Purpose: Transcatheter aortic valve implantation (TAVI) use in high-risk and inoperable patients with symptomatic severe aortic stenosis is supported by increasing empirical evidence. Many factors influence its introduction into healthcare systems. We investigated the adoption and diffusion of TAVI since its introduction into the Italian market in 2007, and we identified the potential drivers of uptake and diffusion at hospital and regional levels.
Methods: Our sample consisted of all Italian hospitals that adopted TAVI since its introduction in 2007. We classified the hospitals according to funding status, ownership, type, teaching status, location and presence of a regional budget balance plan (BBP). We estimated correlation coefficients between the number of TAVIs and variables at the regional level. A nonparametric Mann-Whitney test was used to compare TAVI volumes across groups of hospitals. Last, we run several regressions to estimate the determinants of TAVI adoption in 2012 and the variation of its diffusion between 2011 and 2012.
Results: In total, 7,261 patients underwent TAVI in Italy between 2007 and 2012, corresponding to a penetration rate in 2012 of 18.1%. At the regional level, TAVI is strongly positively correlated with DRG-based funding mechanisms. At the hospital level, tariff-funded hospitals outperform capitation-funded ones; a level of tariff greater than 25,000€ favours a higher volume of implants; private hospitals outperform public hospitals; teaching and research hospitals implant more valves than do general hospitals; non-BBP hospitals perform slightly better; and southern hospitals outperform hospitals in northern and central Italy. However, none of these differences is statistically significant. The regression analysis shows that none of the control variables at the hospital level significantly explains the level of TAVI adoption in 2012, except for the presence of a regional BBP that significantly reduces TAVI adoption. The incremental TAVI adoption between 2011 and 2012 is significantly correlated with geographic area and regional health expenditure per capita.
Conclusions: The diffusion of TAVI in Italy is characterised by great variability. Major drivers of TAVI adoption and diffusion seem to be hospital type, ownership and reimbursement schemes. Our findings suggest that – although relevant – reimbursement policies are not the major driver of new technologies' adoption and that other variables would need to be further investigated (e.g., the role of referring clinicians) to better interpret wide variability of TAVI diffusion.