Background: Validity and reliability of radial artery (RA) assessment techniques, during
coronary artery bypass grafting (CABG), have been at the heart of debates for the last
decades. The correct RA assessment is crucial prior their surgical harvesting to avoid
Objectives: This research aims to evaluate and compare the validity and reliability of the most
commonly adopted RA assessment techniques in CABG surgery.
Data sources: A literature search was undertaken, through five electronic databases, to
access recent studies relating to the assessment of RAs prior to their surgical harvesting during CABG
Review methodology: A systematic review (SR) was conducted to appraise relevant
primary research studies published in the English language between 2010 and 2020. Data
findings were extracted for analysis, quantitative synthesis and conclusions drawn.
Results: Nine studies were included. Modified Allen test (MAT) presenting with reduced
validity and/or reliability was revealed by seven studies. Pulse-oximetry and
plethysmography, used in combination with the MAT, offer more objective results than an
isolated MAT. Ultrasonography provides important insight into the morphological
characteristics of RAs.
Conclusion: Outcomes of this SR suggest ultrasonography screening to be superior in
RA assessment for both validity and reliability; however high-quality research is required to
support these findings.
Keywords: Ultrasonography, modified Allen test, pulse-oximetry, plethysmography, radial
The optimal conduit selection for coronary artery bypass graft (CABG) surgery has been
widely investigated during the development of this surgical procedure. The gold-standard
in CABG was identified in the use of the left internal thoracic artery (LITA) grafted onto the
left anterior descending (LAD) coronary artery. The seminal study suggesting this surgical
approach was produced as long ago as 1986 (Loop, et al., 1986). Nonetheless, in more
recent years Cuminetti, et al. (2017) confirmed that this remains. For multivessel CABG
other conduits, harvested elsewhere in the body, may be used as potential grafts.
Although revascularisation using venous conduits, particularly the greater saphenous vein
(GSV), remains a well-established common practice, the literature suggests arterial
grafting to be superior for long-term graft patency, clinical outcomes and survival rates
(Aldea, et al., 2016; D’Agostino, et al., 2018). Consequently, the radial artery (RA) is
considered one of the first graft choices.
With radial arteries (RAs) becoming more frequently adopted during CABG, the
assessment of these vessels, prior their surgical harvesting, appears to be of vital
importance to avoid post-operative hand ischaemic complications and unnecessary
surgical exposure. Nonetheless, validity and reliability of screening techniques, normally
adopted for the morphological and functional assessment of RAs, is the object of current
debate and their use remains ambiguous and contradictory (Habib, Baetz and Satiani,
2012). There is no protocol or guideline recommending the most effective assessment
technique: the choice for the most accurate and consistent screening technique is yet to
be determined.This research aims to evaluate the most commonly adopted RA assessment techniques
and compare their reliability and generalisability. Recommendation on the most valid and
reliable screening test may be provided, hence the potential to minimise post-operative
acute and chronic complications, including infections, ischaemia and hand impairment in
patients undergoing RA harvesting during CABG surgery
The World Health Organisation (2018) classified cardiovascular diseases as the leading
cause of death worldwide, with ischaemic heart disease (IHD) being the most prevalent.
According to the British Heart Foundation (2020) latest reports, IHD caused 9.43 million
deaths globally in 2016. This is a cardiac disorder caused by atherosclerosis: a chronic
degenerative process of the coronary arteries which results in the coronary artery disease
(CAD). Coronary artery bypass graft surgery is considered the best therapeutic strategy
against CAD, yet the most common cardiac operation worldwide (Melly, et al., 2018).
Coronary artery bypass surgery has undergone significant changes since its first
introduction in 1967. Not only have surgical techniques been developed and perfected
over the years, but also conduit graft choices have been reconsidered (Squiers and Mack,
2018). Radial arteries for example, were originally seldom used, due to high rates of early
graft failure and intimal hyperplasia. However, with the implementation of pharmacological
interventions, such as the use of vasodilators to minimise the risk of RA spasm, and
improvement in the harvesting techniques, RAs were reconsidered as optimum grafts for
coronary revascularisation (Tatoulis and Schwann, 2018). They are one of best possible
graft conduits in fact, second only to the internal thoracic arteries (Yadava, et al., 2016).
