (%) |
| Total | 11,888 | (7.7) | 68,839 | (44.5) | 29,600 | (19.1) | 24,147 | (15.6) | 12,538 | (8.1) | 7635 | (4.9) |
| Denmark | 11,304 | (7.7) | 65,443 | (44.6) | 27,930 | (19.1) | 22,879 | (15.6) | 11,797 | (8.0) | 7228 | (4.9) |
| Norway | 43 | (8.9) | 222 | (45.9) | 81 | (16.7) | 65 | (13.4) | 49 | (10.1) | 24 | (5.0) |
| Sweden | 51 | (10.8) | 197 | (41.8) | 82 | (17.4) | 63 | (13.4) | 41 | (8.7) | 37 | (7.9) |
| Poland | 31 | (6.7) | 208 | (44.8) | 98 | (21.1) | 72 | (15.5) | 36 | (7.8) | 19 | (4.1) |
| Germany | 57 | (6.0) | 388 | (40.9) | 191 | (20.1) | 189 | (19.9) | 85 | (9.0) | 39 | (4.1) |
| Türkiye | 31 | (5.9) | 205 | (39.6) | 116 | (22.2) | 95 | (18.2) | 49 | (9.4) | 27 | (5.2) |
| Former Yugoslavia | 57 | (7.3) | 342 | (43.9) | 162 | (20.8) | 128 | (16.4) | 59 | (7.6) | 31 | (4.0) |
| Eastern Europe | 36 | (8.4) | 181 | (42.4) | 89 | (20.8) | 55 | (12.9) | 46 | (10.8) | 20 | (4.7) |
| Northern Europe | 59 | (8.2) | 328 | (45.5) | 122 | (16.9) | 112 | (15.5) | 54 | (7.5) | 46 | (6.4) |
| Southwestern Europe | 42 | (7.2) | 249 | (42.9) | 131 | (22.6) | 78 | (13.4) | 50 | (8.6) | 30 | (5.2) |
| Western Asia | 40 | (6.4) | 249 | (39.6) | 132 | (21.0) | 105 | (16.7) | 70 | (11.1) | 32 | (5.1) |
| Southern Asia | 47 | (7.2) | 271 | (41.5) | 153 | (23.4) | 82 | (12.6) | 69 | (10.6) | 31 | (4.7) |
| Africa | 17 | (3.6) | 175 | (36.9) | 110 | (23.2) | 96 | (20.3) | 45 | (9.5) | 31 | (6.5) |
| Asia - others | 52 | (10.1) | 215 | (41.7) | 117 | (22.7) | 65 | (12.6) | 53 | (10.3) | 13 | (2.5) |
| Others | 21 | (5.3) | 166 | (41.7) | 86 | (21.6) | 63 | (15.8) | 35 | (8.8) | 27 | (6.8) |
Abbreviations:CPP = cancer patient pathway
*This table provides an overview of all routes, including screening. In the analyses, screening and symptomatic routes are analysed separately
Symptomatic routes to diagnosis according to migrant group
Across all migrant groups and the reference group of Danish origin, the CPP from primary care was the most likely RtD for symptomatic cancers. The CPP from secondary care and the unplanned hospital admission were the second and third most likely routes, with variations across the migrant groups.
When adjusted for sex and age, the predicted probability of being diagnosed with cancer through a CPP from primary care was 48.4% (95% CI 48.1–48.8) for patients of Danish origin, 44.0% (95% CI 40.7–47.3) for migrants from Germany, 39.6% (95% CI 35.1–44.1) for migrants from Türkiye, 40.0% (95% CI 36.1–43.8) for migrants from Western Asia, 43.1% (95% CI 39.2–47.0) for migrants from Southern Asia, 36.1% (95% CI 31.8–40.4) for migrants from Africa, and 42.8% (95% CI 37.8–47.8) for migrants from the Other group. Moreover, migrants from Germany, Türkiye, Western Asia and Africa had the highest predicted probability of being diagnosed through an unplanned hospital admission (Fig. 2a and b, Tables A1-A2 in the Appendix).
Fig. 2.
