Article ID: CJ-14-0972
Background: Physical examination as an initial screening tool to diagnose abdominal aortic aneurysm (AAA) has lost favor over the past 20 years. This multicenter cohort study aimed to determine the prevalence of AAA in elderly Japanese patients with hypertension (HT) and to clarify the diagnostic accuracy of physical examination using a pocket-sized ultrasound imaging device (the “pocket-echo”).
Methods and Results: A total of 1,731 patients with HT aged >60 years from 20 collaborating institutions were enrolled in this study. Abdominal palpation was performed on physical examination, and the pocket-echo was used to confirm the diagnosis of AAA. The abdominal aorta was well visualized in 1,692 patients (98%). AAA was discovered in 69 patients (4.1%), with advanced age and male sex identified as independent risk factors. The prevalence of AAA increased according to age regardless of sex, and reached 9.7% and 5.7%, respectively, in males and females ≥80 years. Overall, 33 cases of AAA were missed on abdominal palpation (sensitivity, 52%), whereas for AAAs >40 mm, the sensitivity was 75%.
Conclusions: We assessed the utility of the pocket-echo and physical examination for diagnosing AAA in Japanese patients with HT aged over 60 years. Our findings highlight the importance of AAA screening programs in high-risk Japanese populations, and confirm the ability of physical examination to detect large, but not small, AAAs.
Rupture of abdominal aortic aneurysms (AAAs) is now considered an important cause of sudden death. Because AAA remains silent until rupture, screening programs have been demonstrated to reduce the frequency of rupture and AAA-related deaths by providing early diagnosis and allowing for elective surgical intervention.1 Physical examination, including abdominal palpation, is generally used as an initial screening method for AAA in daily practice. However, previous investigations, albeit with relatively small numbers of subjects, have reported unsatisfactory results using physical examination for AAA detection.2–5 Ultrasound is currently considered the standard tool for screening of AAA,6–8 and a pocket-sized ultrasound imaging device (the “pocket-echo”) was recently introduced to clinical practice;9–11 however, unfortunately, such advancements in medical imaging techniques have resulted in a decline of the physical examination skills required to diagnose cardiovascular abnormalities.12,13 This multicenter cohort study was designed to assess the utility of the pocket-echo to detect AAAs in a large number of elderly Japanese patients with hypertension (HT), and the results were compared with the findings of physical examination. The aims of this study were (1) to investigate the prevalence of AAA in elderly Japanese patients with HT and (2) to evaluate the accuracy of physical examination in the diagnosis of AAA.
Editorial p ???
The AAA Japan Study was a multicenter, observational investigation, which was started in October 2012. This study was conducted at 20 institutions, including 15 hospitals and 5 private clinics. These institutions were considered to be eligible because they had established, well-experienced physicians and/or sonographers for vascular ultrasound examination. All data, including medical history, medications, and the results of the physical and pocket-echo examinations, were obtained by the attending physicians at each institution and sent to the Osaka Ekisaikai Hospital for analysis. Patient recruitment ended in September 2013. This study was approved by the institutional review board of each hospital. For the 5 private clinics, which did not have their own ethics committee, the approval obtained from the Institutional Review Board of Osaka Ekisaikai Hospital was used. All subjects provided informed consent to participate in this study.
Study PopulationThe study population comprised 1,731 patients with HT aged over 60 years (942 males; mean age, 75±8 years; range, 60–99 years). Pocket-echo and physical examinations were simultaneously performed to screen for the presence of AAA. HT was defined as blood pressure ≥140/90 mmHg on repeated measurements or by the administration of antihypertensive medication to control blood pressure.14 Patients taking antihypertensive medication for conditions other than HT, such as angiotensin-converting enzyme inhibitors for congestive heart failure, were excluded. The initial exclusion criteria were: (1) a history of aortic disease, including AAA, aortic dissection, or aortitis, and (2) the presence of other serious systemic diseases.
Physical and Pocket-Echo ExaminationsAAA was diagnosed if, with the patient in the supine position with knees raised and the abdomen relaxed, an abdominal expansile pulsation was detected between the epigastrium and umbilicus on abdominal palpation.