Not only have RA conduits shown superior patency rates than saphenous vein grafts, but
they are also considered easy to harvest, due to their anatomical location. Nonetheless,the function of this vessel, along with the ulnar artery, is extremely important in the blood
supply of the hand. Therefore, RA harvest may alter vascular and neurological normal
physiology, thus producing devastating results, yet causing considerable morbidity and
impairment (Blitz, Osterday and Brodman, 2013). Moreover, peripheral arterial calcification
involving RAs is another aspect to consider while planning for the harvesting of these
vessels: it may compromise the long-term graft patency, thus impacting on patient’s
survival after CABG (Watchmaker, Watchmaker and Watchmaker, 2019). All the above
emphasises the imperative need for the correct morphological and functional RA
assessment prior to proceed with their surgical harvesting.
In recent years, the steady increase in the use of RA grafts, has established the necessity
to investigate the best assessment technique, improving surgery outcome and standards
of care (Cuminetti, et al., 2017). Nevertheless, in current clinical practice, RA assessment
protocol varies, depending on local institutional policy or surgeon’s preferences. The
assessment technique choice ranges from the pre-operative modified Allen test (MAT), to
the intra-operative use of a pulse-oximetry to measure oxygen saturation changes, after
invasive occlusion of RA. Other institutions use the squirt test, in which blood flow is
visually inspected after invasive occlusion and surgical incision of the distal RA segment
(Habib, Baetz and Satiani, 2012). The inconsistency of a universally agreed approach, the
adoption of invasive and non-measurable, hence subjective, techniques underline the lack
of an evidence-based practice, indicating the urgent need for further research.
The availability of up to date systematic reviews, related to the research topic, was
investigated, consulting the international prospective register of systematic reviews
(PROSPERO) and Cochrane database. The only study found, focused on the reliability of
the MAT and the incidence of acute hand ischaemia after RA puncture. It does not
consider the irreversible and major complications secondary to the surgical harvesting of RAs during CABG, which will form the rationale for this SR. Moreover, it does not reflect
on the use of additional assessment techniques other than the MAT.
A scoping review was undertaken to appraise published literature and have a clear
understanding of current knowledge relating to the research topic.
Available screening tests have been widely investigated with the aim to determine their
validity/reliability. Nonetheless, primary research studies conducted in the last decade,
presented with inconsistent results, thus highlighting a controversial approach in the RA
assessment. The most commonly discussed techniques within the current literature
include: MAT, ultrasonography, pulse-oximetry and digital plethysmography. In addition to
the above, other assessment techniques, such as: angiography hand arteriogram,
computed tomography angiography, intraoperative pressure measurement, laser speckle
contrast analysis and manual or histopathology based morphometry were also
documented within the literature. However, these were more sporadically adopted or used
in combination with MAT or ultrasonography to determine their validity and reliability.
The Allen test was the first developed screening technique to assess the capacity of the
blood supply to the hand, in the event of RA procedures, which would have temporarily or
permanently compromised RA blood flow. These procedures include: arterial puncture for
blood gas sampling, catheterisation for percutaneous coronary interventions, or harvesting
during CABG surgery or formation of a radio-cephalic arteriovenous fistula for
haemodialysis access (Zisquit and Nedeff, 2019). In 1952, twenty-three years after its first
introduction, the practice on how to perform the Allen test was reviewed by Irving Wright,
with the aim to promote more objective results. It was therefore renamed the modified
Allen test (Zisquit and Nedeff, 2019).
The modified Allen test, being considered a cost effective and time efficient technique,
represents the most common RA screening test (Sivaharini, Babu and Mohanraj, 2018).
Although MAT was identified to be a safe, adequate and simple technique to perform
(Ronald, Patel and Dunning, 2005; Yadava, et al., 2015), it was later confirmed to be
prone to observer biases, thus emphasising its subjectivity (Altinsoy, et al., 2017).
