Chambers, Pinkie Pranlal; (2021) The Optimisation of Pre-Chemotherapy Blood Assessments through Prognostic Modelling. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Chambers_10139215_Thesis_sigs_removed.pdf Download (3MB) | Preview

Abstract

Background: Evidence guiding pre-chemotherapy blood assessments would enable accurate patient-planning and support the growing numbers of patients treated with chemotherapy. The aim of this PhD was to guide chemotherapy providers on the appropriate timing of pre-treatment blood assessments and develop a prognostic model to predict dose delays, mitigating the need for multiple assessments. Methods and analysis: A literature review guided retrospective data collection of risk factors for cancer patients receiving chemotherapy from four hospitals in England. Descriptive analysis was used to demonstrate changes in laboratory values of pre-chemotherapy blood tests, specifically neutrophils, when taken at different times. Using multivariable logistic regression, the relationship between potential risk factors and the outcome of a chemotherapy dose-administration delay was determined. Results: The study included 4,604 patients (2,022 breast cancer patients, 1,904 colorectal cancer patients and 678 diffuse large B-cell lymphoma patients) between 1 January 2013 and 1 January 2018. Of these, 616 patients had two neutrophil values within 7 days of treatment. 23% of neutrophils assessed 4-6 days prior to treatment did not meet the required threshold; these were repeated nearer to the treatment time. Among all patients, 628 (14%) experienced a second cycle treatment delay of 7 days or more. Significant variability was noted in the rate of delays at different hospitals ranging from 8% for hospital 4 to 22% for hospital 1 (P<0.005). Fourteen risk factors were pre-selected for the development of the prognostic model and fair predictive performance (concordance index 0.67) with good calibration was found. A net benefit analysis demonstrated the model was most beneficial in predicting patients receiving treatment for colorectal cancer; here the model would have value in 50% of all patients. Conclusions: The use of prognostic modelling offers an alternative to understanding a patient’s likeliness to encounter a dose delay, aiding service providers to plan accordingly and negating the need for inappropriate blood tests.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as CSVJSONXML

Archive Staff Only

View Item