Changes in potential cancer indicator reporting in primary-care during the COVID-19 pandemic

Background In March 2020, the World Health Organisation declared COVID-19 a pandemic. Aim To investigate how the pandemic affected presentation to primary-care with features potentially indicating cancer, and explore how reporting differed by patient characteristics and in face-to-face vs. remote consultations. Design and setting Retrospective cohort study in 21 practices in South-West England. Methods Potential cancer indicators were identied using pre-dened code lists for symptoms, signs, test results and diagnoses. Negative binomial regression models compared reporting of potential cancer indicators in April-July 2020 to April-July 2019. Incidence rate ratios (IRRs), 95% condence intervals (CIs) and p-values are reported. Results During April-July 2019 and 2020, respectively, 152,447/344,184 (44%) and 123,775/350,966 (35%) patients consulted, with 44,662/344,184 (13%) and 29,150/350,966 (8%) reporting a potential cancer indicator (IRR=0.65, 95%CI 0.62-0.68, p<0.001). Reduced indicator reporting was stable across ethnicity, deprivation and shielding status, was greater in children (0-4 years IRR=0.50, 95%CI 0.46-0.55, p<0.001; 5-17 years IRR=0.45, 95%CI 0.41-0.49, p<0.001) and males (IRR=0.61, 95%CI 0.58-0.64), and less marked in patients with mental health conditions (IRR=0.75, 95%CI 0.72-0.79, p<0.001). Indicator reporting dropped for GP face-to-face consultations (IRR=0.88, 95%CI 0.80-0.97, p=0.011) and increased for remote consultations (IRR=1.22, 95%CI 1.11-1.34, p<0.001), but despite this, remained lower in remote consulting than face-to-face in April-July 2020. Conclusion Patient consulting in general, and for potential cancer indicators specically, reduced during


Introduction
In March 2020, the World Health Organisation declared COVID-19 a pandemic 1 and the UK went into lockdown; the public were instructed to "Stay at home, protect the NHS, and save lives". 2 In order to reduce contact and the spread of the virus, primary-care providers were advised to provide all consultations remotely, unless a face-to-face consultation was urgently required. 3 Patients at high risk of severe COVID-19 due to age or pre-existing health conditions were advised to 'shield' and avoid all but essential contact.
GP and nurse consulting rates markedly dropped in April-May 2020 during the period of UK lockdown, but had largely returned to normal by July 2020, 4 with the majority of consultations carried out remotely (mostly via telephone). Further, there was a reduction in weekly reported incidence of asthma, intestinal infectious diseases, and upper and acute respiratory tract infections from the start of lockdown. 5 There are fears that reduced consulting during the lockdown period and changes in consultation provision may have adversely impacted on cancer detection. 6 A Cancer Research UK survey suggested GPs perceive they are receiving fewer reports of cancer symptoms, particularly from older people, than before the pandemic. 6 Coupled with the temporary suspension of cancer screening services, 7,8 reduced symptom reporting could result in late diagnoses, increased workload for cancer services, and poorer patient outcomes. 9 Aims: To investigate how the COVID-19 pandemic affected the number of people presenting to primarycare with symptoms, signs, test results or diagnoses which could potentially indicate cancer, and to explore how this reporting differed by patient characteristics and in face-to-face vs. remote consultations.

Design and setting
A retrospective cohort study in 21 primary-care practices in South-West England. Data were extracted for the RAPCI Study (Rapid COVID-19 intelligence to improve primary-care response), a mixed-methods study on the rapid change to remote consulting in the initial months of the pandemic; primary results are reported elsewhere. 4 Data Routinely collected and anonymised data were provided by One Care, the GP federation in Bristol, North Somerset, and South Gloucestershire. All practices use the EMIS electronic medical records system. Data included demographics (age, sex, ethnicity, and deprivation), clinical characteristics (mental health and shielding status), and all consultations and clinical codes associated with consultations, added to the system by clinical staff between February 2019 and July 2020 inclusive. All patients registered in July 2020 were included. For the analyses, April-July 2020 (i.e. the period following UK lockdown) was compared to April-July 2019.

Consultations
Consultations were de ned as an interaction between a patient and a GP, nurse, or paramedic working in general practice. Consultations recorded by administrators or other health care professionals, and any administrative tasks, were excluded. Remote consultations were those completed by telephone, video, or e-consultation; face-to-face consultations were in GP practices or visits to patients' homes (see appendix 1).