Furthermore, MAT presents with limited sensitivity and specificity, hence showing
significant false-negative and false-positive results (Kohonen, et al., 2010; Gokhroo, et al.,
2016; Altinsoy, et al., 2017; Zhang, et al., 2020). Therefore, there are some studies
suggesting the combined use of MAT with pulse-oximetry or digital plethysmography, in
order to overcome common MAT limitations. Specifically, their use aims to determine a
more quantifiable and objective result, hence a more accurate RA assessment, yet
avoiding more expensive and time consuming techniques (Al-metwalli, 2014; Elwali and
Conversely, the use of ultrasonography was supported by several trials as it enables the
selection of the best quality RA segment. Furthermore, it allows a wider morphological and
physiological assessment of RAs, hence the potential to minimise unnecessary surgical
exposure and post-operative ischaemic complications (Yadava, et al., 2016; Vukovic, et
al., 2017). Ultrasonography was in fact recognised as an accurate method to assess
anatomical aspects of RAs, predicting stenosis and structural variation of the vessel, thus
facilitating the early detection of intraluminal atherosclerotic plaques. Moreover, it permits
the inner diameter measurement of the vessel and the estimation of blood flow within
forearm major arteries, thus predicting radial/ulnar dominance and avoiding hand
ischaemia (Yadava, et al., 2016). Despite the efficiency in the functional assessment of
RAs, ultrasound was considered a time consuming and expensive technique, also
requiring examiners’ advanced skills. Therefore, ultrsonography use was questionable and
not always justified (Yadava, et al., 2015).The correct pre-operative RA assessment may have a positive impact on both short and
long-term surgical outcomes, reducing mortality and morbidity rates after CABG surgery.
The identification of the best RA assessment technique may contribute to a higher quality
of patient-centred care.
• To evaluate and compare screening techniques in assessing radial artery
quality for use in coronary revascularization
• To recommend the most accurate and reliable test within cardiac institutions,
reinforcing evidence-based practice
• To establish the validity and reliability of the MAT, pulse-oximetry,
plethysmography and ultrasonography
• To determine if screening test choice has a significant impact on the outcome of
The research question was formulated as follows:
Is ultrasonography more accurate and reliable than the modified Allen test, pulse-oximetry
and plethysmography, in radial artery assessment for patients undergoing coronary
The literature review, performed in relation to the research topic, revealed contradictory
findings from the primary research studies undertaken. A systematic review represents the
most efficient methodology to compare the findings of several primary studies, answering
the same research question. It enables the reviewer to systematically collect and critically
appraise published literature, and extract and analyse data to produce a synthesis (Boland,
Cherry and Dickson, 2017). A systematic review was therefore considered the most
appropriate approach to answer the research question, yet producing evidence-based
knowledge and highlight the gaps within the literature, which would eventually require
further research. Moreover, an SR methodology allows the appraisal of different study
designs and a larger variety of populations, avoiding concerns about the
generalisability determined by sample size when conducting a primary research study.
Furthermore, standards of clarity, rigor and replication are not compromised (Kane, Butler
and Ng, 2016).
This systematic review includes primary research studies with a quantitative nature.
Numerical data will be extracted, interpreted and synthesised from all selected studies,
through clear prescriptive steps, and a systematic review, with a quantitative design, will
be presented. Research studies with a quantitative methodological philosophy are
epistemologically grounded in positivism (Creswell and Creswell, 2018). This is considered
an objective and scientific paradigm. It refers to an experimental approach of collecting
data; as a consequence, results are not influenced by participants’ judgements and
experiences (Kivunja and Kuyini, 2017). Collecting and analysing empirical data, a
quantitative synthesis is produced, to answer the research question formulated, thus providing an important insight into the choice of most valid and reliable RA assessment
technique, thus impacting clinical practice.