Predicted probability by country/region of origin for being diagnosed with cancer via a CPP from primary care (a) and an unplanned hospital admission (b) based on the unadjusted and age- and sex-adjusted multinomial logistic regression models. In both models, the dashed black line represents the unadjusted predicted probability for patients of Danish origin of being diagnosed through a CPP from primary care or an unplanned hospital admission. Abbreviation: SW Europe = Southwestern Europe
Most migrant groups were more likely than patients of Danish origin to receive their cancer diagnosis from an unplanned hospital admission relative to a CPP from primary care (Fig. 3). For example, the risk was 34% (RRR 1.34, 95% CI 1.12–1.59) higher for migrants from Germany, 67% (RRR 1.67, 95% CI 1.29–2.15) higher for migrants from Türkiye, 30% (RRR 1.30, 95% CI 1.06–1.60) higher for migrants from the former Yugoslavia, 61% (RRR 1.61, 95% CI 1.28–2.02) higher for migrants from Western Asia, and 102% (2.02, 95% CI 1.58–2.60) higher for migrants from Africa. Further, some migrant groups were more likely to receive their cancer diagnosis via a CPP from secondary care or a planned hospital admission compared with patients of Danish origin. For a complete overview, see Table A3 in the Appendix.
Fig. 3.
Forest plot depicting the relative risk ratios (RRRs) with 95% CI for each migrant group compared with the reference group of Danish origin for presenting in each symptomatic route relative to presenting through a CPP from primary care. Adjusted for sex and age
Screening route according to migrant background
For breast cancer, migrants had 30% lower odds of being diagnosed through a screening route compared with patients of Danish origin (odds ratio (OR) 0.70, 95% CI 0.60–0.82) (Table 3). Migrants also had lower odds of being diagnosed with cervical and colorectal cancer through a screening route. However, these estimates were not statistically significant.
Table 3.
Proportions of patients diagnosed via screening and odds ratio for being diagnosed with cervical, breast or colorectal cancer through the screening route compared with symptomatic routes according to migrant background
| Proportion | Unadjusted | Adjusted for age and sex* | |||
|---|---|---|---|---|---|
| % | OR | 95% CI | OR | 95% CI | |
| Cervical cancer | |||||
| Danish origin | 60.9 | 1 | (ref) | 1 | (ref) |
| Migrant | 56.2 | 0.83 | (0.55–1.24) | 0.84 | (0.56–1.25) |
| Breast cancer | |||||
| Danish origin | 57.3 | 1 | (ref) | 1 | (ref) |
| Migrant | 48.0 | 0.69 | (0.59–0.80) | 0.70 | (0.60–0.82) |
| Colorectal cancer | |||||
| Danish origin | 29.4 | 1 | (ref) | 1 | (ref) |
| Migrant | 27.9 | 0.93 | (0.78–1.12) | 0.93 | (0.78–1.12) |
Abbreviations:OR = odds ratio, CI = confidence interval
*Colorectal cancer was adjusted for age and sex. Cervical and breast cancer were adjusted for age
Discussion
Main findings
This register-based nationwide cohort study comprised almost 155,000 cancer patients, including 8066 migrants with cancer. We found that the RtD varied across migrant groups. Migrants from Germany, Türkiye, Western Asia, Africa, Southern Asia and the Other group had a lower predicted probability of being diagnosed via the CPP from primary care compared with patients of Danish origin. Additionally, the first four groups also had a higher predicted probability of presenting via unplanned hospital admission. The RRR estimates revealed a higher risk of being diagnosed through an unplanned hospital admission, a planned hospital admission or a CPP from secondary care relative to the CPP from primary care for several migrant groups compared with patients of Danish origin. Additionally, we saw significantly lower odds of being diagnosed with breast cancer via the screening route among migrants.
Methodological considerations
We used national administrative and health registers covering the entire Danish population, acknowledged for their reliability, completeness, and high data validity [28, 36]. This was a major strength in our study. As the algorithm [24] is based on these registers, misclassification of RtD is considered unlikely. We excluded relatively few observations due to missing data.
