A Systematic Review of the Methodology Employed to Calculate Abdominal Aortic Aneurysm Growth Rate

Abstract

Once detected, abdominal aortic aneurysms (AAA) are surveyed with periodic ultrasound scans until they reach an intervention threshold, based on maximal AAA size. The rate of AAA growth varies greatly between patients. There has been significant research interest in pharmacological interventions to attenuate AAA growth and a lack of good-quality evidence for surveillance scan intervals. However, studies to date have used differing methods of growth rate estimation which have been analysed together in meta-analyses. We questioned the validity of this approach and systematically reviewed the methods currently used for AAA growth estimation in the literature and considered their relative merits and limitations. We reviewed 23 studies that met our inclusion criteria, containing a total of 9769 patients and identified three methods of growth rate estimation in use: (i) a simple distance/time calculation, (ii) linear regression modelling and (iii) linear multilevel modelling. Multilevel modelling had significant advantages over the other two methods as it allowed a linear model to reflect individualized growth rates of each patient. However, all methods in use presumed AAA growth to be linear and this is not necessarily the case. Further work using models which allow for nonlinear (e.g. quadratic) growth patterns is required. Consensus on a standardized model would allow valid pooling of data between centres. The UK National AAA Screening Programme will provide an important data source for ongoing work in this area.