Pocket-echo examination was performed using the Vscan (GE Medical Systems, Milwaukee, WI, USA), which was equipped with a broadband-width phased array transducer with a mean transmitting frequency of 1.7 MHz and a mean receiving frequency of 3.8 MHz.15 This device has gain- and depth-control features, as well as measurement packages that include linear measurement caliper capabilities. The entire abdominal aorta was visualized in the transverse and longitudinal planes. AAA was defined as a focal enlargement of the abdominal aorta with a maximum external diameter greater than 30 mm or more than 1.5-fold the diameter of the proximal aorta.16
Statistical AnalysisCategorical variables are presented as number (%) and continuous variables as mean±standard deviation. The chi-square test was used for comparison of categorical variables, whereas continuous variables were compared by the unpaired t-test or Mann-Whitney U test according to the type of data distribution. Multivariate logistic regression analysis was performed to identify independent factors associated with the presence of AAA. Baseline variables that showed a univariate relationship with outcome were entered into the multivariate analysis. Differences were considered significant at P<0.05.
The interobserver variability for the ultrasound measurements was analyzed in 15 subjects by 2 independent, blinded observers. Moreover, the intraobserver variability was analyzed in another 15 subjects by the same observer at 2 different time points. The results were analyzed by both least-squares fit linear regression analysis and the Bland-Altman method.
Among the total 1,731 patients, 39 patients (2.3%) were excluded from the analysis because of inadequate quality of the pocket-echo images, resulting in 1,692 patients (98%) being included in the final analysis. The clinical characteristics of the 2 groups were similar, except for weight and body mass index (Table S1), with the patients with inadequate image quality on pocket-echo being heavier (P=0.047) and having a greater body mass index (P=0.015) than those with adequate image quality.
Prevalence and Determinants of AAAOf the 1,692 analyzed patients, 747 (44%) were enrolled in hospitals and 945 (56%) patients in private clinics. These consisted of 1,374 outpatients (81%) and 318 inpatients (19%); 89% (n=1,506) received antihypertensive agents, namely β-blockers, diuretics, calcium-channel blockers, angiotensin-converting enzyme inhibitors, and angiotensin-receptor blockers in 559 (33%), 451 (27%), 926 (55%), 150 (9%), and 920 (54%) patients, respectively (Table 1). All these patients had very well controlled blood pressure.
| AAA (+) (n=69) |
AAA (−) (n=1,623) |
P value | |
|---|---|---|---|
| Age, years | 79±8 | 74±8 | <0.001 |
| Sex (male), n (%) | 48 (70) | 873 (54) | 0.013 |
| Height, cm | 159±9 | 157±10 | 0.071 |
| Weight, kg | 59±12 | 59±12 | 0.97 |
| Body mass index, kg/m2 | 23.1±3.6 | 23.7±3.6 | 0.16 |
| Heart rate, beats/min | 73±13 | 70±12 | 0.054 |
| Systolic blood pressure, mmHg | 132±19 | 132±16 | 0.99 |
| Diastolic blood pressure, mmHg | 71±12 | 73±10 | 0.13 |
| Hyperlipidemia, n (%) | 39 (57) | 933 (57) | 0.87 |
| Impaired glucose tolerance, n (%) | 21 (30) | 473 (29) | 0.82 |
| Smoking status | 0.027 | ||
| Current smoker, n (%) | 9 (13) | 122 (8) | |
| Former smoker, n (%) | 31 (45) | 569 (35) | |
| Never smoked, n (%) | 29 (42) | 932 (57) | |
| Atrial fibrillation, n (%) | 9 (13) | 255 (16) | 0.55 |
| History of coronary artery disease, n (%) | 35 (51) | 605 (37) | 0.024 |
| History of peripheral artery disease, n (%) | 7 (10) | 77 (5) | 0.043 |
| History of stroke or transient ischemic attack, n (%) | 10 (14) | 131 (8) | 0.059 |
| Medications | |||
| β-blocker | 31 (45) | 528 (33) | 0.032 |
| Diuretic | 28 (41) | 423 (26) | 0.008 |
| Calcium-channel blocker | 38 (55) | 888 (55) | 0.95 |
| Angiotensin-converting enzyme inhibitor | 7 (10) | 143 (9) | 0.70 |
| Angiotensin-receptor blocker | 36 (52) | 884 (54) | 0.71 |
Data are presented as mean value±standard deviation or n (%) of patients. AAA, abdominal aortic aneurysm.
