The online version of this article can be found at:
2011 16: 415Vasc Med
Stephen A Badger, Claire Jones, Jane McClements, Louis L Lau, Ian S Young and Christopher C Patterson
Surveillance strategies according to the rate of growth of small abdominal aortic aneurysms
On behalf of:
Society for Vascular Medicine
can be found at:
Additional services and information for
What is This?
- Nov 29, 2011Version of Record >>
by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013vmj.sagepub.comvmj.sagepub.comvmj.sagepub.comvmj.sagepub.comvmj.sagepub.comvmj.sagepub.comvmj.sagepub.comvmj.sagepub.comDownloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from
16(6) 415 –421
© The Author(s) 2011
Reprints and permission: sagepub.
The role for abdominal aortic aneurysm (AAA) screening
is now irrefutable, adding impetus to national programmes.1
Although screening will prevent rupture of large aneu-
rysms, a large number of small aneurysms would be diag-
nosed. Medical optimization and risk factor modification
are designed to retard the rate of growth and prevent dis-
ease progression. There is an association between initial
size and subsequent rate of growth, although the nature of
this is debateable.2–5 Other inconsistent predictors of pro-
gression include smoking, hypertension, diabetes mellitus,
age, sex and certain classes of drugs.2–8
The Aneurysm Detection and Management (ADAM)
trial and the UK Small Aneurysm Trial (UK SAT) showed
that AAA < 5.5 cm should have ultrasonographic surveil-
lance as follow-up, rather than early surgery.9,10 While the
validity of the results has been questioned, the applicability
of the conclusions in the endovascular era have also been
challenged.11–14 However, the frequency of serial measure-
ments is variable between institutions, with no consistent
evidence to support guidelines.
It could, therefore, be hypothesized that the rate of growth
and associated surveillance intervals could be determined by
risk stratification and aneurysm size. The primary aim of the
study was to calculate the average rate of growth of small
aneurysms. The secondary aims were to identify any risk fac-
tors for increased rate of growth and to determine appropriate
screening intervals, according to the initial size of the AAA.
Patients and methods
The details of all patients who underwent an ultrasound
coded as ‘ultrasound of aorta’ over a 13-month period from
January 2005 to January 2006 were obtained from the
Surveillance strategies according to the
rate of growth of small abdominal aortic
Stephen A Badger1, Claire Jones1, Jane McClements1,
Louis L Lau1, Ian S Young2 and Christopher C Patterson2
The management of small abdominal aortic aneurysms (AAA) is by ultrasound surveillance. The study aimed to calculate
their growth rate, identify risk factors and determine appropriate screening intervals. The local screening programme
and hospital records were used to identify patients with a small (< 5.5 cm) AAA. The dates and maximum diameter of
serial scans of patients with two or more scans were obtained. Patients were subdivided by 0.5 cm increments above
3.0 cm. The rate of growth was calculated by linear regression for each patient using both the absolute measurements
and logarithmically (ln) transformed measurements. The 95th centile of growth rate within each subgroup was used to
estimate the minimum time to grow to 5.5 cm. A total of 252 were included. The mean (± SD) AAA size on the initial
scan was 3.9 (± 0.7) cm. Statin use and initial size were predictive factors for the growth rate. The median rate of growth
increased according to size from 0.075 to 0.432 cm/year for AAA < 3.5 cm and > 5.0 cm, respectively. It also steadily
increased for ln measurements from 0.022 (or 2.2%/year) to 0.078 or (7.8%/year). The minimum time (months) to
reach 5.5 cm was 61, 17, 11 and 5 for AAA < 3.5 cm, 3.5–3.9 cm, 4.0–4.4 cm and 4.5–4.9 cm, respectively. Based on ln
measurements, the times were similar at 60, 17, 10 and 4 months. In conclusion, the rate of growth increased steadily
with AAA size. An aneurysm < 3.5 cm does not require a repeat scan for 5 years, while those measuring 3.5–3.9 cm and
4.0–4.4 cm require a repeat scan after 17 and 11 months.