We included all international migrants in Denmark, resulting in heterogeneity in the migrant population. To accommodate this, we categorised the migrant population into subgroups by country or region of origin, aiming for detailed and granular classification. However, to ensure sufficient power, most migrants were grouped into broader (sub)regions, which may have masked potential differences in our analyses. Still, with 14 migrant groups, we were able to show meaningful differences. When analysing the screening route, we had to combine all migrant groups due to limited numbers in each group, which prevented us from showing differences between groups in this analysis.
Our aim was to describe group-level associations rather than estimate causal effects between the migrant groups and RtD. Therefore, we refrained from using fully adjusted regression models, which is in line with epidemiological guidance [43]. Thus, we did not adjust for variables such as income and tumour stage, as this could have masked important differences between the groups. However, we adjusted for sex and age, since the patient groups varied notably in age and sex distributions, and these factors influence both diagnostic patterns and cancer risk. This approach enhanced the comparability between migrant groups and the reference group while preserving relevant variations.
Although we did not stratify the results by cancer types, Table 1 suggests some variation in the diagnostic spectrum between the reference group of Danish origin and the migrant groups. This could partly be due to differences in age, but it is also in line with previous research showing considerable variation in cancer types across different migrant or ethnic minority groups [44, 45]. Further research should aim to determine whether the most common cancer types among potentially vulnerable migrant groups are linked to a higher likelihood of unplanned hospital admission.
Comparison with other studies
A recent UK study on ethnic inequalities in routes to diagnosis of cancer found that the majority of patients (36%) were diagnosed through a fast-track referral from the GP (known as the urgent suspected cancer route, the former Two-Week-Wait) or an elective GP referral (23%) [46]. This is similar to the finding in our study, where the majority of all cancer patients (45%) were diagnosed through a CPP from primary care. Martins et al. [46] focused on ethnic groups based on ethnicity information from primary care and hospital records, as opposed to migrant background. Therefore, the results are not directly comparable to ours. However, they found that Black and Other patients were less likely to be diagnosed via screening (5% for both) compared with White (8%), Asian (11%) and Mixed (9%) patients. This is partly in line with our findings, as we saw lower odds of being diagnosed with breast cancer via the screening route in the migrant group compared with patients of Danish origin. Unlike our findings, the UK study found that diagnosis through an emergency route was more common among White (19%) and Other (28%) patients than among Asian (14%), Black (15%) and Mixed (16%) patients.
A US study from 2024 on emergency diagnosis among older adults found that the proportion of emergency department involvement within 30 days before a cancer diagnosis was highest for non-Hispanic Black patients (28%), Native Hawaiian or Other Pacific Islander patients (25%), and American Indian or Alaska Native patients (26%), whereas it was lowest among non-Hispanic White patients (23%) [47]. Another US study from 2022 showed that African Americans made up a larger share of patients diagnosed following an emergency visit within six months before their cancer diagnosis than among those not diagnosed after an emergency visit (12% versus 7%, CI for difference of 5.0% = 2% to 6%) [48]. These findings should be understood in the context of the American healthcare system, which differs markedly from the Danish model. Without a gatekeeping function in primary healthcare, emergency diagnosis may be the only available route for some patients. However, despite universal health coverage and equal access to healthcare in Denmark [28], our study still showed variations in RtD across migrant groups, with several migrant groups having a higher risk of being diagnosed via an unplanned hospital admission relative to a CPP from primary care compared with patients of Danish origin.
A study from New Zealand in 2023 found ethnic disparities in the rate of emergency hospitalisation within 30 days before lung cancer diagnosis, with the Pacific population (62%) and the Māori population (54%) having higher rates than the majority European population (47%) (age-standardised proportions) [49]. Although not directly comparable to our study regarding ethnic groups and cancer type(s), the finding of ethnic disparities in emergency presentation of lung cancer corroborates our results, where several migrant groups had a higher likelihood of being diagnosed through an unplanned hospital admission across all investigated cancer types compared with patients of Danish origin.
Danish and Nordic studies have shown low participation rates in cancer screening programmes among migrants [11, 50–52], which likely results in fewer cancer diagnoses being detected through screening compared to other diagnostic routes. In our study, migrants had reduced odds of being diagnosed with breast, colorectal and cervical cancer through screening compared with patients of Danish origin, although these figures were not statistically significant for colorectal and cervical cancer. However, we analysed migrants as a combined group, and future studies should seek to explore this in more detail.