Overall, AAA was observed in 69 patients (4.1%), including 34 patients evaluated in hospital and 35 patients in private clinics. The prevalence of cases of AAA in hospital and the private clinics was 4.6% and 3.7%, respectively (P=0.38). Table 1 shows the characteristics of patients with and without AAA. Patients with AAA were older (P<0.001) and more often male (P=0.013), smokers (P=0.027), and had a history of coronary artery (P=0.024) and peripheral artery disease (P=0.043). There were no differences between the 2 groups in heart rate or blood pressure. Multivariate logistic regression analysis confirmed that age (P<0.001) and sex (P=0.006) were significant independent determinants of the presence of AAA, whereas smoking status was not a significant predictor of AAA in the multivariate analysis (Table 2).
| Variable | Odds ratio | 95% CI |
|---|---|---|
| Model 1 | ||
| Age, years | 1.1** | 1.05–1.12 |
| Male sex | 2.2* | 1.25–3.72 |
| Current smoker | 1.9 | 0.90–4.08 |
| History of coronary artery disease | 1.5 | 0.90–2.41 |
| History of peripheral artery disease | 1.6 | 0.68–3.64 |
| Model 2 | ||
| Age, years | 1.1** | 1.05–1.12 |
| Male sex | 1.7 | 0.93–3.25 |
| Current or former smoker | 1.7 | 0.93–3.01 |
| History of coronary artery disease | 1.4 | 0.85–2.29 |
| History of peripheral artery disease | 1.6 | 0.70–3.72 |
Both models included age, male sex, current smoking, and history of coronary artery disease and peripheral artery disease. In model 1, smoking status included only current smokers. In model 2, smoking status included both current and previous smokers. *P<0.01, **P<0.001. AAA, abdominal aortic aneurysm; CI, confidence interval.
The age- and sex-specific results are shown in Figure 1. The prevalence of AAA gradually increased with age in both men and women. Further, men had a higher prevalence of AAA than women in each age category. Thus, the highest prevalence of AAA was observed in men aged ≥80 years (9.2%), whereas the lowest prevalence was seen in women aged 60–69 years (0.6%).
Age- and sex-specific prevalence of abdominal aortic aneurysms (AAAs).
The distribution of AAA size is summarized in Figure 2. The mean diameter of the AAA was 34±7 mm, ranging from 25 to 62 mm. There were 12 AAAs >40 mm (17% of all AAAs, 0.7% of the overall population).
Size distribution of abdominal aortic aneurysms (AAAs).
Of the 69 patients with AAAs, 36 were correctly diagnosed by abdominal palpation and 33 were missed using this technique (sensitivity, 52%). Abdominal palpation detected 9 of 12 AAAs with diameter >40 mm, showing an improved sensitivity (75%) relative to abdominal palpation in the overall population. The diagnostic value of abdominal palpation is summarized in Table 3. A total of 30 of 57 AAAs with diameter <40 mm (53%) were missed by abdominal palpation.
| Pocket-echo | ||
|---|---|---|
| AAA (+) | AAA (−) | |
| A. Overall population | ||
| Physical examination | ||
| AAA (+) | 36 | 25 |
| AAA (−) | 33 | 1,598 |
| B. AAA >40 mm | ||
| Physical examination | ||
| AAA (+) | 9 | 27 |
| AAA (−) | 3 | 1,653 |
(A) Agreement=97%, sensitivity=52%, specificity=98%, PPV=59%, and NPV=98%. (B) Agreement=98%, sensitivity=75%, specificity=98%, PPV=25%, and NPV=99.8%. AAA, abdominal aortic aneurysm; NPV, negative predictive value; PPV, positive predictive value.
Furthermore, the body mass index of the 25 AAA patients identified using abdominal palpation but not confirmed using the pocket-echo (overestimated patients) was 22.5±3.2, which tended to be lower than that of the population as a whole (23.7±3.6, P=0.088). In contrast, the 33 patients identified using the pocket-echo but not using abdominal palpation (underestimated patients) had a body mass index similar to that of the overall population (23.5±4.1 vs. 23.7±3.6, P=0.99).
Excellent correlation was observed in both the interobserver (r=0.98) and intraobserver (r=0.99) measurements. From the Bland-Altman analysis, the inter- and intraobserver variabilities were determined as 0.57 mm and 0.27 mm, respectively.
