Impact of COVID‐19 on cancer incidence, presentation, diagnosis, treatment and survival in Northern Ireland

The COVID‐19 pandemic had a major impact on cancer patients and services but has been difficult to quantify. We examined how the entire cancer pathway—from incidence, presentation, diagnosis, stage, treatment and survival—was affected in Northern Ireland during April–December 2020 compared to equivalent 2018–2019 periods using retrospective, observational cancer registry data from the Northern Ireland Cancer Registry (NICR). There were 6748 cancer cases in April–December 2020 and an average 7724 patients in April–December 2018–2019. Incident cases decreased by 13% (almost 1000). Significant differences were found across age cohorts and deprivation quintiles, with reductions greatest for younger people (<55 years; 19% decrease) and less deprived (22% decrease). A higher proportion had emergency admission (16%‐to‐20%) with lower proportions diagnosed pathologically (85%‐to‐83%). There was a significant stage shift, with lower proportions of early stage (29%‐to‐25%) and higher late‐stage (21%‐to‐23%). Lower proportions received surgery (41%‐to‐38%) and radiotherapy (24%‐to‐22%) with a higher proportion not receiving treatment (29%‐to‐33%). One‐year observed‐survival decreased from 73.7% to 69.8% and 1‐year net‐survival decreased from 76.1% to 72.9%, with differences driven by five tumours; Lung (40.3%‐to‐35.0%), Head‐and‐Neck (77.4%‐to‐68.4%), Oesophageal (53.5%‐to‐42.3%), Lymphoma (81.1%‐to‐75.2%) and Uterine cancer (87.4%‐to‐80.4%). Our study reveals profound adverse impact of COVID‐19 on the entire cancer patient pathway, with 13% fewer cases, greater emergency admissions and significant stage‐shift from early to more advanced‐stage disease. There was major treatment impact with lower rates of surgery and radiotherapy and higher proportions receiving no treatment. There were significant reductions in 1‐year survival. Our study will support service recovery and protect cancer services in future pandemics or disruptions.


What's new?
Population-level assessments of the impact of the COVID-19 pandemic on the cancer patient pathway, from presentation to survival, are lacking.Here, the authors compared data on cancer incidence, presentation, diagnosis, stage, treatment and survival in Northern Ireland from April through December 2020 to equivalent time periods in 2018 and 2019.Overall, there were 13% fewer cancer cases and more emergency admissions in 2020.A significant shift from early to more advanced stage disease at time of diagnosis was observed, and higher proportions of patients went untreated.The findings could inform efforts to minimize cancer care disruptions in future pandemics.

| INTRODUCTION
On March 11, 2020, the World Health Organization declared coronavirus disease 2019 (COVID-19) a pandemic. 1The COVID-19 pandemic overwhelmed health and cancer care systems and forced radical re-organization and redistribution of healthcare staff, treatment and patient care across the world, including Northern Ireland and the United Kingdom.Many non-essential healthcare services were adapted, suspended or shutdown, with concerns ongoing regarding the impact of the pandemic on the availability and usage of health and cancer care services. 2The pandemic also caused huge societal disruption with multiple lockdowns and widespread restrictions. 3,4ncer remains a leading cause of death and early intervention can improve patient outcomes. 510][11][12][13][14][15][16][17][18][19][20] Recent studies assessing COVID-19 impact in NI reported reductions in pathology-confirmed cancers compared to Scotland and Wales and reduced emergency hospital admissions for Northern Ireland cancer patients. 7,8A recent study found reduced lung, breast and colorectal pathology-confirmed cancers in Northern Ireland and the Netherlands compared to Aotearoa/New Zealand, which reflect differences in disruptions to screening and healthcare services, patient presentations to primary and secondary care and wider societal restrictions. 21ile studies have examined parts of the cancer patient pathway, we are not aware of a population-based study which examines the impact of the COVID-19 pandemic on the entire cancer patient pathway, from presentation to diagnosis through to treatment and finally to survival.Our study utilises pre-pandemic and inter-pandemic data to describe changes in finalised cancer registrations in Northern Ireland, to assess changes by demographic (age, sex) and socioeconomic (deprivation) characteristics, to investigate differences in mode of presentation, basis of diagnosis and stage distribution at diagnosis, to examine changes in treatments received and differences in overall and net survival 1-year following diagnosis.

