Cost Effectiveness of Using Array-CGH for Diagnosing Learning Disability

Objective To undertake a cost-effectiveness analysis of using microan-ay comparative genomic hybridisation (array-CGH) as a first-line test versus as a second-line test for the diagnosis of causal chromosomal abnormalities in patients referred to a NHS clinical genetics service in the UK with idiopathic learning disability, developmental delay and/or congenital anomalies. Methods A cost-effectiveness study was conducted. The perspective is that of a UK NHS clinical genetics service provider (with respect to both costs and outcomes). A cohort of patients (n = 1590) referred for array-CGH testing of undiagnosed learning disability and developmental delay by a single NHS regional clinical genetics service (South East Thames Regional Genetics Service), were split into a before-and-after design where 742 patients had array-CGH as a second-line test (before group comparator intervention) and 848 patients had array-CGH as a first-line test (after group evaluated intervention). The mean costs were calculated from the clinical genetics testing pathway constructed for each patient including the costs of genetic testing undertaken and clinical appointments scheduled. The outcome was the number of diagnoses each intervention produced so that a mean cost-per-diagnosis could be calculated. The cost effectiveness of the two interventions was calculated as an incremental cost-effectiveness ratio to produce an incremental cost-per-diagnosis (in 2013 GBP). Sensitivity analyses were conducted by altering both costs and effects to check the validity of the outcome. Results The incremental mean cost of testing patients using the first-line testing strategy was GBP241.56 (95 % CIs GBP256.93 to GBP226.19) and the incremental mean gain in the percentage diagnoses was 0.39 % (95 % CIs 2.73 to 3.51 %), which equates to an additional 1 diagnosis per 256 patients tested. This cost-effectiveness study comparing these two strategies estimates that arrayCGH first-line testing dominates second-line testing because it was both less costly and as effective. The sensitivity analyses conducted (adjusting both costs and effects) supported the dominance of the first-line testing strategy (i.e. lower cost and as effective). Conclusions The first-line testing strategy was estimated to dominate the second-line testing strategy because it was both less costly and as effective. These findings are relevant to the wider UK NHS clinical genetics service, with two key strengths of this study being the appropriateness of the comparator interventions and the direct applicability of the patient cohort within this study and the wider UK patient population.