Development of a Transplantation Risk Index in Patients With Mechanical Circulatory Support: A Decision Support Tool.
Johnston LE, Grimm JC, et al.JACC Heart Failure 2016; 4(4): 277-286.
Aims
To develop a risk index for mechanical circulatory support (MCS) patients using data from the United Network for Organ Sharing (UNOS) that accurately predicts 1-year mortality after orthotopic heart transplantation (OHT).
Interventions
A sample of standard transplant analysis and research files provided by UNOS and based on organ procurement and transplantation network data current through December 31, 2013 were randomly divided into an 80% model derivation cohort and a 20% validation cohort. All plausible variables for predicting 1-year mortality were evaluated in the derivation sample.
Participants
6,036 OHT patients aged ≥ 18 years who underwent transplantation after 2000.
Outcomes
The primary outcome measured was 1-year mortality. Secondary measured outcomes were 90-day and 3-year mortality and overall survival time.
Follow-up
1 year
CET Conclusions
This manuscript reports the development of a survival risk index after cardiac transplantation in patients on a mechanical circulatory support device. The authors develop and validate a model using 9 recipient and 4 donor variables, using a randomly split sample from the UNOS database. The model predicts post-transplant survival, with a moderate C-statistic between the derivation and validation cohorts of 0.69. The methodology appears sound, although there is no validation in an external cohort at present. There are some limitations to practical use. As with many of the renal transplantation risk indices, the model incorporates cold ischaemia time, which is not available at the time of decision making. The survival predictions can only be interpreted in light of the likely survival remaining on mechanical support, which is not included in the model. Finally, devices are evolving over time and follow-up with some of the current devices is short, meaning that developments in mechanical support may affect later applicability of the model.
Quality notes
Quality assessment not appropriate
Trial registration
None