A comprehensive search of all available literature was performed and a scoping review
undertaken. A high quality review is in fact performed by appraising all available evidence,
relating to the research question (Parahoo, 2014). Therefore, primary and secondary
research, abstracts, as well as grey literature and conference proceedings were
considered, and available data searched adopting a pragmatic and systematic approach.
Five electronic databases were consulted to retrieve studies, published between 2010 and
2020, considering RA assessment techniques during CABG surgery, including MEDLINE,
PubMed, CINAHL, SCOPUS and Embase. These databases were specifically selected as
they provide relevant resources relating to biomedical health sciences, yet pertaining to
the research subject. MEDLINE and PubMed are two search platforms offering over
twenty-seven million biomedical and life science related records. The CINAHL database
provides nursing and allied health literature. SCOPUS is considered the largest platform
for peer-reviewed abstract literature linked with technology, medicine and social science
fields. Embase is an electronic database covering international biomedical literature
(Parahoo, 2014). The full-text of studies unavailable on the mentioned databases was
obtained from the British Library, the academic institution library and the reviewer’s clinical
institution research department.
The selection of keywords in the Medical Subject Headings (MeSH) terms, Boolean search
operators (AND, OR) and word truncation (asterisk wildcard) are of fundamental
importance to combine and retrieve variants of a word stem (Baumann, 2016). The MeSH
terms adopted are reported in table 1.
Four combined searches were performed, in every database, including above MeSH terms
and following the strategy highlighted in table 2, producing more relevant search results
This allowed the combination of different key-words, including:
• one RA assessment technique
• forearm arteries
• coronary artery bypass grafting
An overview of the four combined searches is presented in table 3.
Results of the database searches are reported in appendix A.
Additional records were identified through reference harvesting and hand searching of
relevant specialist journals. Complimentary searches were conducted in Google Scholar
and ProQuest, allowing the author to search the grey literature to retrieve any other
studies, abstracts, theses and conference proceedings, with the aim to identify further
resources answering the research question.
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
flowchart (Figure 1) was used to retrace the steps of the search process undertaken, yet
highlighting the search outcome (Moher, et al., 2015). Initially, 686 articles were retrieved
from the database searches, 2 additional articles were identified through reference
harvesting and hand searching. Duplicates were removed using RefWorks software,
leaving 466 articles. Abstract were screened and full-text reviewed, removing 447 and 10
articles respectively. Nine articles were included in the quantitative synthesis.
Once all available articles were screened, a methodical and impartial study selection was
performed. The first phase in the selection process involves the definition of a well-structured review question. Eriksen and Frandsen (2018) recommended the use of the
PICO (Population/Problem Intervention Comparison Outcome) framework to formulate a
research question. It is in fact a largely adopted tool in reviews with a quantitative design
(Fain, 2017). Therefore, the research question of the presented SR was structured by
using this model, as reported in Table 4.
Inclusion and exclusion criteria were established and applied, narrowing the search results
to those studies that specifically answered the research question, yet excluding less
relevant evidence. Bettany-Saltikov (2016) suggests that high quality SRs are obtained by
following this prerequisite. A rigorous study selection requires the definition of explicit and
clear inclusion and exclusion criteria. These set out the boundaries and define the
attributes for a study to be included. The PICO framework facilitates the selection of
evidence-based resources (Watt and Eng, 2015). Therefore, the inclusion and exclusion
criteria were established following the PICO system and they are presented in table 5
Applying inclusion and exclusion criteria to the 19 articles, resulted in 10 being excluded
with reasons (appendix B). The remaining 9 articles were selected for the quantitative
The nine remaining primary research studies were critically appraised and their quality
assessed. Downes, et al. (2016) emphasised the importance of critical appraisal in the
design, analysis and reporting of a research study, prior to implementation in clinical
practice. Assessing research quality is an aspect of vital importance, as it enables the
reviewer to investigate if a study was conducted applying the appropriate measures to
minimise bias, thus increasing the standard. Similarly, the quality of the studies will impact
on the quality of the systematic review (Boland, Cherry and Dickson, 2017). Therefore, a
rigorous quality assessment is required, as it may influence the systematic review shape,
conclusion and outcome. According to Heale and Twycross (2015), appraising rigour in
quantitative studies is obtained by measuring validity and reliability, hence assessing the
accuracy and the consistency of the results. Moreover, reproducibility and generalisability
of the studies are other important parameters, to highlight their trustworthy (Saunders,
Lewis and Thornhill, 2019). Generally, articles published in peer-reviewed journal are
expected to be robust in their design and conclusions drawn. However, their quality level
may vary; therefore, a comprehensive quality appraisal is recommended to evaluate
strengths and limitations of the studies.
