ABSTRACT
BackgroundModelling the long-term effects of disruption of cancer services and minimising any excess cancer mortality due to the Covid-19 pandemic is of great importance. Here we adapted a stage-shift model to inform service planning decisions within NHS Scotland for the Detect Cancer Early tumours, breast, colorectal and lung cancer which represent 46% of all cancers diagnosed in Scotland. Methods & DataLung, colorectal and breast cancer incidence data for years 2017-18 were obtained from Public Health Scotland Cancer Quality Performance Indicators (QPI), to define a baseline scenario. The most current stage-specific 5-year survival data came from 2009-2014 national cancer registry and South East Scotland Cancer Network (SCAN) QPI audit datasets. The Degeling et al., inverse stage-shift model was adapted to estimate changes in stage at diagnosis, excess mortality and life-years lost from delays to diagnosis and treatment due to Covid-19-related health services disruption. Three and 6-month periods of disruption were simulated to demonstrate the model predictions. ResultsApproximately, 1-9% reductions in stage I/II presentations leading up to 2-10% increases in stage III/IV presentations are estimated across the three cancer types. A 6-month period of service disruption is predicted to lead to excess deaths at 5 years of 32.5 (31.1, 33.9) per 1000 cases for lung cancer, 16.5 (7.9, 24.3) for colorectal cancer and 31.6 (28.5, 34.4) for breast cancer. ConclusionsDisruption of cancer diagnostic services can lead to significant excess deaths in following years. Increasing diagnostic and capacity for cancer services to deal with the backlog of care are needed. Real time monitoring of incidence and referral patterns over the disruption and post-disruption period to reduce excess deaths including more rapid incidence data by stage and other key tumour/clinical characteristics at presentation for key cancer cases (on a quarterly basis). Real time monitoring in cancer care and referral patterns should help inform what type of interventions are needed to reduce excess mortality and whether different population subgroups require public health messaging campaigns. Specific mitigation measures can be the subject of additional modelling analysis to assess the benefits and inform service planning decision making.
ABSTRACT
ObjectivesThe Coronavirus-19 (COVID-19) pandemic continues to impose formidable challenges on healthcare services. The dramatic curtailment of endoscopy and CT colonography capacity has adversely impacted on timely diagnosis of colorectal cancer (CRC). We describe a COVID-adapted pathway rapidly implemented to mitigate risk and maximise cancer diagnosis in patients referred with symptoms of suspected CRC during the pandemic. DesignThe "COVID-adapted pathway" integrated multiple quantitative faecal immunochemical tests (qFIT), to enrich for significant colorectal disease. CT with oral contrast was used to detect gross pathology. Patients reporting high-risk symptoms were triaged to qFIT+CT and the remainder underwent initial qFIT. Prospective data collection comprised referral category, symptoms, blood results, medical history, time to first test, qFIT and CT results. SettingTertiary colorectal unit which manages over 500 cancer patients annually. ParticipantsAll patients referred as urgent suspicious of cancer (USOC) were included. Overall 422 patients (median age 64 years, 220 females) were triaged using this pathway. Main outcome measuresOutcomes comprised cancer detection frequency. ResultsCompared to the same time period (1st April - 31st May) in 2017-2019, we observed a 43% reduction in primary care referrals with suspected CRC (1071 referrals expected reducing to 609). Overall 422 patients (median age 64 years, 220 females) were triaged using this pathway. Most (84{middle dot}6%) were referred as USOC. Of the 422 patients, 202 (47{middle dot}9%) were triaged to CT and qFIT, 211 (50{middle dot}0%) to qFIT only, eight (1{middle dot}9%) to outpatient clinic, and one to colonoscopy. Fifteen (3{middle dot}6%) declined investigation and seven (1{middle dot}7%) were deemed unfit. We detected 13 cancers (3{middle dot}1%); similar to the mean cancer detection rate from all referrals in 2017-2019 (3{middle dot}3%). ConclusionsThe response to the COVID-19 pandemic resulted in a marked reduction in referrals and cessation of key diagnostic services. Although this COVID-adapted pathway mitigated the adverse effects on diagnostic capacity, the overall reduction in expected diagnoses is very substantial. It is clear that the adverse impact of measures taken to constrain the pandemic will lead to many undetected cancers due to the decrease in referrals. Trial registrationNot applicable
ABSTRACT
PurposeThe COVID-19 pandemic posed an unprecedented challenge to healthcare systems around the world. To mitigate the risks of those referred with possible colorectal cancer during the pandemic we implemented a clinical pathway which required a customised data management system for robust operation. Here, we describe the principal concepts and evaluation of the performance of a spreadsheet-based data management system. MethodsA system was developed using Microsoft Excel(R) 2007 aiming to retain the spreadsheets inherent intuitiveness of direct data entry. Data was itemised limiting entry errors. Visual Basic for Applications (VBA) was used to construct a user-friendly interface to enhance efficiency of data entry and segregate the data required for operational tasks. This was done with built-in loop-back data entry. Finally data derivation and analysis was performed to facilitate pathway monitoring. ResultsFor a pathway which required rapid implementation and development of a customised data management system, the use of a spreadsheet was advantageous due to its user-friendly direct data entry capability. Its function was enhanced by UserForm and large data handling by data segregation using VBA macros. Data validation and conditional formatting minimised data entry errors. Computation by the COUNT function facilitated live data monitoring on a dashboard. During the three months the pathway ran for, the system processed 36 nodal data points for each of the included 837 patients. Data monitoring confirmed its accuracy. ConclusionLarge volume data management using a spreadsheet system is possible with appropriate data definition and VBA programmed data segregation. Clinicians regular input and optimisation made the system adaptable for rapid implementation.
