Cost-effectiveness analysis of using the TBX6 -associated congenital scoliosis risk score (TACScore) in genetic diagnosis of congenital scoliosis

BACKGROUND We previously reported a novel clinically distinguishable subtype of congenital scoliosis (CS), namely, TBX6-associated congenital scoliosis (TACS). We further developed the TBX6-associated CS risk score (TACScore), a multivariate phenotype-based model to predict TACS according to the patient's clinical manifestations. In this study, we aimed to evaluate whether using the TACScore as a screening method prior to performing whole-exome sequencing (WES) is more cost-effective than using WES as the first-line genetic test for CS. METHODS We retrospectively collected the molecular data of 416 CS patients in the Deciphering disorders Involving Scoliosis and COmorbidities (DISCO) study. A decision tree was constructed to estimate the cost and the diagnostic time required for the two alternative strategies (TACScore versus WES). Bootstrapping simulations and sensitivity analyses were performed to examine the distributions and robustness of the estimates. The economic evaluation considered both the health care payer and the personal budget perspectives. RESULTS From the health care payer perspective, the strategy of using the TACScore as the primary screening method resulted in an average cost of $1074.2 (95%CI: $1044.8 to $1103.5) and an average diagnostic duration of 38.7d (95%CI: 37.8d to 39.6d) to obtain a molecular diagnosis for each patient. In contrast, the corresponding values were $1169.6 (95%CI: $1166.9 to $1172.2) and 41.4d (95%CI: 41.1d to 41.7d) taking WES as the first-line test (P < 0.001). From the personal budget perspective, patients who were predicted to be positive by the TACScore received a result with an average cost of $715.1 (95%CI: $594.5 to $835.7) and an average diagnostic duration of 30.4d (95%CI: 26.3d to 34.6d). Comparatively, the strategy of WES as the first-line test was estimated to have significantly longer diagnostic time with an average of 44.0d (95%CI: 43.2d to 44.9d), and more expensive with an average of $1193.4 (95%CI: $1185.5 to $1201.3) (P < 0.001). In 100% of the bootstrapping simulations, the TACScore strategy was significantly less costly and more time-saving than WES. The sensitivity analyses revealed that the TACScore strategy remained cost-effective even when the cost per WES decreased to $8.8. CONCLUSIONS This retrospective study provides clinicians with economic evidence to integrate the TACScore into clinical practice. The TACScore can be considered a cost-effective tool when it serves as a screening test prior to performing WES.