| METHODS
Data on all cancer cases (excluding non-melanoma skin cancer  22 is the date of first histological or cytological confirmation of malignancy.Cancer data were coded using the International Classification of Diseases (ICD-10), 23 with codes C00-C97 (excluding C44) used to identify relevant cases, and specific cancer types classified using the codes outlined in Table S1.
Staging classifications were based on TNM Version 8. 24 Socioeconomic deprivation was determined using the 2017 Northern Ireland multiple deprivation measure (NIMDM), 25 with patients assigned to deprivation quintiles based on postcode-derived area of residence, 26 ranging from quintile 1 (most deprived) to quintile 5 (least deprived).A very small proportion of records (0.03% in 2018-2020) had an unknown, incomplete or invalid postcode and deprivation quintiles could not be assigned to these areas.Method of admission to hospital was derived from the Patient Administration System (PAS) which includes records of all inpatient hospital admissions in Northern Ireland.Treatment data (surgery, chemotherapy, radiotherapy, hormone therapy) were determined from multiple systems including PAS, Regional Information System for Oncology & Haematology (RISOH), Radiotherapy information system (ARIA) and prescription information from the Business Services Organisation (BSO).

| Evaluating changes over time
To investigate if COVID-19 impacted cancer incidence, the number of cancer cases in Northern Ireland (ex NMSC) were compared for the period April to December 2020 with the average of equivalent periods for April to December 2018-2019, which were used for baseline comparison.The change in distribution over time for specific groups was evaluated using the χ 2 test.Statistical significance is reported for P-values <.05, but note is made of P-values which would not be significant if adjustments for multiple comparisons, using the Bonferroni correction, were applied.

| Survival analysis
Patient survival up to 1-year post diagnosis was evaluated using two measures.Observed survival examines the time between diagnosis and death from any cause, while age-standardised net survival provides an estimate of patient survival which has been adjusted to take account of deaths unrelated to cancer.Observed survival is calculated using the Kaplan-Meier approach with statistically significant differences evaluated using the log-rank test.Net survival is calculated using the Pohar-Perme method, 27 with calculations completed using the stns module in STATA. 28Given that survival from cancer is dependent upon age at diagnosis, net survival estimates are also age-standardised using the direct method with the standard populations suggested by Corazziari et al, 29 but collapsed to four age groups due to the small number of events in the Northern Ireland population for specific age ranges.Observed and net survival values are both accompanied by 95% confidence intervals.
Data on lung, breast, prostate and colorectal cancer and skin melanoma alongside data on tumour sites that had significant survival differences (head and neck, oesophageal, uterine cancer and lymphoma), are shown in Tables 1-4 and Figure 1A-C, with complete data on all cancers reported in Tables S1-S8 and Figures S1-S40.

| RESULTS
Overall, there was a 12.6% (976 patients) decrease in the number of cancer cases in April-December 2020 (6748 patients) compared to the same period in 2018-2019 (7724 patients) (Table 1).

| Patient characteristics and demographics
The number of male cases decreased by 11.0% (3908 in 2018-2019 to 3480 in 2020) and female cases by 14.4% (3816-3268), but this change was not significant (Table 1).The number of cases aged 0 to 54 decreased by 18.6%, from 1359 in 2018-2019 to 1106 in 2020, while the number of cases aged 75 and over decreased by 7.7% (2617-2416) and this change was significant without correction for multiple comparisons (P = .020)(Table 1).The largest proportional decrease in cases (À21.8%)occurred in the least deprived quintile while the smallest decrease (À7.4%) occurred in the most deprived quintile.Consequently, there was a smaller proportion of cases from the least deprived quintile in 2020 compared to 2018-2019 (18.6% vs 20.8%), and a larger proportion from the most deprived quintile (19.4% in 2020 vs 18.3% in 2018-2019), and this change was significant after correction for multiple comparisons (P = .003)(Table 1).