We know from a previous Danish study that age is associated with the RtD [34]. The authors found that both the oldest and youngest patients were less likely than other age groups to be diagnosed through a CPP from primary care, and that older patients were more likely than younger patients to be diagnosed with cancer following an unplanned hospital admission. The associations between age and RtD were generally similar for men and women. In our study, most migrant groups had a lower median age than the Danish reference group. Once adjusted for sex and age, we saw that several migrant groups had a higher likelihood of being diagnosed through an unplanned hospital admission compared with the Danish reference group. This larger variation after adjustment was expected, since younger age is associated with a lower risk of unplanned hospital admission [34].
The finding that migrants from Germany had a higher predicted probability for diagnosis through an unplanned hospital admission and a lower predicted probability for diagnosis through a CPP from primary care compared with patients of Danish origin was surprising. However, migrants from Germany had a higher median age than all other groups, and the differences narrowed after adjusting for sex and age. Thus, the results could partly be explained by older age and, consequently, a higher risk of unplanned hospital admission among migrants from Germany.
Compared with patients of Danish origin, migrants from Türkiye, Western Asia, and Africa had the highest likelihood of being diagnosed through an unplanned hospital admission relative to a CPP from primary care. Migrants from Western Asia and Africa are primarily refugees from different time periods, whereas migrants from Türkiye constitute a large group of labour migrants in Denmark. Thus, the potentially most vulnerable migrant groups have different geographical origins and distinct migration histories. However, research indicates that ethnic minority groups in general have lower awareness of cancer symptoms and experience difficulties seeking medical help [5], which may partly explain the observed diagnostic pattern. Further, social inequality in cancer outcomes is well-known [53, 54] and should also be considered in relation to our findings.
Generalisability
Due to differences in healthcare organisations and welfare systems across countries, the findings of our study may not be directly applicable to other countries. Further, the recording of ethnicity data varies widely across Europe [55], and migration patterns differ globally and regionally. However, despite these differences, our results supplement studies from other healthcare systems, indicating a universal issue; many migrant groups have a higher risk of being diagnosed with cancer through unplanned hospital admissions. Thus, adapting the description of the RtD and migrant/ethnic groups to other healthcare systems could still be meaningful and relevant.
Conclusion
This nationwide cohort study found that RtD varied across migrant groups. Notably, migrants from Germany, Türkiye, Western Asia and Africa were less likely to be diagnosed via a CPP from primary care and more likely to be diagnosed via an unplanned hospital admission compared with patients of Danish origin. For all migrant groups combined, we also found significantly lower odds of being diagnosed via screening for breast cancer. The findings confirm our hypotheses, as we expected the RtD to vary across migrant groups and migrants to be less likely to be diagnosed through a CPP initiated in primary care compared with patients of Danish origin. The findings indicate that migrants, to some extent, are diagnosed via other routes than patients of Danish origin, which may impact the cancer prognosis. The increased likelihood of being diagnosed through unplanned hospital admission warrants further investigation, as well as future research could explore the link between RtD and prognosis for migrant groups in potentially vulnerable situations.
Supplementary Information
Acknowledgements
We thank Statistics Denmark for data storage and the Danish Healthcare Quality Institute for allowing access to data from the Danish Breast Cancer Group, the Danish Colorectal Cancer Database, and the Danish Quality Database for Cervical Cancer Screening. We also thank Professor Marie Louise Nørredam, Danish Research Centre for Migration, Ethnicity and Health, for her valuable insights on categorising migrant groups. Lastly, our thanks go to Kaare Rud Flarup for assistance with data management and to Lone Niedziella for proofreading.
