Randomised controlled trials have aimed to assess the effectiveness of stereotactic ablative radiotherapy (SABR) with curative intent versus surgical resection for individuals diagnosed with early-stage non-small cell lung cancer (NSCLC) but have failed to recruit sufficient numbers of patients. Non-randomised studies for early-stage NSCLC have reported mixed outcomes following curative SABR versus surgical resection, but did not fully address confounding by indication. The Surgery Or RadioTherapy for early-stage cancer study (SORT) will assess the comparative effectiveness of SABR with curative intent versus surgical resection for NSCLC with a target trial emulation approach, as this can reduce biases in observational studies that aim to estimate the causal effect of interventions.

The SORT study will use the National Cancer Registry for individuals diagnosed with early-stage NSCLC in England during 2015–2020 (inclusive) who received SABR with curative intent or surgical resection. These data will be linked to Hospital Episode Statistics, National Radiotherapy Data Set and the Systemic Anti-Cancer Therapy dataset to obtain information on clinical and sociodemographic characteristics and the treatment received. This target trial emulation will define study population eligibility criteria and regimens for SABR with curative intent and surgical resection. We will reduce the risk of residual confounding with instrumental variable analyses that will exploit geographical variation across the National Health Service in England in the use of SABR with curative intent versus surgical resection for early-stage NSCLC. The primary outcome will be 3-year all-cause mortality after treatment initiation. Secondary outcomes will include 3-month, 6-month, 12-month and 24-month all-cause and lung-cancer mortality, time to death, numbers of hospitalisations, incremental costs and incremental cost-effectiveness.

Ethical approval was obtained from the London School of Hygiene and Tropical Medicine Research Ethics Committee (reference number 29 717–1). Results will be disseminated to clinicians, patients, policy-makers and researchers.