abdominal aortic aneurysm; rate of growth
1 Vascular and Endovascular Surgery Department, Belfast City Hospital,
2 Centre for Public Health, Queens University Belfast, Belfast, UK
Belfast City Hospital
Belfast BT9 7AB
423971 VMJ16610.1177/1358863X11423971Badger SA et al.Vascular Medicine
Vascular Medicine 16(6)
Belfast City Hospital Department of Radiology. This was
supplemented by a list of patients diagnosed with AAA as a
result of attendance at the screening programme from
August 2004 to August 2006. To ensure complete capture
of patients, the small aneurysm database in the unit was
Inclusion criterion was patients with an aneurysm meas-
uring less than 5.5 cm, while those greater than 5.5 cm were
excluded. The computerized radiology records were
searched for each patient and the results of the follow-up
ultrasound and CT scans were obtained. The rate of growth
was then calculated for all patients with two or more scans,
at least 3 months apart. The overall cohort of patients was
divided into subgroups depending on the initial maximal
aortic diameter: 3.0–3.4 cm, 3.5–3.9 cm, 4.0–4.4 cm, 4.5–
4.9 cm, 5.0–5.4 cm. Exclusion criteria were patients with a
normal aorta, miscoded as an aneurysm; those who had a
large AAA; missing data or patient who had only a single
scan in the radiology records.
Past medical and drug history at the time of initial diag-
nosis was recorded, where available. The risk factors
recorded included smoking (current or former), hyperten-
sion requiring treatment, ischaemic heart disease, diabetes
mellitus, chronic obstructive pulmonary disease, hyperlipi-
daemia and a family history of AAA. The drugs included in
the study analysis were antiplatelets, statins, warfarin and
beta blockers. Although these risk factors and medications
may have changed during the follow-up period, the meas-
urements were standardized to the start of the follow-up.
An estimated glomerular filtration rate was also obtained
from the biochemistry records to quantify any renal impair-
ment; a value of less than 60 ml/min was taken as indicat-
ing renal impairment.
The data were stored in Microsoft Excel and subsequently
analysed by SPSS Version 15 (SPSS Inc., Chicago, IL,
USA). Individual rates of growth were calculated by linear
regression both on the original scale of size measurement
and also on a (natural) logarithmically transformed scale.
The former provided an absolute rate of increase (cm/year).
The latter provided a proportionate rate of increase (%/year),
after anti-logarithmic transformation, which could be
expressed as the percentage increase per year, as such a
proportionate increase of 1.045 was equivalent to a 4.5%
increase per year (%/year). Growth rates were closer to nor-
mally distributed when calculated on the logarithmic scale
than on the original scale. The influence of each factor on
rate of growth, on the logarithmic scale, was assessed ini-
tially by independent samples t-test and one-way analysis
of variance, and confirmed in a two-level random effects
model fitted using the xtmixed command in Stata release 11
(StatCorp College Station, TX, USA). In light of some non-
normality of growth rates, even when calculated on the
logarithmic scale, the 95th centile was used to calculate the
appropriate surveillance intervals in each subgroup defined
by initial AAA size. Taking 5.5 cm as the target size, the
minimum time taken to reach this size for each subgroup
could then be estimated. The 95th, rather than the 50th cen-
tile was used, so that even the most rapid aneurysm growth
would be captured. Thus, the recommended surveillance
intervals should be regarded as a minimum, with each strat-
egy tailored according to the patient’s individual growth
pattern. Finally, the individual patient growth patterns were
graphically examined to validate the calculated surveil-
After the exclusion criteria were applied, 252 patients
formed the study cohort. This comprised of 214 male and
38 female patients. The number of follow-up scans ranged
from 2 to 17 (median 5). The mean (SD) size of AAA on the
initial scan taken at the time of diagnosis was 3.9 (0.7) cm.
The screening programme provided 94 patients, while the
remaining 158 were from the Department of Radiology
The distribution of risk factors at the time of initial diagnosis
is detailed in Table 1. The influence of each risk factor on the
subsequent rate of growth is detailed in Table 1. The only
significant (p < 0.05) predictor of rate of growth was statin
therapy: patients receiving statins having a slower growth
(geometric mean 1.045 or 4.5% per year) than patients who
did not (geometric mean 1.075 or 7.5% per year).
