Combining simple blood tests to identify primary care patients with unexpected weight loss for cancer investigation: Clinical risk score development, internal validation, and net benefit analysis

Abstract

Background Unexpected weight loss (UWL) is a presenting feature of cancer in primary care. Existing research proposes simple combinations of clinical features (risk factors, symptoms, signs, and blood test data) that, when present, warrant cancer investigation. More complex combinations may modify cancer risk to sufficiently rule-out the need for investigation. We aimed to identify which clinical features can be used together to stratify patients with UWL based on their risk of cancer. Methods and findings We used data from 63,973 adults (age: mean 59 years, standard deviation 21 years; 42% male) to predict cancer in patients with UWL recorded in a large representative United Kingdom primary care electronic health record between January 1, 2000 and December 31, 2012. We derived 3 clinical prediction models using logistic regression and backwards stepwise covariate selection: Sm, symptoms-only model; STm, symptoms and tests model; Tm, tests-only model. Fifty imputations replaced missing data. Estimates of discrimination and calibration were derived using 10-fold internal cross-validation. Simple clinical risk scores are presented for models with the greatest clinical utility in decision curve analysis. The STm and Tm showed improved discrimination (area under the curve ≥ 0.91), calibration, and greater clinical utility than the Sm. The Tm was simplest including age-group, sex, albumin, alkaline phosphatase, liver enzymes, C-reactive protein, haemoglobin, platelets, and total white cell count. A Tm score of 5 balanced ruling-in (sensitivity 84.0%, positive likelihood ratio 5.36) and ruling-out (specificity 84.3%, negative likelihood ratio 0.19) further cancer investigation. A Tm score of 1 prioritised ruling-out (sensitivity 97.5%). At this threshold, 35 people presenting with UWL in primary care would be referred for investigation for each person with cancer referred, and 1,730 people would be spared referral for each person with cancer not referred. Study limitations include using a retrospective routinely collected dataset, a reliance on coding to identify UWL, and missing data for some predictors. Conclusions Our findings suggest that combinations of simple blood test abnormalities could be used to identify patients with UWL who warrant referral for investigation, while people with combinations of normal results could be exempted from referral.

Symptoms and signs of lung cancer prior to diagnosis: Comparative study using electronic health records

“…Confirmation of our findings using datasets from other healthcare systems in the U.S. are needed and could be enhanced by more advanced machine learning modelling to incorporate additional clinical variable including quantitative data such as changes in body weight or results of routinely collected laboratory tests, given emerging evidence for associations between weight loss and minor deviations of hemoglobin or platelet count with incident cancer. 25 Given the low uptake of low dose CT screening for lung cancer in the U.S., our findings provide support for revising current priorities to improve early diagnosis of lung cancer. 26…”

BLOod Test Trend for cancEr Detection (BLOTTED): protocol for an observational and prediction model development study using English primary care electronic health records data

“…The number of proposed predictor variables in this study is 70, expected mean follow-up is 11 years (obtained from a second study using blood tests from CPRD 16 ), 0.046 Cox-Snell R 2 (from our recent model 12 ) and expected 0.9 shrinkage factor. The 6-month event rate was 0.014 in our recent model (908 of 63,973 diagnosed 12 ). We found no study that reported 1-year risk of any cancer in patients visiting primary care, so we used a conservative approach to obtain the 1-year event rate: as the outcome window is doubled from 6 months, we also doubled the event rate to give ~1,800 events.…”

“…Simple clinical scores including age-group, sex, and seven simple primary care blood tests (albumin, alkaline phosphatase, C-reactive protein, haemoglobin, liver enzymes, platelets, and total white cell count) could be used to select patients with unexpected weight loss who do and do not warrant further cancer investigation 12 . Internal validation of these risk scores have shown good discrimination between patients with and without cancer, were well calibrated at the levels of risk that decisions to investigate are made in primary care, and have shown superior clinical utility compared to models including only age, sex, and symptoms 12 . It remains unclear whether these findings are valid in external NHS and primary care datasets from abroad and whether similar scores could be developed for patients with other cancer symptoms.…”

“…However, multivariable cancer-risk prediction models widely accessible to National Health Service (NHS) GPs incorporate multiple symptoms and risk factors but not blood test results 10, 11 . Simple clinical scores including age-group, sex, and seven simple primary care blood tests (albumin, alkaline phosphatase, C-reactive protein, haemoglobin, liver enzymes, platelets, and total white cell count) could be used to select patients with unexpected weight loss who do and do not warrant further cancer investigation 12 . Internal validation of these risk scores have shown good discrimination between patients with and without cancer, were well calibrated at the levels of risk that decisions to investigate are made in primary care, and have shown superior clinical utility compared to models including only age, sex, and symptoms 12 .…”

Cancer Special Issue: Early detection and minimal residual disease

“…This is proof of concept that cfDNA can be leveraged as a biomarker for monitoring treatment response in patients with MPNST. Brian D Nicholson and colleagues demonstrated that risk scores based on combinations of risk factors and routine blood tests can be used to stratify patients with unexpected weight loss based on their risk of cancer [ 7 ]. They found that these combined risk scores showed superior clinical utility–compared to the symptoms-only model–to discriminate between patients with and without cancer.…”