In this multicenter cohort study, we used the pocket-echo to evaluate the prevalence of AAA in a large number of elderly Japanese patients with HT. The main strength of this study was the large population available for examining the diagnostic accuracy of abdominal palpation for AAA in the current era. We found that physical examination with abdominal palpation is still useful for screening large AAAs. As another strength, this study demonstrated that AAAs are relatively common in Japanese populations at high risk of AAA (namely, elderly patients with HT), even when blood pressure is well controlled by medication. The prevalence of AAA increases with age regardless of sex, and reached 9.7% and 5.7% in men and women aged ≥80 years, respectively.
Screening ModalitiesPrevious investigations over the past 10 years have examined the utility of abdominal palpation for the diagnosis of AAA.2–5 However, the widespread adoption of medical devices and their increasing sophistication has resulted in a declining ability to diagnose cardiovascular disease based on physical examination alone.12,13 This calls for a reappraisal of the diagnostic accuracy of abdominal palpation for AAA in the current era. In this study, we found that approximately half of all AAAs that were too small for likely rupture were missed by abdominal palpation, which suggests a need for screening programs using imaging modalities, as small AAAs increase in size over time and therefore need to be followed up.17,18 In contrast, abdominal palpation detected 75% of all AAAs >40 mm, and careful physical examination thus remains a possible initial screening method in a selected population that is ineligible for ultrasound screening and/or intensive management, such as patients of a very advanced age.19,20
The current screening strategy for AAA relies on systemic ultrasound examination.6–8 A pocket-sized echocardiographic imaging device was recently introduced to clinical practice, and has been demonstrated to provide immediate diagnostic information as part of the clinical examination. Previous studies have reported the accuracy of the pocket-echo for the measurement of cardiac chamber size, as well as for their functional status.9–11 Subsequently, the cost-effectiveness of the pocket-echo for screening various cardiac diseases has been reported.21 Dijos et al showed an excellent correlation between the abdominal aortic diameter determined by standard echocardiography and the pocket-echo, resulting in 100% agreement for the AAA diagnosis.22 Considering the portability and low cost of pocket-sized imaging devices, similar to stethoscopes, the pocket-echo may alter the current screening strategies for AAA by providing a supplement to physical examination.
Prevalence of AAAMost of the available data regarding the prevalence of AAA has been derived from studies in Western countries.19,23–26 Further, there is a paucity of data regarding the prevalence of AAA in patients aged over 80 years. Bruce et al in the USA performed ultrasound screening of AAA in patients with HT aged over 70 years, and reported that AAA occurred in 11 of 125 patients (8.8%).27 Similarly, in a population-based investigation from Australia, the prevalence of AAA was 7.2% in patients aged between 65 and 83 years;19 further, in another population-based investigation from Norway, AAA was found in 19.8% and 5.2% of men and women aged 75–84 years, respectively.25 Conversely, few studies have examined the prevalence of AAA in Japanese patients. Takai et al did not find AAAs >30 mm in 348 subjects with normal blood pressure,28 which concurred with a retrospective analysis by Salem et al, in which AAA was observed in only 0.45% of Asian residents in the UK.29 Furthermore, Adachi et al also performed AAA screening in 1,591 subjects in a Japanese rural community.30 They found that, although the prevalence of AAA was 0.3% overall, it was 7.7% in subjects with HT, which is consistent with our detection of AAA in 4.1% of patients with HT aged over 60 years. Taken together, the previous results suggest that our findings (9.7% and 5.7% prevalence, respectivelyin men and women aged ≥80 years) were not too high, and that the prevalence of AAA in Japanese populations at high risk of AAA seems to be similar to that in Western populations. Moreover, in this study, the prevalence of AAA was found to increase according to age, regardless of sex. AAA was detected in 5.7% of women aged over 80 years, which was similar to the results for men aged 70–79 years (5.7%) (Figure 1). Thus, we believe that AAA screening is justifiable to reduce AAA-related mortality in both male and female elderly Japanese people with HT, even when their blood pressure is well controlled with antihypertensive medication. Further, simultaneous examination of the aorta using echocardiography in patients with coronary artery disease would be valuable for the screening of AAA.