Rate of growth and surveillance intervals
The overall cohort was subcategorized according to initial
AAA diameter on diagnosis (Tables 2 and 3). The number
of patients in each subgroup decreased with increasing size
from 72 in the < 3.5 cm category to 20 in the 5.0+ cm cat-
egory. The mean rate of growth (cm/year), based on the
absolute measurements, increased across the five catego-
ries of initial size from 0.09 cm/year to 0.41 cm/year. The
corresponding geometric mean proportionate rates of
increase from the analysis of logarithmically transformed
data ranged from 1.025 (or 2.5%/year) to 1.075 (or 7.5%/
year) and differed significantly (one-way analysis of vari-
ance, p = 0.002). The distribution of rates of growth is illus-
trated in Figure 1.
The 95th centile of rate of growth was calculated within
each subgroup defined by initial size and used to estimate a
screening interval as shown in Tables 2 and 3. Based on the
absolute measurements, those patients with AAA 3.0–3.4
cm should have a repeat ultrasound scan in 61 months.
Those with AAA 3.5–3.9 cm should have their next scan in
17 months, while those in the next subgroup should be
scanned after 11 months. The patients with AAA 4.5–4.9
cm need a repeat scan in 5 months. Based upon ln rate of
growth, the times were similar at 60, 17, 10 and 4 months,
respectively. Examination of the plots for individuals
showed that no patient breeched the 5.5 cm threshold within
Badger SA et al.
these intervals, thus providing internal validation of the cal-
culations (Figure 2).
Analysis of the correlation structure of the repeated
size measurements revealed that correlations between
measurements declined with elapsed time. The initial size
measurement showed a correlation of 0.79 with measure-
ments taken the following year, but this reduced to 0.45 for
measurements taken 10 or more years later. Consequently,
the unstructured option was selected for the correlation
matrix in the Stata xtmixed command. The multilevel ran-
dom effects analysis showed that a quadratic term in time
was significant when AAA size was analyzed (p < 0.001)
but was not significant when AAA size was analysed on a
logarithmic scale (p = 0.17), indicating that growth rates
were closer to linear when analyzed on a logarithmic scale.
The analysis also confirmed that screen-detected cases
Table 1. The influence of risk factors on the rate of growth in 252 small AAA patients
Prevalence n/N (%) Average growth rate (%/year)b
Age > 70 years
Ischaemic heart disease
Chronic obstructive pulmonary disease
Chronic renal failure (eGFR < 60 ml/min/1.73 m2)
Family history of AAA
aSome factors were recorded only in 150 patients whose clinical details were available.
bDerived from the geometric mean of proportionate increases estimated from regression of logarithmically transformed sizes.
Table 2. Rates of growth derived for each patient by linear regression analysis and summarized according to initial AAA size
together with an estimate of time to attain a size of 5.5 cm assuming a growth rate at the 95th centile
Initial size (cm) n
Growth rate (cm/year)Max initial
Growth to attain
5.5 cm (cm)
Time to attain
5.5 cm (months)
Mean 95% CI95th centile
Table 3. Rates of growth derived for each patient by linear regression analysis of logarithmically transformed size and summarized
according to initial AAA size together with an estimate of time to attain a size of 5.5 cm assuming a growth rate at the 95th centile
Initial size (cm) n
Growth rate (%/year)Max initial
Growth to attain
5.5 cm (ln cm)
Time to attain
5.5 cm (months)
95% CI95th centile
Vascular Medicine 16(6)
were not significantly different in their growth rates (p =
0.16) but that cases treated with statins had significantly
slower growth rates (p = 0.005). Finally, the growth rates
were found to be faster in the aneurysms of largest initial
size (p = 0.004). These results therefore confirmed the find-
ings obtained using the simpler statistical analyses of slopes
The available evidence on the aetiology of AAA shows that
it is multifactorial. Most studies have assessed each risk fac-
tor with regard to disease presence. Recognized predisposing
factors include advanced age, male sex, cigarette smoking,
elevated cholesterol, hypertension and other atherosclerotic
disease manifestations.7,15–17 Less evidence is available for
the risk factors for the rate of growth of AAA. The results of
this study revealed some trends towards independent predic-
tors of the rate of growth that are of clinical importance.