The nine studies included in this SR were critically appraised through a validated and
updated quality assessment tool: the modified Downs and Black checklist (appendix C).
This checklist was considered the most appropriate tool, allowing the assessment of both
randomised and non-randomised research studies, of healthcare interventions, presenting
with a quantitative methodological approach (Downs and Black, 1998). Validity and
reliability of this scale were evaluated to be appropriate by the National Collaborating
Centre for Methods and Tools. Overall, this scale obtained a strong methodological rating
(National Collaborating Centre for Methods and Tools, 2020). Acknowledged to be
applicable to the retrieved studies and not compromising the standards of the review
process, the modified Downs and Black checklist was adopted within this SR. Outcomes of
the critically appraised studies are presented in the results section of this SR (table 8).
Activities performed in a clinical environment respect the ethical mandate to promote
wellbeing to people receiving care. Healthcare professionals (HCP) abide by their code of
ethics and professional conduct, constantly respecting this principle (Nursing and
Midwifery Council, 2018). Consequently, primary research studies, conducted on
healthcare interventions, must respect the same ethical mandate. Therefore, ethical
approval must be obtained by the Research Ethics Committee prior to undertake primary
researches (Polit, 2017). A systematic review is considered the highest level of research
design in the hierarchy of evidence (Murad, et al., 2016; Thoma, et al., 2019). However, it
does not undertake primary research, and therefore, no ethical approval is required.
Nonetheless, complying with academic regulations, the Faculty Research Ethics Panel of
the university was informed prior to commencing this SR. A stage 1 Research Ethics
Application Form (appendix D) was completed and submitted to the faculty along with
evidence confirming the successful completion of the Research and Professional Ethics
online course (appendix E).
Research on healthcare interventions is regulated by common law, other than by
professional regulatory bodies, including the General Medical Council
and the Medical Research Council (Medical Research Council, 2007), and
different statutes, such as the Human Rights Act 1998 and the Mental Capacity Act 2005.
All the above ensure medical practice meets ethically approved criteria of safeguarding
rights of health, safety and wellbeing of patients keen to participate in research studies.
Researchers’ enthusiasm and the interest of the society must never prevail over the
wellbeing, security and interests of participants (Al Tajir, 2018).
Decision making in healthcare research is influenced by the ethical principles of autonomy,
beneficence, non-maleficence, justice, fidelity, veracity and confidentiality (Parahoo, 2014).
Considering these principles, researchers balance the risk of harm and the potential
benefit produced by the research process. Safety of participants must never be
compromised and risk of harm maintained at minimal level, regardless of the benefits
brought from the research to the community. Similarly, researchers owe the duty to inform
participants regarding the aim of the study, what his/her involvement will lead to.
Furthermore, risks and benefits of the research must be disclosed. A study respecting
these ethical principles is recognised as high quality evidence (Parahoo, 2014).
A further regulation to orientate HCP’s practice, in the field of medical research, is
represented by the Declaration of Helsinki. This is a statement of ethical principles,
produced by the World Medical Association (WMA), providing directives to HCP on how to
conduct healthcare research involving human subjects (WMA, 2018). Primary research
studies, included in this SR, replicated these principles.
The nine articles included in this SR were critically appraised to establish their ethical
quality, other than the rigour adopted in their research process. The ethical appraisal
involved the assessment of whether these studies sought ethical approval by a Research
Ethics Committee and if voluntary and informed consent was gained by all participants.