Outcomes
Pre-existing code lists 10,11 were used to identify potential cancer indicators associated with a consultation. Indicators were collated from clinical features of undiagnosed cancer (symptoms, signs, abnormal test results or diagnoses) listed in the National Institute for Health and Care Excellence guidance on the recognition and referral of suspected cancer (NG12) 12 , using robust methods 11 (see appendix 1).
Individual potential cancer indicator were categorised by the percentage of patients reporting them in April-July 2019: most commonly (≥0.5%), less commonly (0.1% to <0.5%), rarely (0.02% to <0.1%), and very rarely reported (<0.02%). We separate these because the most commonly reported indicators include symptoms which often indicate minor illness rather than cancer (e.g. cough), whereas the less commonly/rarely reported indicators were more likely to be associated with cancer (e.g. weight loss, lumps and masses).

Explanatory variables
Age (in July 2020) was categorised: 0-4, 5-17, 18-49, 50-69, 70-84, and 85+ years old. Deprivation quintiles were calculated using index of multiple deprivation score (IMD) deciles recorded in patient records, based on Lower Super Output Areas of residence. Ethnicity was derived by mapping descriptions from primary-care records to ve categories: white, Asian, black, mixed, and other (see appendix 1).
Presence of a mental health condition included severe mental illness (de ned according to the Quality and Outcomes Framework rules 13 ), diagnosed depression, or prescribed anti-depressants (excluding tricyclics) in the three months prior to July 2020. Sex and shielding status (as of July 2020) were obtained directly from primary-care records.

Statistical analysis
Number and percentages of patients reporting potential cancer indicators are presented. Further, consultation rates (and percentages with indicators) are reported per 1,000 registered patients. Practice list sizes were based on July 2020 data, and adjusted to account for historic list sizes using NHS digital data 14 ('adjusted list size'; see appendix 1).
Changes in proportions of patients presenting (per practice) with any potential cancer indicator in April-July 2020 compared to April-July 2019 were investigated using negative binomial regression models; incidence rate ratios (IRRs) and 95% con dence intervals (CIs) are reported. Consultation year was tted as a xed effect, GP practice as a random effect, and adjusted practice list size (per level of covariate where appropriate) as the offset. Fixed effects for each categorical variable (age, sex, ethnicity, IMD quintile, mental health status, and shielding status), along with the interaction between each covariate and consultation year, were separately tted to the model; interaction p-values are presented and results only presented separately for each level of a covariate if p<0.05. Model validity was checked using standard methods; outliers which disrupted model t were removed.
To investigate consultation provision (i.e. face-to-face vs. remote consulting), we modelled proportions of consultations with potential cancer indicators. Separate models were tted for GPs and nurses/paramedics consultations. Negative binomial regression models were tted with number of consultations as the outcome, a xed effect for consultation provision and an interaction with year, GP practice as a random effect, and total numbers of consultations per practice per level of consultation provision as the offset.
For individual potential cancer indicators, unadjusted IRRs comparing April-July 2020 to 2019 are presented to help interpretation, but due to large numbers of indicators and issues with multiple testing, no modelling was performed. consultation rates, dramatically dropped in April 2020 (following UK lockdown), and while consultation rates had recovered by July 2020, the proportion of patients who consulted had not. Further, the percentage of patients and consultations reporting potential cancer indicators also dropped in April 2020; both had increased by July 2020, but neither recovered to pre-COVID levels.
Potential cancer indicator reporting differed by consultation provision for both GP and nurse/paramedic consultations (p-values <0.001 and 0.007, respectively;  Table 3). Despite this increase in potential cancer indicator reporting in remote consultations, reporting was still less common than in GP face-to-face consultations in both years ( Table 3).  Table 4).

Summary of ndings
In April 2020, following UK lockdown, the proportion of patients consulting, and those reporting potential cancer indicators, reduced substantially compared to April 2019. By July 2020, although consultation rates had returned to previous levels, the proportion of patients consulting, and the proportion reporting potential cancer indicators remained lower than the previous year. The reduction in indicator reporting was particularly evident in children and males, and less evident in patients with mental health conditions. Among adults, the greatest reduction was in patients aged 50-84. Indicator reporting increased in remote consultations in 2020 compared with 2019, and decreased in face-to-face consultations. Despite this, indicator reporting remained higher in GP face-to-face than remote consultations in 2020, suggesting the signi cant increase in remote consulting during the pandemic could be contributing to the reduced reporting of potential cancer indicators. Of the most commonly reported indicators, chest infections, fever, throat pain, coughs and fatigue reduced most dramatically.

