Development and validation of multivariable machine learning algorithms to predict risk of cancer in symptomatic patients referred urgently from primary care: a diagnostic accuracy study

Nonspecific Signs and/or Symptoms of Cancer: A Retrospective, Observational Analysis from a Secondary Care, US Community Oncology Dataset

To help determine the unmet need for improved diagnostic tools to evaluate patients with nonspecific signs and/or symptoms (NSSS) and suspicion of cancer, we examined patient characteristics, diagnostic journey, and cancer incidence of patients with NSSS within The US Oncology Network (The Network), a secondary care community oncology setting. This retrospective, observational cohort study included patients aged ≥40 years with ≥1 NSSS in their problem list at their first visit within The Network (the index date) between 1 January 2016 and 31 December 2020. Patients were followed longitudinally with electronic health record data for initial cancer diagnosis, new noncancer diagnosis, death, end of study observation period, or 12 months, whichever occurred first. Of 103,984 patients eligible for inclusion, 96,722 presented with only 1 NSSS at index date; 6537/103,984 (6.3%) were diagnosed with 1 primary cancer within 12 months after the index date; 3825/6537 (58.5%) with hematologic malignancy, and 2712/6537 (41.5%) with solid tumor. Among patients diagnosed with cancer (n = 6774), the median time to cancer diagnosis after their first visit within The Network was 5.13 weeks. This study provides a real-world perspective on cancer incidence in patients with NSSS referred to a secondary care setting and highlights the unmet need for improved diagnostic tools to improve cancer outcomes.

The National Health Service urgent cancer referral pathway for suspected urological cancers: early economic evaluation of a risk prediction test

OBJECTIVES

In the UK, the number of patients urgently referred for suspected cancer is increasing, and providers are struggling to cope with demand. We explore the potential cost-effectiveness of a new risk prediction test - the PinPoint test - to triage and prioritize patients urgently referred with suspected urological cancers.

METHODS

Two simulation models were developed to reflect the diagnostic pathways for patients with (i) suspected prostate cancer, and (ii) bladder or kidney cancer, comparing the PinPoint test to current practice. An early economic analysis was conducted from a UK National Health Service (NHS) perspective. The primary outcomes were the percentage of individuals seen within 2 weeks and health care costs. An exploratory analysis was conducted to understand the potential impact of the Pinpoint test on quality-adjusted life years gained.

RESULTS

Across both models and applications, the PinPoint test led to more individuals with urological cancer being seen within 2 weeks. Using PinPoint only to prioritize patients led to increased costs overall, whereas using PinPoint to both triage and prioritize patients led to cost savings. The estimated impact on life years gained/lost was very small and highly uncertain.

CONCLUSIONS

Using the PinPoint test to prioritize urgent referrals meant that more individuals with urological cancer were seen within 2 weeks, but at additional cost to the NHS. If used as a triage and prioritization tool, the PinPoint test shortens wait times for referred individuals and is cost saving. More data on the impact of short-term delays to diagnosis on health-related quality of life is needed.

Systems Biology in Cancer Diagnosis Integrating Omics Technologies and Artificial Intelligence to Support Physician Decision Making

Cancer is the second major cause of disease-related death worldwide, and its accurate early diagnosis and therapeutic intervention are fundamental for saving the patient's life. Cancer, as a complex and heterogeneous disorder, results from the disruption and alteration of a wide variety of biological entities, including genes, proteins, mRNAs, miRNAs, and metabolites, that eventually emerge as clinical symptoms. Traditionally, diagnosis is based on clinical examination, blood tests for biomarkers, the histopathology of a biopsy, and imaging (MRI, CT, PET, and US). Additionally, omics biotechnologies help to further characterize the genome, metabolome, microbiome traits of the patient that could have an impact on the prognosis and patient's response to the therapy. The integration of all these data relies on gathering of several experts and may require considerable time, and, unfortunately, it is not without the risk of error in the interpretation and therefore in the decision. Systems biology algorithms exploit Artificial Intelligence (AI) combined with omics technologies to perform a rapid and accurate analysis and integration of patient's big data, and support the physician in making diagnosis and tailoring the most appropriate therapeutic intervention. However, AI is not free from possible diagnostic and prognostic errors in the interpretation of images or biochemical-clinical data. Here, we first describe the methods used by systems biology for combining AI with omics and then discuss the potential, challenges, limitations, and critical issues in using AI in cancer research.

An exploratory assessment of the impact of a novel risk assessment test on breast cancer clinic waiting times and workflow: a discrete event simulation model

Background

Breast cancer clinics across the UK have long been struggling to cope with high demand. Novel risk prediction tools – such as the PinPoint test – could help to reduce unnecessary clinic referrals. Using early data on the expected accuracy of the test, we explore the potential impact of PinPoint on: (a) the percentage of patients meeting the two-week referral target, and (b) the number of clinic ‘overspill’ appointments generated (i.e. patients having to return to the clinic to complete their required investigations).

Methods

A simulation model was built to reflect the annual flow of patients through a single UK clinic. Due to current uncertainty around the exact impact of PinPoint testing on standard care, two primary scenarios were assessed. Scenario 1 assumed complete GP adherence to testing, with only non-referred cancerous cases returning for delayed referral. Scenario 2 assumed GPs would overrule 20% of low-risk results, and that 10% of non-referred non-cancerous cases would also return for delayed referral. A range of sensitivity analyses were conducted to explore the impact of key uncertainties on the model results. Service reconfiguration scenarios, removing individual weekly clinics from the clinic schedule, were also explored.

Results

Under standard care, 66.3% (95% CI: 66.0 to 66.5) of patients met the referral target, with 1,685 (1,648 to 1,722) overspill appointments. Under both PinPoint scenarios, > 98% of patients met the referral target, with overspill appointments reduced to between 727 (707 to 746) [Scenario 1] and 886 (861 to 911) [Scenario 2]. The reduced clinic demand was sufficient to allow removal of one weekly low-capacity clinic [N = 10], and the results were robust to sensitivity analyses.

Conclusion

The findings from this early analysis indicate that risk prediction tools could have the potential to alleviate pressure on cancer clinics, and are expected to have increased utility in the wake of heightened pressures resulting from the COVID-19 pandemic. Further research is required to validate these findings with real world evidence; evaluate the broader clinical and economic impact of the test; and to determine outcomes and risks for patients deemed to be low-risk on the PinPoint test and therefore not initially referred.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-08665-0.