Abbreviations
- AHD
Anne Harbo Dahl
- AMFH
Anne Mette Fløe Hvass
- AZF
Alina Zalounina Falborg
- CCI
Charlson Comorbidity Index
- CI
Confidence interval
- CPP
Cancer patient pathway
- CW
Christian Wejse
- DCR
Danish Cancer Registry
- DNPR
Danish National Patient Registry
- GAI
Generative artificial intelligence
- GDPR
General Data Protection Regulation
- GL
Georgios Lyratzopoulos
- GP
General practitioner
- HJ
Henry Jensen
- ICD-10
International Classification of Diseases, 10th revision
- IQI
Interquartile Interval
- LFV
Line Flytkjær Virgilsen
- OECD
Organisation for Economic Co-operation and Development
- OR
Odds ratio
- PK
Per Kallestrup
- RRR
Relative risk ratio
- RtD
Routes to diagnosis
- UK
United Kingdom
- UN
United Nations
- US
United States
- VPN
Virtual private network
Authors' contributions
AHD, LFV, HJ and PK initially conceived the study idea. AHD was responsible for managing the data, conducting the data analysis and drafting the manuscript. Throughout the project, AHD, LFV, HJ, PK, AZF, CW, AMFH and GL provided valuable feedback and contributed to editing multiple drafts. Ultimately, all authors reviewed and approved the final version of the manuscript.
Funding
The project was funded by the Danish General Practitioners’ Foundation for Education and Development (PLU-Fonden), the Health Foundation (Helsefonden), the Danish Cancer Research Foundation (Dansk Kræftforskningsfond), the Danish Foundation for General Practice (Fonden for Almen Praksis) and the Department of Public Health, Aarhus University.
Data availability
The datasets generated and analysed during the current study are not publicly available. All data are securely stored at Statistics Denmark and can only be accessed by pre-approved researchers who have applied for access to data via a research institution. All access to data is operated through a secure remote-access virtual private network (VPN). Furthermore, Statistics Denmark restricts the reporting of data that could identify individuals. Thus, due to privacy regulations at Statistics Denmark and the Danish Data Protection Act, no data from this project can be shared publicly.
Declarations
Ethics approval and consent to participate
The project is formally registered in the Record of Processing Activities at the Research Unit for General Practice in Aarhus (ID No. 951) in compliance with the Danish Data Protection Act No. 502 of 23 May 2018 [56] and the General Data Protection Regulation (GDPR) of the European Union [57]. As no biomedical interventions were carried out during the study, no approval was required from the Committee on Health Research Ethics of the Central Denmark Region. Statistics Denmark approved the project and the use of the registers. Participation consent from patients was not necessary as the Danish Data Protection Act exempts statistical or scientific studies of significant societal importance.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Refugee and migrant health overview. https://www.who.int/health-topics/refugee-and-migrant-health#tab=tab_1. Accessed 21 January 2025.
- 2.Sironi A, Bauloz C, Emmanuel M. Glossary on Migration. International migration law, No. 34. Geneva: international organization for migration (IOM); 2019. [Google Scholar]
- 3.Castaneda H, Holmes SM, Madrigal DS, Young ME, Beyeler N, Quesada J. Immigration as a social determinant of health. Annu Rev Public Health. 2015;36:375–92. [DOI] [PubMed] [Google Scholar]
- 4.Lebano A, Hamed S, Bradby H, Gil-Salmeron A, Dura-Ferrandis E, Garces-Ferrer J, et al. Migrants’ and refugees’ health status and healthcare in Europe: a scoping literature review. BMC Public Health. 2020;20(1):1039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Niksic M, Rachet B, Warburton FG, Forbes LJL. Ethnic differences in cancer symptom awareness and barriers to seeking medical help in England. Br J Cancer. 2016;115(1):136–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Licqurish S, Phillipson L, Chiang P, Walker J, Walter F, Emery J. Cancer beliefs in ethnic minority populations: a review and meta-synthesis of qualitative studies. Eur J Cancer Care (Engl). 2017;26(1). [DOI] [PubMed]
- 7.Darko N, Millet N, Usman A, Teece L, Moss EL. Exploring the perspectives of underrepresented voices: perceptions and experiences of uterine cancer for black African, Caribbean, black British, and mixed-black women in the UK to develop strategies for early symptom presentation. Gynecol Oncol. 2024;180:132–8. [DOI] [PubMed] [Google Scholar]
