Comparing the national cancer waiting times dataset with other linked electronic health records to measure treatment timeliness: A national cohort study of kidney cancer in England

Timely treatment is a vital component of high-quality cancer care. We compared a national purpose-specific dataset of cancer waiting times (CWT) with routinely collected hospital datasets (‘comparator datasets’) as sources of data on the timeliness of kidney cancer treatment in England. 11,937 patients diagnosed with kidney cancer between 1st July 2022 and 31st December 2023 were included, using national linked cancer registry and CWT records. The primary analysis compared CWT records and those in comparator datasets (Hospital Episode Statistics, Systemic Anti-Cancer Therapy, and Radiotherapy datasets) in terms of the number, percentage, and timing of patients receiving treatment. Using CWT records, a secondary analysis assessed the effects of the ‘waiting time adjustment policy’ to account for patient treatment deferral on distributions of waiting times at patient and provider levels. 6971 patients were identified as having received treatment (within 100 days of decision-to-treat) according to either CWT or comparator datasets. Of these patients, 6434 (92.3%) had treatment according to both CWT and comparator datasets, with 5836 patients (90.7% of 6434) identified as receiving treatment on the same day across both dataset groups, demonstrating strong agreement. Using the waiting time adjustment policy, the mean number of days to treatment decreased from 29.2 to 27.3 days at the patient-level and from 24.3 to 22.6 days at the provider-level. This study demonstrates strong agreement in treatment records from a national, purpose-specific, and comparator datasets. Strengths of both data sources can be harnessed in linked data to evaluate treatment timeliness, enabling more detailed assessments of specific treatments and better definition of the start of treatment pathways. • Routine and purpose-specific datasets had 92.7% agreement on treatment record presence. • Records agreed on the exact day of treatment for nine of every ten cancer patients. • Strengths of both data sources can be harnessed with patient-linked data in England. • Linked data facilitates detailed assessment and improvement of cancer pathways.