Microsimulation-informed guideline: an example of lung cancer screening in China

Article type
Authors
Li S1, Shi Q2, Ji G3, Zhang J4, Li L4, Cheng Y4, Shao J3, Zhang H1, Liu Y4, He X4, Li W4, Liu D4
1Department of Endocrinology and Metabolism, MAGIC China Centre, West China Hospital of Sichuan University, Chengdu, Sichuan, China
2Department of Endocrinology and Metabolism, MAGIC China Centre, West China Hospital of Sichuan University, Chengdu, Sichuan, China; Faculty of Science and Engineering, University of Groningen, Groningen, Netherland
3Health Management Center, General Practice Medical Center, West China Hospital, Sichuan University, Chengdu, Sichuan, China
4Department of Respiratory and Critical Care Medicine, State Key Laboratory of Respiratory Health and Multimorbidity, Institute of Respiratory Health, Center of Precision Medicine, The Research Units of West China, West China Hospital, Sichuan University, Chengdu, Sichuan, China
Abstract
"Context
Screening for lung cancer, one of the most prevalent malignancies in China, is critical but may introduce additional radiological and psychological risks for a wide range of residents. With diverse incidence patterns with the western world and limited resource in large-scale trials, China calls for native guidelines with novel screening strategies to inform the clinicians the indication of low-dose computerized tomography (LDCT) in Chinese smokers and non-smokers.
Methods
A systematic review of risk calculators of lung cancer without the image data and a microsimulation using the individual data of a large population-based cohort (West China Lung Nodule screening cohort, WCLN) informs the guideline. We validated and compared the discrimination and calibration of all available lung cancer risk calculators in WCLN cohort for the best-performed calculators. Using GRADE Evidence-to-Decision (EtD) framework, the guideline panels will select the best applicable risk calculator for further microsimulation.
With the guideline panel- and patient partner-informed benefits and harms (burdens) of LDCT screening, the microsimulation incorporates a nature history component for the outcomes without screening, and an intervention component for screening strategies. Anchoring the literature reports, the panel informs a range of three input parameters (a.k.a., the risk for screening initiation and the intervals and the terminating time of screening). With a large representative cross-sectional study in China, the panel selected model after recalibration estimates the individual-level incidence of lung cancer and its associated health outcome such as cancer-related death. Individuals with the same population receive simulated screening strategies with different input parameters. With the literature informed evidence and parameters within the guideline time frame (20 years), the simulation informs the panel for their guideline recommendations by estimating the benefits and harms (burdens) of screening strategies with three input parameters.
Conclusion
Rigorously designed microsimulations could be a source of evidence for clinical practice guidelines of diagnosis or public health policy. The certainty rating of simulated evidence remains a challenge in the guideline development.
"