It is interesting to note that smoking did not emerge as
a predictor of rate of growth. Over 90% of AAA patients
have a history of smoking, with half of them continuing to
do so at the time of diagnosis.17 Several studies, including
UK SAT, have demonstrated a correlation thereafter
between the expansion and continued smoking.2,7,18 It has
been previously reported in a subgroup of this present
cohort that smoking was important in AAA formation.19
The lack of continuing association with subsequent growth
would suggest that tobacco is perhaps more important as a
trigger to formation, rather than persistent promotion of
subsequent growth, although only cautious conclusions
can be drawn since almost 40% of patients lacked docu-
mented smoking history. Although current smoking is
associated with an increased growth rate, this does not per-
sist with lifetime measures of exposure, which show no
evidence of a dose-dependent relationship.2 Brady et al.
demonstrated a 15–20% increase in growth rate for current
smokers, but considered this insufficient to warrant more
frequent screening intervals.2 While the present study pro-
vides important new evidence about the role of smoking, it
would be rash to conclude that smoking is not involved
and that smoking cessation advice is not vital in these
patients, particularly when the much larger UK SAT indi-
Statin therapy at the time of AAA diagnosis showed a
definite trend of slower rate of growth in this study. It is
known to retard progression of atherosclerosis, with
improved clinical outcomes. This appears to be as a result
of a reduction in both atherogenic lipoproteins and other
pleiotrophic effects, such as lowering C-reactive protein
levels.20 Interestingly, elevated cholesterol is associated
with the presence of AAA, but not the rate of gro
wth.2,15,16,18,21 Nevertheless, the use of statins do slow the
progression of AAA. This is substantiated by the present
results, where although there was no relationship with ele-
vated lipids (p = 0.14), there was slower growth with the
use of statins (p = 0.005). This trend may have become
stronger if it was possible to compensate for inevitable con-
tamination from commencement of statins during the study
period. Similar results were demonstrated by Schlosser
et al., with statin-reduced growth retardation, independent
of cholesterol levels.22,23 Therefore, statins appear to stabi-
lize the aortic wall by other means apart from cholesterol
No other Framingham cardiovascular risk factors in
this study, as listed in Table 1, emerged as predictors of
growth, as would be expected. Previous studies have
shown diabetes and peripheral vascular disease to retard
the growth rate by up to 30%, further enforcing the
hypothesis that atherosclerosis has only a minor role in
The rate of AAA growth in the UK SAT ranged from 0.1
to 0.61 cm/year, which are not dissimilar to the range shown
by the present study, as stratified by initial size. The size-
related growth rate was shown previously, with 0.21 vs
0.47 cm/year calculated for AAA 3.0–3.9 cm and 4.0–4.9
cm, respectively.6 This more rapid expansion is associated
with reaching 5.0 cm and undergoing surgical repair.
Subgroup analysis in this present study showed a gradation
of growth rates, with rates similar to those reported by oth-
ers.3,6,25 A bimodal distribution of the rate of AAA growth
may exist, helping to explain why about 25% of small AAA
fail to grow at all.24 Within this growth pattern, an increased
rate was shown to be associated with significant clinical
AAA-related events. As a result, the authors recommended
Figure 1. The distribution of rates of growth from linear
regression analysis of logarithmically transformed size
measurements broken into subgroups of initial size and showing
the 95th centile of growth rate in each subgroup.
Badger SA et al.
that the growth rate, as well as aortic diameter, should be
accounted for when determining the screening interval.
Insufficient evidence is available at present to be certain if
early growth rate is an important determinant of later
growth rate and warrants more research. This clearly has
major implications for the service provision and cost-
effectiveness of any screening programme, since the inten-
sity of follow-up scans impacts upon workforce planning.
This postulated growth pattern is not universally accepted,
with staccato growth suggested by some researchers.2,26
The present study has a few weaknesses. First, it is well
recognized that participants in a screening programme
often tend to be more health conscious, with a better risk
profile. Since a large proportion of this study cohort was
derived from the local screening programme, the biasing
effect of these patients may have altered the results of the
growth rates. In addition, the lack of significance for most
risk factors to influence the growth rate may be due to this
subgroup of patients. Second, the retrospective collection
of data is likely to have results in an over-representation of
slow-growing aneurysms, thus altering the calculations,
with the additional problem of missing data.