Strengths and limitations
To our knowledge, this is the rst papers to assess the effect of the COVID-19 pandemic on potential cancer symptom reporting in UK primary-care. The analysis is based on a comprehensive list of potential cancer indicators, used in previous research. 10,11 Our analysis included large numbers of patients (>350,000) from a diverse range of backgrounds, and as changes due to the pandemic have affected the whole country, ndings are likely to be generalisable across England. Except for ethnicity, missing patient characteristics data were low. There are several limitations which pertain to the recording of potential cancer indicators. Firstly, some symptoms and signs may have been reported in free text rather than using clinical codes, 15 or may have been recorded in administration notes not associated with consultations. This may have resulted in systematic under-reporting of potential cancer indicators; however, this is likely to be similar in both years so should not affect the comparative ndings. A further limitation was that new SNOMED CT codes added since the changeover from Read codes to SNOMED were not in our pre-de ned lists, an issue which would affect April-July 2020 but not 2019; however, we checked this for common indicators cough and back pain, and found new codes were rarely used.

Comparison with other literature
Few other studies have looked at the content of primary-care consultations following UK lockdown. One analysis of primary-care data from a deprived urban population, found that diagnoses of common conditions decreased substantially between March and May 2020, suggesting patients may have undiagnosed conditions resulting from changes in access post-lockdown. 16 Further, data from the Royal College of General Practitioners research and surveillance centre showed a marked reduction in weekly reported incidence of asthma, intestinal infectious diseases, and upper and acute respiratory tract infections from week 12 of 2020. 5 Our study adds to these ndings for cancer-speci c indicators.
Cancer screening programmes, estimated to account for about 5% of cancer diagnoses, were suspended in the wake of the pandemic, 7 making symptom-based diagnosis more important. UK lockdown could have affected symptom-based reporting via the move to remote consulting, as research suggests telephone and video consultations result in fewer problems being identi ed than face-to-face consultations. 17 Lockdown could also have in uenced patients to only contact primary-care if they thought their problem was serious, which is likely re ected in the reduced proportion of patients consulting found in our study.
Early in the pandemic, GPs predicted patients with well-recognised red-ag symptoms, such as a new lump or rectal bleeding, would continue to present to primary-care, but vaguer cancer symptoms such as fatigue, change in bowel habit, and weight loss might be dismissed by patients as trivial and not presented to primary-care. 7 This was supported by our ndings, which show that common symptoms reduced more substantially than less common ones, and reporting of indicators such as weight loss and lumps were maintained. Many of the more common symptoms (such as chest infections and sore throats) will be attributable to causes such as viral infections, rather than cancer. This is likely the reason for the larger reduction of indicators observed in children, as the closure of schools and nurseries will have reduced the infection rate for viral illnesses. A recent survey from Cancer Research UK suggested GPs are particularly worried about older people not consulting with potential cancer symptoms compared to before the pandemic. 6 This was partly supported by our ndings; among adults, we noted the largest drop in 50 to 84-year olds, but patients aged 85+ had less of a reduction than this group, perhaps indicating the success of increased focus on this groups by GPs during the pandemic. 4 In 2015-2017, over 50% of cancers were in people aged 70+ but only 0.2% in children aged 0-4, 18 so GPs concerns being focused on the older age groups is not surprising.

Implications for policy and practice
Findings suggest that patients are less likely to report potential cancer indicators than before the COVID-19 pandemic, particularly for more common symptoms such as fever and coughs. In the context of repeated lockdowns, it is therefore important that the general-public (particularly men and adults aged 50-84) are advised to still consult with primary-care for persistent symptoms. Further, GPs and nurses should be encouraged to ask more probing questions during remote consulting, as they could miss symptoms which may have previously been picked up from non-verbal cues and possibly a more open discussion face-to-face.

Declarations
West (ARC West) at University Hospitals Bristol and Weston NHS Foundation Trust and One Care. Professor Salisbury is a National Institute for Health Research (NIHR) Senior Investigator. The funder had no role in the study design, collection and analysis of data, or the writing of the manuscript. The views expressed are those of the authors and not necessarily those of, the NIHR or the Department of Health Transparency statement: All authors a rm that the manuscript is an honest, accurate, and transparent account of the study being reported, that no important aspects of the study have been omitted, and that any discrepancies from the study as originally planned have been explained.
Dissemination declaration: Results will not be disseminated to study participants as they were not directly involved or consented into the study and it would therefore not be appropriate.
Contributions of authors: LS, MM, RD, JH, and CS contributed to the conception and design of the study, SP provided the potential cancer indicator code lists and interpretation of these, RL extracted the data, MM lead the project administration, and LS performed the analysis and drafted the manuscript. All authors contributed to the organisation and conduct of the study, the interpretation of study data and results, and critiqued the manuscript for important intellectual content.