- 8.Hvidberg L, Pedersen Anette F, Wulff CN, Vedsted P. Cancer awareness and socio-economic position: results from a population-based study in Denmark. BMC Cancer. 2014;14:581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tatari CR, Andersen B, Brogaard T, Badre-Esfahani SK, Jaafar N, Kirkegaard P. Perceptions about cancer and barriers towards cancer screening among ethnic minority women in a deprived area in Denmark - a qualitative study. BMC Public Health. 2020;20(1):921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Deding U, Henig AS, Salling A, Torp-Pedersen C, Bøggild H. Sociodemographic predictors of participation in colorectal cancer screening. Int J Colorectal Dis. 2017;32(8):1117–24. [DOI] [PubMed] [Google Scholar]
- 11.Hertzum-Larsen R, Kjær SK, Frederiksen K, Thomsen LT. Participation in cervical cancer screening among immigrants and Danish-born women in Denmark. Prev Med. 2019;123:55–64. [DOI] [PubMed] [Google Scholar]
- 12.Jones CE, Maben J, Lucas G, Davies EA, Jack RH, Ream E. Barriers to early diagnosis of symptomatic breast cancer: a qualitative study of Black African, Black Caribbean and White British women living in the UK. BMJ Open. 2015;5(3):e006944. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Martins T, Abel G, Ukoumunne OC, Price S, Lyratzopoulos G, Chinegwundoh F et al. Assessing Ethnic inequalities in diagnostic interval of common cancers: a population-based UK cohort study. Cancers (Basel). 2022;14(13):3085. [DOI] [PMC free article] [PubMed]
- 14.Martins T, Down L, Samuels A, Lavu D, Hamilton W, Abel G et al. Understanding ethnic inequalities in cancer diagnostic intervals: a cohort study of patients presenting suspected cancer symptoms to GPs in England. Br J Gen Pract. 2025;75(754):e333-e40. [DOI] [PMC free article] [PubMed]
- 15.Norredam M, Krasnik A, Pipper C, Keiding N. Differences in stage of disease between migrant women and native Danish women diagnosed with cancer: results from af population-based cohort study. Eur J Cancer Prev. 2008;17:185–90. [DOI] [PubMed] [Google Scholar]
- 16.Fry A, White B, Nagarwalla D, Shelton J, Jack RH. Relationship between ethnicity and stage at diagnosis in England: a national analysis of six cancer sites. BMJ Open. 2023;13(1):e062079. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Askari A, Nachiappan S, Currie A, Latchford A, Stebbing J, Bottle A, et al. The relationship between ethnicity, social deprivation and late presentation of colorectal cancer. Cancer Epidemiol. 2017;47:88–93. [DOI] [PubMed] [Google Scholar]
- 18.Harvey-Sullivan A, Ali S, Dhesi P, Hart J, Painter H, Walter FM et al. Comparing cancer stage at diagnosis between migrants and non-migrants: a meta-analysis. Br J Cancer. 2025;132(2):158-67. [DOI] [PMC free article] [PubMed]
- 19.McPhail S, Johnson S, Greenberg D, Peake M, Rous B. Stage at diagnosis and early mortality from cancer in England. Br J Cancer. 2015;112(Suppl 1):S108–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Maringe C, Walters S, Butler J, Coleman MP, Hacker N, Hanna L, et al. Stage at diagnosis and ovarian cancer survival: evidence from the International Cancer Benchmarking Partnership. Gynecol Oncol. 2012;127(1):75–82. [DOI] [PubMed] [Google Scholar]
- 21.Maringe C, Walters S, Rachet B, Butler J, Fields T, Finan P, et al. Stage at diagnosis and colorectal cancer survival in six high-income countries: a population-based study of patients diagnosed during 2000–2007. Acta Oncol. 2013;52(5):919–32. [DOI] [PubMed] [Google Scholar]
- 22.Jensen H, Torring ML, Vedsted P. Prognostic consequences of implementing cancer patient pathways in Denmark: a comparative cohort study of symptomatic cancer patients in primary care. BMC Cancer. 2017;17(1):627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Elliss-Brookes L, McPhail S, Ives A, Greenslade M, Shelton J, Hiom S, et al. Routes to diagnosis for cancer - determining the patient journey using multiple routine data sets. Br J Cancer. 2012;107(8):1220–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Danckert B, Falborg AZ, Christensen NL, Frederiksen H, Lyratzopoulos G, McPhail S, et al. Routes to diagnosis and the association with the prognosis in patients with cancer - a nationwide register-based cohort study in Denmark. Cancer Epidemiol. 2021;74:101983. [DOI] [PubMed] [Google Scholar]
- 25.McPhail S, Swann R, Johnson SA, Barclay ME, Abd Elkader H, Alvi R, et al. Risk factors and prognostic implications of diagnosis of cancer within 30 days after an emergency hospital admission (emergency presentation): an International Cancer Benchmarking Partnership (ICBP) population-based study. Lancet Oncol. 2022;23(5):587–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.STROBE. Strengthening the reporting of observational studies in epidemiology. https://www.strobe-statement.org/. Accessed 20 August 2025.