| Hospital admission type and basis of diagnosis
A significant change in the pattern of hospital admission was found with a higher proportion diagnosed following emergency admission with reduced proportion with pathological diagnoses (67.7%-61.4%),and increased proportion with clinical diagnoses (31.9%-37.6%).

| Stage
There was a significant change in case distribution by stage at diagnosis for all cancers overall, with a reduction in proportion with Stage  2).For all other tumour sites, including stomach, pancreatic, liver, gallbladder, uterine, ovarian, cervical, bladder, testicular and thyroid the change in overall case distribution by stage was not significant (Table S6).

| Treatment
There were significant changes in treatments received by cancer patients in the initial pandemic period.There was a significant reduction in the proportion of patients undergoing surgery, from 41.3% to 38.2% (P < .001)and a significant reduction in the proportion receiving radiotherapy, from 24.1% to 22.3% (P = .005).There was an increase in the proportion receiving chemotherapy (25.7%-26.8%),although this difference was not significant, and no significant difference in the proportion receiving hormone therapy (18.9%-18.0%).
The largest change was found in those receiving no treatment, which increased from 29.1% in 2018-2019 to 33.4% in 2020 (P < .001)(Table 1).T A B L E 3 (Continued) These findings were mirrored for certain tumour sites, with significant reduction in the proportion of lung cancer patients receiving surgery (12.5%-8.1%,P < .001)and radiotherapy (33.6%-26.7%,P < .001),no change in the proportion receiving chemotherapy, and a large, significant increase in the proportion not receiving any treatment (46.3%-54.2%,P < .001)(Table 3).There was a significant or hormone therapy (Table S8), and a small but significant increase in the proportion not receiving any treatment (2.5%-3.7%,P = .047).
For colorectal patients there was no significant difference in the proportions receiving surgery, chemotherapy or radiotherapy or in those not receiving any treatment (Table 3).
The most common finding across tumour sites was an increased proportion of patients not receiving any treatment (melanoma [24.6%-39.5%],HAN [12.1%-19.1%]).Conversely for liver cancer there was a significant increase in patients received any treatment compared to previous years (63.3%-51.5%,P = .046).
For some tumour sites, it appears a modal shift occurred; for example for uterine cancer patients a reduced proportion received surgery (79.3%-68.8%,P = .007)while an increased proportion received radiotherapy (from 22.1% to 33.5%, P = .003).A significantly increased proportion of kidney cancer patients received chemotherapy (5.5%-10.6%,P = .027)(Table S8).Other tumour sites such as pancreatic, gallbladder, oesophageal, ovarian, stomach, testicular, bladder and cervical cancer also found no significant difference in the proportions receiving different treatment modalities (Table S8).