The optimal interval period between two consecutive
scans is not clear.27 A Danish study into this subject used
a diameter of 50 mm as the endpoint for determining
screening intervals.28 The decisive variable was the initial
diameter, with the resultant different rate of growth. Their
recommendations were that rescanning should be in 4, 2
and 1 year for AAA 3.0–3.4 cm, 3.5–3.9 cm and 4.0–5.0
cm, respectively. These recommendations are in keeping
with those proposed by several groups and may help to
reduce the psychological distress associated with repeated
scanning.29 Cook and Galland also recommended annual
screening for aneurysms < 40 mm, while those greater
than this should be scanned every 6 months.30 Current
guidelines for the UK National Screening Programme are
for a repeat ultrasound in 1 year if < 4.5 cm and in 3
months if greater than this. A recent study of 1743 patients
demonstrated that if intervals of 36, 24, 12 and 3 months
were adopted for aneurysms of 35, 40, 45 and 50 mm,
respectively, the risk of breeching the 5.5 cm threshold at
re-screening was less than 1%.2 It is, however, important
to distinguish between the clinical threshold for surgery
and the imaging threshold of 5.5 cm. While the small
Figure 2. Individual growth patterns for each patient plotted on a logarithmic vertical scale, in 0.1 cm increments of initial size,
showing a horizontal 5.5 cm size threshold and the estimated screening intervals derived from the data.
Vascular Medicine 16(6)
aneurysm trials indicate that aneurysms greater than this
diameter should be considered for surgery, the growth pat-
tern is gradual and breeching this does not represent a
clinically adverse event. The decision to operate should be
based upon the individual patient, with co-morbidities
taken into consideration as well as size, thus introducing
an inevitable flexibility to the threshold for surgery
according to the clinical situation. Therefore, any guidelines
for screening intervals need to be practical and convenient
for both patient and clinician. Thus, the patients whose
AAA is 4.5 cm or more could be followed up in 3–6
months, as suggested by the UK National Screening
Programme and ADAM study, with the smaller aneurysms
requiring less intensive intervals according to the present
In conclusion, these results have demonstrated an
increasing rate of growth according to initial AAA size,
even after logarithmical transformation. Statin use was the
only variable, which showed retardation of aneurysm
growth. It would therefore be prudent to tailor screening
intervals according to the AAA size and previous growth
The project concept was originally conceived by the late Prof
Chee Soong, whose further input and valuable guidance the
authors sadly miss.
This research received no specific grant from any funding agency
in the public, commercial, or not-for-profit sectors.
1. Thompson SG, Ashton HA, Goa L, Scott RAP. Screening
men for abdominal aortic aneurysm: 10 year mortal-
ity and cost effectiveness results form the randomised
Multicentre Aneurysm Screening Study. Br Med J 2009;
2. Brady AR, Thompson SG, Fowkes FG, Greenhalgh RM,
Powell JT. Abdominal aortic aneurysm expansion: risk fac-
tors and time intervals for surveillance. Circulation 2004;
3. Brown PM, Sobolev B, Zelt DT. Selective management of
abdominal aortic aneurysms smaller than 5.0cm in a prospec-
tive sizing program with gender-specific analysis. J Vasc
Surg 2003; 38: 762–765.
4. Stonebridge PA, Draper T, Kelman J, et al. Growth rate
of infrarenal aortic aneurysms. Eur J Vasc Endovasc Surg
1996; 11: 70–73.
5. Golledge J, Muller J, Coomans D, Walker PJ, Norman PE.
The small abdominal aortic aneurysm. Eur J Vasc Endovasc
Surg 2006; 31: 237–238.
6. Vega de CM, Gomez R, Estallo L, Rodriguez L, Baquer M,
Barba A. Growth rate and associated factors in small abdom-
inal aortic aneurysms. Eur J Vasc Endovasc Surg 2006; 31:
7. Chang JB, Stein TA, Liu JP, Dunn ME. Risk factors associ-
ated with rapid growth of small abdominal aortic aneurysms.
Surgery 1997; 121: 117–122.
8. Biancari F, Mosorin M, Anttila V, Satta J, Juvonen J, Juvonen
T. Ten-year outcome of patients with very small abdominal
aortic aneurysm. Am J Surg 2002; 183: 53–55.
Lederle FA, Wilson SE, Johnson GR, et al. Immediate repair
compared with surveillance of small abdominal aortic aneu-
rysms. N Engl J Med 2002; 346: 1437–1444.
10. UK Small Aneurysm Trial Participants. Mortality results for
randomised controlled trial of early elective surgery or ultra-
sonographic surveillance for small abdominal aortic aneu-
rysms. Lancet 1998; 352: 1649–1655.
11. Branchereau A. Small aortic aneurysms: is evidence evi-
dent? Eur J Vasc Endovasc Surg 2004; 27: 363–365.