- 27.Immigrants in Denmark. 2024. [In Danish: Indvandrere i Danmark 2024]. Available at: www.dst.dk/publ/indvandrereidk
- 28.Schmidt M, Schmidt SAJ, Adelborg K, Sundboll J, Laugesen K, Ehrenstein V, et al. The Danish health care system and epidemiological research: from health care contacts to database records. Clin Epidemiol. 2019;11:563–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Danish Health Authority. Cancer Patient Pathways. [In Danish: Sundhedsstyrelsen. Pakkeforløb for kræft.] https://www.sst.dk/da/Fagperson/Sygdomme-lidelser-og-behandling/Kraeftsygdom/Pakkeforloeb-for-kraeft. Accessed 21 January 2025.
- 30.Gjerstorff ML. The Danish Cancer Registry. Scand J Public Health. 2011;39(7 Suppl):42–5. [DOI] [PubMed] [Google Scholar]
- 31.Denmark S. Statistics Denmark. Immigrants and their descendants. [In Danish: Indvandrere og efterkommere]. https://www.dst.dk/en/Statistik/emner/borgere/befolkning/indvandrere-og-efterkommere. Accessed 01 November 2024.
- 32.UN Methodology. Standard country or area codes for statistical use (M49). https://unstats.un.org/unsd/methodology/m49/. Accessed 1 November 2024.
- 33.Virgilsen LF, Vedsted P, Falborg AZ, Pedersen AF, Prior A, Jensen H. Routes to cancer diagnosis for patients with pre-existing psychiatric disorders: a nationwide register-based cohort study. BMC Cancer. 2022;22(1):472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Danckert B, Christensen NL, Falborg AZ, Frederiksen H, Lyratzopoulos G, McPhail S, et al. Assessing how routes to diagnosis vary by the age of patients with cancer: a nationwide register-based cohort study in Denmark. BMC Cancer. 2022;22(1):906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Virgilsen LF, Falborg AZ, Vedsted P, Prior A, Pedersen AF, Jensen H. Unplanned cancer presentation in patients with psychiatric disorders: a nationwide register-based cohort study in Denmark. Cancer Epidemiol. 2022;81:102293. [DOI] [PubMed] [Google Scholar]
- 36.Schmidt M, Schmidt SA, Sandegaard JL, Ehrenstein V, Pedersen L, Sorensen HT. The Danish national patient registry: a review of content, data quality, and research potential. Clin Epidemiol. 2015;7:449–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Christiansen P, Ejlertsen B, Jensen MB, Mouridsen H. Danish breast cancer cooperative group. Clin Epidemiol. 2016;8:445–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Ingeholm P, Gogenur I, Iversen LH. Danish colorectal cancer group database. Clin Epidemiol. 2016;8:465–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Rygaard C. The Danish quality database for cervical cancer screening. Clin Epidemiol. 2016;8:655–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Hjorth CF, Kjaerulff TM, Thomsen MK, Cronin-Fenton D, Dalton SO, Olsen MH, et al. SEPLINE: socioeconomic position in epidemiological research-a national guideline on Danish registry data. Clin Epidemiol. 2025;17:593–624. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. [DOI] [PubMed] [Google Scholar]
- 42.Harrell, FE. 2015. Regression Modeling Strategies, with Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis. Second edition. New York: Springer. 10.1007/978-3-319-19425-7.