| Survival
Significant reductions in short-term survival were found, with

| DISCUSSION
[12][13][30][31][32] Overall, there was a 13% (almost 1000) reduction in new (incident) cancer cases with variations by tumour type, ranging from reductions of 39% for melanoma to 3.6% for stomach cancer and reductions of 7.5% for lung, 11% for female breast, 12% for colorectal and 14.5% for prostate cancer, respectively.In the prepandemic period cancer cases had risen steadily in NI and were projected to continue increasing, so the reduction in cancer cases was unprecedented and shows the profound impact of COVID on cancer diagnoses. 33However, this was less than the 23% reduction in pathologically diagnosed (PD) cancers previously reported for Northern Ireland in 2020 and also less than the 27% reduction in cancer diagnoses during the COVID-19 pandemic reported in a recent meta-analysis of population-based studies. 17,34However, these studies often relied exclusively on cancers identified from pathological samples, which highlights the importance of assessing finalised, complete population-based cancer registrations to obtain a completer and more accurate picture, as in the current study.
Some cancers, namely pancreatic and gallbladder, did not experience a reduction in cases, which may be due to their severity of symptoms, but which also suggest that Hepato-Pancreato-Biliary (HPB) pathways may have maintained during the initial stages of the pandemic.A study in the Netherlands also reported pancreatic cancer incidence in the second quarter of 2020 to be similar to 2017-2019, and indeed reported increased incidence in the fourth quarter of 2020. 35ere were significant differences across age cohorts and deprivation quintiles, with reductions greatest for younger people (<55 years;19% decrease) and those residing in least deprived areas (22% decrease).Younger women were impacted by the pause on population-based breast and cervical cancer screening programs, which are available to women under 55 years in Northern Ireland.
Younger patients may also have had less severe disease symptoms, which may have been easier to ignore during initial periods of draconian COVID restrictions.Furthermore, those of working age experienced huge, sudden lifestyle disruptions, including working from home (WFH) and managing childcare and home-schooling, which may have led to symptoms being ignored and avoidance of healthcare settings.Those from less deprived areas may have strictly followed, and even over-interpreted, stay-at-home advice and avoided attending healthcare settings even when symptoms developed.
The significantly higher proportion with an emergency admission before diagnosis (16% compared to 20%) may have been due to severe restrictions in access to GP and non-emergency HC settings at early stages of pandemic and sudden alterations to how primary care services were delivered with moves to online bookings and phone and video consultations. 36The significantly lower proportions diagnosed pathologically (85%-to-83%) may have been due to restricted access to diagnostics in early stages of the pandemic.Reductions in small biopsy diagnostic procedures and cancer resections were reported in the early pandemic period in an Irish region, suggesting a COVID-19 impact on both diagnostic and therapeutic services. 37ere was significant stage shift to more advanced disease, with lower proportions of cases diagnosed at an early stage (29% compared to 25%) and higher proportions diagnosed at a late-stage (21% compared to 23%).Pauses, interruptions and reduced capacity with breast, bowel and cervical cancer screening programs, which identify early-stage cancers, have been highlighted as a potential factor in delayed diagnosis and reduced detection of early-stage cancers. 30,38wer proportions of early-stage breast cancers, and higher proportions of later stage, have been associated with pandemic-related reduced breast screening volumes. 30In Northern Ireland the Department of Health (DOH) paused routine cancer screening services on April 7, 2020, with breast cancer screening temporarily paused for 4 months. 39,40The increase in later-stage diagnosis may be associated

[
NMSC]) diagnosed in Northern Ireland in 2018-2020 were extracted from NICR; a population-based registry with complete coverage of Northern Ireland's population of 1.91 million people, which uses pathology reports, hospital admissions and death registrations to identify cancers.Date of cancer diagnosis, based upon European Network of Cancer Registries recommendations, with delays in presentation of people with symptoms due to COVID-related healthcare and societal restrictions.Delays in presentation associated with COVID have been reported in a hospital-based study from Barcelona, which suggested their findings of significantly longer intervals between symptoms onset and cancer diagnosis in the early COVID period (median 39 days) compared to the pre-pandemic period (20 days), were a possible factor associated with the shift to later disease stage.An English study which found a reduction of onethird in prostate cancer cases also reported a higher proportion with advanced stage disease (Stage IV: 21.2% vs 17.4%) in 2020 compared to 2019.41Reduced 1-year survival for all cancers overall was driven by five tumours; Lung Head and Neck, Oesophageal, Lymphoma and Uterine cancer.Regarding lung cancer, the similarity and overlap of symptoms with COVID-19 have created particular challenges, with individuals with a new cough and breathlessness advised to isolate and test for T A B L E 1 Number and proportion of all cancer cases diagnosed in April-December of 2018-2020 by sex, age bands, deprivation quintile, basis of diagnosis, stage at diagnosis, admission type and treatment received-comparing between April-December 2020 and April-December 2018-2019.
Observed and net survival for patients with cancer diagnosed in April-December of 2018-2020 by period of diagnosis-comparing between April-December of 2020 and April-December 2018-2019.
Note: The values in bold are shown for the number and proportion of treatment types which were significantly different between the 2 time periods, by tumour site.*P< .05.T A B L E 4