John de Heide

Chapter 6 98 chose the PADIT score calculator (14), as this score has been validated in several independent cohorts supporting the generalizability of its use with a C-statistic ranging between 0.63 and 0.76 (3,21,22). Besides the US Health claims database study, the number of patients in these validation cohorts ranged from 1000 to 2675 patients. In our study population (n = 2333), the PADIT score also provided good discriminative ability with a C-statistic of 0.70. Patients with a PADIT score of ≥ 7 comprised 17.4% of our study population and had a 1-year standard-of-care infection rate of 1.23%. This infection rate (≥ 1%) seems to justify an antibiotic envelope based on cost-effectiveness studies (12). 4.3. Clinical implications The different cost-effectiveness studies evaluating incremental cost-effectiveness ratios of the TYRX™ envelope used different costs of the antibacterial envelope depending on the specific country (USA, $669; Germany, €945; Italy, €945; England, £800) (7,12). Currently, there is no reimbursement for the antibacterial envelope in the Netherlands. Almost half of our study population fulfilled the inclusion criteria for WRAP-IT; however, the 1-year standard-of-care infection rate was < 0.5% in this specific cohort. This renders the use of an antibacterial envelope less cost-effective for our patient population based on WRAP-IT inclusion criteria. Restricting the use of antibacterial envelopes to high-risk patients according to the PADIT score (≥ 7) will be more cost-effective in our tertiary center because less than 20% of the patients will require an antibacterial envelope. Prospective randomized data should evaluate whether patient selection for an antibacterial envelope based on a high PADIT score (including clinical and procedural factors) is more cost-effective in comparison to using the eligibility criteria for WRAP-IT (mainly based on type of procedure). 4.4. Study limitations The present study has the known limitations inherent to a retrospective study design. Despite the retrospective study design, the variables needed for the PADIT calculator were readily available from the medical records. Furthermore, the primary endpoint comprised hospitalization for CIED infection which is an event that is usually well documented. The fact that our center is a regional endocarditis and lead extraction tertiary referral center reduces the risk of missing a clinically relevant endpoint. We did not focus on minor CIED infections not requiring hospitalization; this may explain

RkJQdWJsaXNoZXIy MTk4NDMw