- 43.Lesko CR, Fox MP, Edwards JK. A framework for descriptive epidemiology. Am J Epidemiol. 2022;191(12):2063–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Hjerkind KV, Larsen IK, Aaserud S, Moller B, Ursin G. Cancer incidence in non-immigrants and immigrants in Norway. Acta Oncol. 2020;59(11):1275–83. [DOI] [PubMed] [Google Scholar]
- 45.Delon C, Brown KF, Payne NWS, Kotrotsios Y, Vernon S, Shelton J. Differences in cancer incidence by broad ethnic group in England, 2013–2017. Br J Cancer. 2022;126(12):1765–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Martins T, Abel G, Ukoumunne OC, Mounce LTA, Price S, Lyratzopoulos G et al. Ethnic inequalities in routes to diagnosis of cancer: a population-based UK cohort study. Br J Cancer. 2022;127(5):863-71. [DOI] [PMC free article] [PubMed]
- 47.Thompson CA, Sheridan P, Metwally E, Peacock Hinton S, Mullins MA, Dillon EC et al. Emergency department involvement in the diagnosis of cancer among older adults: a SEER-Medicare study. JNCI Cancer Spectr. 2024;8(3):pkae039. [DOI] [PMC free article] [PubMed]
- 48.Pettit N, Sarmiento E, Kline J. Disparities in outcomes among patients diagnosed with cancer in proximity to an emergency department visit. Sci Rep. 2022;12(1):10667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Gurney J, Davies A, Stanley J, Signal V, Costello S, Dawkins P, et al. Emergency presentation prior to lung cancer diagnosis: a national-level examination of disparities and survival outcomes. Lung Cancer. 2023;179:107174. [DOI] [PubMed] [Google Scholar]
- 50.Njor SH, Heinavaara S, Stefansdottir H, Nygard M, Guethmundsdottir EM, Bhargava S, et al. Differences in mammography screening attendance among non-Western immigrants in Denmark, Finland, Iceland and Norway. Prev Med Rep. 2023;36:102516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Badre-Esfahani S, Larsen MB, Seibaek L, Petersen LK, Blaakaer J, Andersen B. Low attendance by non-native women to human papillomavirus vaccination and cervical cancer screening - a Danish nationwide register-based cohort study. Prev Med Rep. 2020;19:101106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Larsen MB, Mikkelsen EM, Rasmussen M, Friis-Hansen L, Ovesen AU, Rahr HB, et al. Sociodemographic characteristics of nonparticipants in the Danish colorectal cancer screening program: a nationwide cross-sectional study. Clin Epidemiol. 2017;9:345–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Dalton SO, Schüz J, Engholm G, Johansen C, Kjær SK, Steding-Jessen M, et al. Social inequality in incidence of and survival from cancer in a population-based study in Denmark, 1994–2003: Summary of findings. Eur J Cancer. 2008;44(14):2074–85. [DOI] [PubMed] [Google Scholar]
- 54.Vaccarella S, Georges D, Bray F, Ginsburg O, Charvat H, Martikainen P, et al. Socioeconomic inequalities in cancer mortality between and within countries in Europe: a population-based study. Lancet Reg Health Eur. 2023;25:100551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Villarroel N, Davidson E, Pereyra-Zamora P, Krasnik A, Bhopal RS. Heterogeneity/granularity in ethnicity classifications project: the need for refining assessment of health status. Eur J Public Health. 2019;29(2):260–6. [DOI] [PubMed] [Google Scholar]
- 56.Danish Data Protection Act. Act No. 502 of 23 May 2018. https://www.datatilsynet.dk/media/7753/danish-data-protection-act.pdf. Accessed 11 September 2025.
- 57.General Data Protection Regulation GDPR. https://gdpr-info.eu/. Accessed 11 September 2025.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and analysed during the current study are not publicly available. All data are securely stored at Statistics Denmark and can only be accessed by pre-approved researchers who have applied for access to data via a research institution. All access to data is operated through a secure remote-access virtual private network (VPN). Furthermore, Statistics Denmark restricts the reporting of data that could identify individuals. Thus, due to privacy regulations at Statistics Denmark and the Danish Data Protection Act, no data from this project can be shared publicly.