Study LimitationsFirst, the physical and pocket-echo examinations were performed by experienced physicians or sonographers at each institution, and the results were sent to the study center for analysis. These processes are close to those pertaining to real clinical situations, but inter-institution variability in the examinations may exist and may have influenced the results of this study. Second, smoking status was associated with AAA in the univariate, but not the multivariate analysis, and this may derive at least partly from the lack of detailed information about smoking status (duration of smoking history and/or years after quitting smoking)31 and/or discrepancies in AAA formation owing to genetic and/or lifestyle differences.32,33 Third, because this was a cohort investigation, the effect of antihypertensive medication on AAA occurrence was not estimated.34 Fourth, currently computed tomography is a mainstream modality for measuring AAA size. However, the AAA Japan study was not designed to validate the accuracy of the pocket-echo, and the results of other imaging modalities were not included. There might be the over- or underestimation in measuring AAA size compared with other imaging modalities. Finally, the risk of AAA formation may not fully reflect the risk of rupture. The prognosis and cost-effectiveness should be estimated to determine the benefits of conducting screening programs in the Japanese population, considering the development of interventional techniques such as endovascular repair of AAA.35,36
The AAA Japan study, a large multicenter cohort study, showed that AAA is present in 4.1% of Japanese patients with HT aged over 60 years. Pocket-echo may supplement physical examination for AAA detection, particularly for the identification of small AAAs.
We sincerely thank Kentaro Shibayama, MD (Heart Center, Tokyo Bay Urayasu/Ichikawa Medical Center) for the ultrasound measurements, and Akiomi Inoue, PhD (Department of Mental Health, Institute of Industrial Ecological Sciences, University of Occupational and Environmental Health, School of Medicine) for providing feedback and criticism of our statistical analysis methods.
The authors declare no conflict of interest.
This work was supported in part by a research grant from the Osaka Foundation for the Prevention of Cancer and Cardiovascular Diseases.
Investigators Participating in the AAA Japan Study
Shota Fukuda, MD (Department of Medicine, Osaka Ekisaikai Hospital, Osaka, Japan); Hiroyuki Watanabe, MD (Heart Center, Tokyo Bay Urayasu/Ichikawa Medical Center, Urayasu, Japan); Katsuomi Iwakura, MD (Division of Cardiology, Sakurabashi Watanabe Hospital, Osaka, Japan); Masao Daimon, MD (Department of Clinical Laboratory, The University of Tokyo Hospital, Tokyo, Japan); Yukio Abe, MD (Department of Cardiology, Osaka City General Hospital, Osaka, Japan); Hiroki Oe, MD (Department of Cardiovascular Medicine, Okayama University Graduate School of Medicine, Okayama, Japan); Kenichiro Otsuka, MD (Department of Cardiovascular Medicine, Osaka City University Graduate School of Medicine, Osaka, Japan); Ryo Otsuka, MD (Otsuka Clinic, Ikoma, Japan); Ryoko Kitada, MD (Department of Cardiology, Asakayama General Hospital, Sakai, Japan); Hajime Kihara, MD (Kihara Cardiovascular Clinic, Asahikawa, Japan); Konomi Sakata, MD (Second Department of Internal Medicine, Kyorin University School of Medicine, Mitaka, Japan); Kengo Suzuki, MD (Division of Cardiology, Department of Internal Medicine, St. Marianna University School of Medicine, Kawasaki, Japan); Yoshihiro Seo, MD (Cardiovascular Division, Faculty of Medicine, University of Tsukuba, Tsukuba, Japan); Tstutomu Takagi, MD (Takagi Cardiology Clinic, Kyoto, Japan); Masahiro Tamashiro, MD (Department of Cardiology, Tomishiro Central Hospital, Tomisugu, Japan); Koki Nakanishi, MD (Department of Cardiovascular Medicine, Baba Memorial Hospital, Sakai, Japan); Eiichi Hyodo, MD (Nishinomiya Watanabe Cardiovascular Center, Nishinomiya, Japan); Keigo Yasukawa, MD (Dr.GON Crinic, Miyakojima, Japan); Satoshi Yamada, MD (Department of Cardiovascular Medicine, Hokkaido University Graduate School of Medicine, Sapporo, Japan); Ken Yoshida, MD (Yoshida Clinic, Osaka, Japan); Hiroshi Ito, MD (Department of Cardiovascular Medicine, Okayama University Graduate School of Medicine, Okayama, Japan); and Junichi Yoshikawa, MD (Nishinomiya Watanabe Cardiovascular Center, Nishinomiya, Japan).
