Introduction
Abdominal aortic aneurysm (AAA) is present when the vessel has a diameter of 30 mm or more.1-4Western world prevalence has been reported between 1.2% to 3.0%.1,5-11Age and gender influence greatly AAA prevalence, being more common in males, particularly over 65 years old, where it can achieve a prevalence as high as 4.0% to 8.0%.1,2,12This is not surprising considering that smoking, arterial hypertension and dyslipidemia are well known risk factors for AAA,1,3,13-15all medical conditions often encountered in western male seniors.
Although over 80% of AAA are asymptomatic, it is believed that 1 in every 3 AAA may progress to rupture.16-17This is a life-threatening event, with mortality rate reaching up to 80% to 90%.2,18AAA rupture hazard is greater in large aneurysms, particularly those above 50 mm.2 Elective surgical repair should be considered for vessels over 55 mm.2,18Mortality rate for elective surgery is reportedly low, both for open and endovascular techniques (5.0% and 2.0%, respectively).2
Abdominal ultrasound is an accessible, safe, efficient and low-cost screening method for AAA.3,18,19Its limitations are that it is operator dependent, and it can be more difficult to perform in obese patients or if there is marked intestinal gas presence.2,3,19
Due to its high mortality burden, low complications rates of elective surgical repair and a simple and low-cost screening method, several countries (such as United States of America, Sweden and United Kingdom), have implemented national screening programs for senior males. Reduced mortality not only due to AAA rupture but also from all causes of mortality has been reported.20,21As most countries, Portugal does not have a national screening program for AAA.
In 2015, 534 patients admitted to our internal medicine ward were male over 65 years old and would therefore qualify as ideal candidates to AAA screening. Also, as expected in an internal medicine ward, AAA risk factors are very common in all our patients. Only 16 male patients over 65 years old had a diagnosis of AAA at ward admission. Considering western world’s prevalence for this gender and age group, we could be underdiagnosing between 5 to 27 AAA per year.
In the absence of a national screening program, hospital admission could be an opportunity to perform AAA screening by ultrasound, without significantly increasing hospitalization cost or length of stay. However, in this type of screening we will probably identify older and more fragile patients. Point of care ultrasound is available at our ward, although there is no previous report to our knowledge of internal medicine physicians doing this procedure. We considered that, with a short training period, internal medicine residents would be able to perform quality abdominal ultrasound for AAA screening, and therefore increase our diagnostic accuracy of this life-threatening condition
Material and Methods
Five internal medicine residents (IMR) without previous experience in ultrasound received a total of 5 hours of training, divided in two sessions, provided by an Internal Medicine attending with point of care ultrasound experience and a vascular surgeon attending.
Male patients over 65 years old admitted to general Internal Medicine Wards at our tertiary care academic hospital, without known AAA, were candidates to opportunistic AAA screening. The only exclusion criterion was the patient refusal to participate in the study. The ultrasounds were made preferentially in the morning, and if was not possible the aorta imagining because of inference of bowel gas, the exam was repeated in other time.
After informed consent was obtained, protocol dictated that each patient would be evaluated in different times by two IMR, in order to assess non specialist interobserver consistency. A subset of patients was also randomly evaluated by vascular surgery attendant or abdominal computed tomography (CT) scan, which was made for other motives, and were used to compare the IMR evaluation to a gold standard measurement of AAA. All examinations were made without knowing the result of previous evaluations. If at least one of the IMR diagnosed a AAA, a gold standard evaluation of this patient was mandatory, although that examinator was blinded for the first measurement.
Abdominal ultrasounds were performed using Siemens ACUSON P500® ultrasound with an CH5-2 transducer. After the evaluation along abdominal aorta, transversal diameter was determined below renal arteries, since AAA occur more frequently in this location, and it was considered the easiest method for non-experienced operators. It was considered AAA a measurement of 30 mm or more. Measurement differences were evaluated, both between IMR observations and also IMR-gold standard. Risk factors for AAA were recorded.
Statistical analysis was performed using SPSS v26.0. Absolute and relative frequencies were used for qualitive variables and mean, and standard deviation (SD) were used for quantitative variables. Interobserver measurement agreement for quantitative variables was evaluated by intraclass correlation coefficient (ICC) and 95% confidence interval (95% CI). Cohen’s Kappa coefficient was used to evaluate agreement between observers in the determination of AAA presence (qualitative variable). ICC > 0.9 and Cohen Kappa > 0.8 were considered as a strong concordance between observers. A significance level of 0.05 was considered for all analysis.
The study was approved by the ethical committee.
All positive cases were sent for follow-up at Vascular Surgery consultation.
Results
From November 2017 to April 2019, a total of 98 patients gave consent to participate in our study. All patients were evaluated by 2 IMR, except for one that was evaluated by 1 IMR and the Vascular Surgeon attendant. In 24 patients, IMR evaluation was compared to a gold standard exam (12 abdominal CT performed for other reasons + 12 vascular surgeon ultra-sounds). A total of 207 abdominal ultrasounds were performed (195 by IMR and 12 by vascular surgeon). The number of ultrasounds performed by each IMR were 55, 44, 43, 38 and 14.
