Methodological challenges when quantifying a mortality reduction with screening for abdominal aortic aneurysm when incidence is falling: a registry study




Poster session 2


Monday 24 October 2016 - 15:30 to 16:00


All authors in correct order:

Johansson M1, Brodersen J2, Siersma V2, Marklund B1, Juhl Jørgensen K3
1 University of Gothenburg, Sweden
2 University of Copenhagen, Denmark
3 Nordic Cochrane Centre, Denmark
Presenting author and contact person

Presenting author:

Minna Johansson

Contact person:

Abstract text
Background: Screening for abdominal aortic aneurysms (AAA) has been implemented in Sweden, the UK and the USA based on a relative reduction in disease-specific mortality of about 50% in randomised trials, which translates into a 0.5% absolute mortality reduction (1) However, these estimates were based on populations with a much higher disease prevalence than today; due to reduced smoking, the incidence of AAA has fallen over 70% (2), resulting in reduced absolute benefit and probably a worsened benefit/harm-ratio (3). Additionally, AAA screening has been claimed to result in reduced mortality from other diseases due to life-style modifications and treatment of cardiovascular risk factors following a AAA-diagnosis. However, these claims are debated (4), and meta-analyses have not shown a significant reduction in total mortality (1).

Objectives: To estimate the effect of organised AAA screening in Sweden on disease-specific and total mortality.

Methods: We are conducting a study based on national Swedish registry data using anonymised, individual patient data for disease-specific AAA mortality and total mortality. The Swedish screening programme was gradually implemented from 2006 to 2015, which makes it possible to compare a screened versus a non-screened cohort.

Results: We will discuss the methodological challenges created by substantially diminishing disease incidence in our on-going register study, and how we have tackled them. Preliminary results will be presented.

Conclusions: The balance of benefits and harms of AAA screening today is unknown. The gradual implementation of AAA screening in Sweden presents a unique possibility for evaluation of the screening programme, but substantially declining disease incidence complicates analyses.

1. Guirguis-Blake JM, et al. Ann Intern Med 2014; 160: 321–9.
2. Darwood R, et al. J Vasc Surg 2012; 56: 8–13.
3. Johansson M, et al. BMJ 2015; 350: h825.
4. Johansson M, et al. Lancet 2016; 387: 308-10.