12. Ouriel K. The PIVOTAL study: a randomised comparison
of endovascular repair versus surveillance in patients with
smaller abdominal aortic aneurysms. J Vasc Surg 2009; 49:
13. Patel MS, Brown DA, Wilson SE. Relevance of the ADAM
and UK Small Aneurysm Trial data in the age of endovascu-
lar aneurysm repair. Arch Surg 2009; 144: 806–810.
14. Wellborn MB, Yau FS, Modrali JG, et al. Endovascular
repair of small abdominal aortic aneurysms: a paradigm
shift? Vasc Endovasc Surg 2005; 39: 381–391.
15. Lederle FA, Johnston GR, Wilson SE, et al.; Aneurysm
Detection and Management (ADAM) Veterans Affairs
Cooperative Study Group. Prevalence and associations of
abdominal aortic aneurysm detected through screening. Ann
Intern Med 1997; 126: 441–449.
16. Iribarren C, Darvinian JA, Go AS, Firemen BH, Lee CD,
Grey DP. Traditional and novel risk factors for clinically
diagnosed abdominal aortic aneurysm: the Kaiser multipha-
sic health check-up cohort study. Ann Epidemiol 2007; 17:
17. Powell JT, Worrelll P, MacSweeney ST, Franks PJ,
Greenhalgh RM. Smoking as a risk factor for abdominal aor-
tic aneurysm. Ann N Y Acad Sci 1996; 800: 246–248.
18. Lindholt JS, Heegaard NH, Vammen S, Fasting H, Henneberg
EW, Heickendorff L. Smoking, but not lipids, lipoprotein(a)
and antibodies against oxidised LDL, is correlated to the
expansion of abdominal aortic aneurysms. Lancet 1994; 344:
19. Badger SA, O’Donnell ME, Sharif MA, McMaster C, Young
IS, Soong CV. The role of smoking in abdominal aortic
aneurysm development. Angiology 2009; 60: 115–119.
20. Gotto AM Jr. Role of C-reactive protein in coronary risk
reduction: focus on primary prevention. Am J Cardiol 2007;
21. MacSweeney ST, Ellis M, Worrell PC, Greenhalgh RM,
Powell JT. Smoking and growth rate of small abdominal aor-
tic aneurysms. Lancet 1994; 344: 651–652.
22. Schlosser FJV, Tangelder MJD, Verhagen HJM, et al., on
behalf of the SMART study group. Growth predictors and
prognosis of small abdominal aortic aneuryms. J Vasc Surg
2008; 47: 1127–1133.
23. Agmon Y, Khandheria BK, Meissner I, et al. Is aortic dilata-
tion an atherosclerosis-related disease? J Am Coll Cardiol
2003; 42: 1076–1083.
24. Thompson AR, Cooper JA, Ashton HA, Hafez H. Growth
rates of small abdominal aortic aneurysms correlate with
clinical events. Br J Surg 2010; 97: 37–44.
Badger SA et al.
25. Santilli SM, Littooy FN, Cambria PA, et al. Expansion rates
and outcomes for the 3.0cm to the 3.9cm infrarenal abdomi-
nal aortic aneurysm. J Vasc Surg 2002; 35: 666–671.
26. Kurvers H, Vieth FJ, Lipsitz EC, et al. Discontinuous, sta-
catto growth of abdominal aortic aneurysms. J Am Coll Surg
2004; 199: 709–715.
27. Wilmink AB, Hubbard CS, Day NE, Quick CRG. The inci-
dence of small abdominal aortic aneurysms and the change in
normal infrarenal aortic diameter: implications for screening.
Eur J Vasc Endovasc Surg 2001; 21: 165–170.
28. Lindholt JS, Vammen S, Juul S, Fasting H, Henneberg EW.
Optimal interval screening and surveillance of abdominal aor-
tic aneurysms. Eur J Vasc Endovasc Surg 2000; 20: 369–373.
29. Collin J, Heather B, Walton J. Growth rates of subclinical
abdominal aortic aneurysms – implications for review and
re-screening programmes. Eur J Vasc Endovasc Surg 1991;
30. Cook TA, Galland RB. A prospective study to define the
optimum rescreening interval for small abdominal aortic
aneurysm. Cardiovasc Surg 1996; 4: 441–444.