Mean age was 77.7 (SD = 9.2) years old and 99.0% were Caucasian. Except for 3 patients, all had at least 1 risk factor for AAA other than age and gender. Risk factor prevalence is presented in Table 1.
Risk Factor | n (%) |
---|---|
Caucasian | 97 (99) |
Smoking (past and present) | 75 (76.5) |
Dyslipidemia | 60 (61.2) |
Arterial hypertension | 75 (76.5) |
Family history of AAA | 1 (1.0) |
Past diagnosis of other aneurysmatic lesions | 5 (5.1) |
Coronary disease | 36 (36.7) |
Cerebrovascular disease | 28 (28.6) |
Peripheral arterial disease | 14 (14.3) |
AAA - abdominal aortic aneurysm
A total of 8 AAA were diagnosed, resulting in an 8.2% prevalence. A total of 24 patients were evaluated both by 2 IMR and gold standard evaluation, resulting in a total of 47 interactions (24 first IMR evaluation-gold standard + 23 evaluation by the second IMR-gold standard). Only 1 IMR examination failed to identify an AAA when both the gold standard and the other IMR confirmed its presence, but the maximum difference between measures was only 3.1 mm. The mean of the measurements made by IMR was 19.1 mm and by gold standard was 26.8 mm. There was good agreement regarding AAA presence between IMR examinations and IMR/gold standard (Table 2).
n | Cohen’s Kappa (SE) | p | |
Between IMR | 97 | 0.918 (0.082) | <0.001 |
Between IMR/gold standard | 47 | 0.950 (0.049) | <0.001 |
SE: standard error; IMR: internal medicine residents
Considering IMR AAA screenings, there was a mean difference of 1.8 mm (SD 1.88; min 0.10 mm; max 9.40 mm) between measurements. Good measurement agreement was observed between the different IMR observers. When comparing IMR to gold standard evaluations, there was a mean difference of 3.34 mm (SD 3.84; min 0.10 mm; max 17.90 mm) between measurements.
Likewise, there was a good agreement between observers (Table 3).
n | ICC | 95% CI | p | |
Between IMR | 97 | 0.956 | 0.935-0.970 | <0.001 |
Between IMR/gold standard | 47 | 0.957 | 0.923-0.976 | <0.001 |
95% CI: 95% confidence interval; ICC: intraclass correlation coefficient; IMR: internal medicine residents
The life-long impact of this opportunistic AAA screening is yet to be determined. From the 8 patients diagnosed with AAA, 4 had indication for elective surgical repair. Two of them underwent surgery without any reported perioperative compli-cation and 1 refused any intervention. One patient had significant comorbidity that would recommend against AAA surgical repair. The other four patients died from unrelated medical conditions.
Discussion
Our study showed a high prevalence of AAA in hospitalized males over 65 years old. In fact, our prevalence of 8.2% matches the upper limit estimated for western countries and is much higher than a previous reported 2.4% prevalence in a Portuguese non-hospitalized population.2,22This can certainly result from the significant presence of AAA risk factors in our population,23 something we believe can also be encountered in other European Internal Medicine wards. Also, it confirms that AAA may be an underdiagnosed condition in male seniors admitted to internal medicine wards.
IMR without previous experience in ultrasound, submitted to a 5-hour training program, were able to accurately perform abdominal ultrasound for AAA screening, with the advantage of not taking time off the patient´s usual observation, as it can be integrated into the objective examination. They could correctly diagnose AAA and its size comparatively to gold standard evaluations and also had a good agreement between them. We have not analysed individual performance of each of the IMR since there was a good agreement between examinations both for diagnose of AAA presence and AAA measurement. Recently, Brakel et al24 have reported the use of a 4-point aorta scan ultrasound by unexperienced health practitioners to determine the presence AAA. Comparatively to our study, there was a greater number of discrepancies between the unexperienced health practitioners and control evaluation by radiologist. However, other studies have already proved the effectiveness of teaching ultrasound to non-radiologists, but in the emergency department or in outpatients,25,26 showing a different type of patients.
Our study has limitations. Only 1 in each 4 patients of our study population were evaluated by a gold standard exam, either an abdominal ultrasound performed by a vascular surgeon or abdominal CT scan. Also, since by protocol gold standard evaluation was mandatory if any of the IMR diagnosed an AAA, these patients are overrepresented in the subset of patients submitted to gold standard evaluation.
Since individuals with high comorbidity burden can usually be found in Internal Medicine wards, more judicious patient selection maybe in order and protocols with Vascular Surgery Department should be established and revised as new techniques for AAA surgical repair development. Our study shows that AAA screening can be implemented in internal medicine wards by the available health staff, without significant costs or time constrains.
Conclusion
Point of care opportunistic AAA screening can be performed in Internal Medicine wards in selected patients with known risk factors. This intervention may provide an early diagnosis, in time of preventing the catastrophic event of an AAA rupture, particularly in countries without a national AAA screening program.