All studies clearly respected anonymity, privacy and confidentiality of participants,
throughout the research. However, three out of the nine studies did not have documented
ethical approval and/or consent statements from participants. One of these studies
(Zhang, et al., 2020) had a retrospective nature; hence data were collected from clinical
databases. In these circumstances it is rather difficult to gain patient consent. Therefore,
Ethics Committees investigate if such research is sufficiently valid to justify the breach of
autonomy, yet establishing the absence of any other alternatives to perform the research
(Hope, Savulescu and Hendrick, 2008). In this study patient consent was waived, however
ethical approval was granted by the Ethics Committee.
Wu, et al. (2019) revealed that reporting of ethical consideration in peer-reviewed
international journals is currently developing compared to earlier years, when ethical
standards were not clearly defined, and therefore, many clinical trials did not report ethical
considerations. However, improvements are required as this tendency is still a matter of
concern: as evidence of ethical review are sometimes still omitted (Wu, et al., 2019). This
would suggest incomplete reporting from Kohonen, et al. (2010) and Gokhroo, et al.
(2016), which however does not mean ethical principles were not respected or ethical
approval was not granted. The outcomes of the ethical appraisal process are reported in Table 6.
Data extraction process was performed, following the appraisal of rigour and ethical
assessment of the selected studies. It is considered beneficial to undertake quality
assessment first, as exclusion of primary studies with poor quality, may avoid unnecessary
data extraction (Fleeman and Dundar, 2017).
The aim of the data extraction is to summarise and present relevant data addressing and
relating to the research question (Boland, Cherry and Dickson, 2017). A bespoke data
table was designed and piloted to facilitate collection of both descriptive and analytical
data, to highlight study characteristics and outcomes
Data extraction was conducted by the reviewer independently; however accuracy of the
process was monitored by the reviewer’s academic supervisor. It is recommended that a
cross-check is conducted by two reviewers independently, in order to minimise errors and
bias, yet determining more robust results (Büchter, Weise and Pieper, 2020). Similarly,
contacting authors for clarification of missing data is another strategy with the potential to
improve the quality of the review
Variables to extract from the studies were identified. Dependent variables were
represented by: intra-operative findings (pressure measurement, manual morphometry),
post-operative findings (histopathology based morphometry), morphological examination
through Doppler ultrasound, forearm/hand arteriogram and computed tomography
angiography of upper limbs. Independent variables include the use of ultrasonography,
modified Allen test, pulse oximetry or plethysmography during RA assessment.
Different methods may be used in an SR to analyse, summarise and present data extracted
from primary research. These include meta-analysis, meta-synthesis and narrative
analysis (Siddaway, Wood and Hedges, 2019). Meta-analysis is a statistical method able
to derive conclusions with a low margin of error, as suggested by Gurnsey (2017).
However, it may only be performed if extracted data is sufficiently similar, thus the
combination of primary research appears a sensible process to undertake. According to
Denscombe (2012) the use of descriptive statistical analysis, if conducted with rigour and
transparency, is a valid strategy to present data and findings of primary studies. Therefore,
a meta-analysis is not an essential component of a review, as corroborated by Boland,
Cherry and Dickson (2017).
The heterogeneity of study characteristics, the diversity in their adopted protocols and
the inconsistency of research outcomes precluded the combination of extracted data, thus
preventing the use of a statistical analysis. The reviewer opted for an alternative analysis
method: a narrative synthesis is presented.
After applying inclusion criteria to the findings of the literature search, nine
primary research studies were selected for the final review. These studies were
consequently critically appraised through the modified Downs and Black checklist for their
quality assessment (appendix C), where results are synthesised in Table 8. Study
characteristics and outcomes were collected using the piloted data extraction tool.
Findings were combined and presented in Table 9. Furthermore, ethical considerations,
bias and limitations were considered for each study.
The Downs and Black checklist was first developed in 1998, with the aim of producing a
valid and reliable quality assessment scale to appraise randomised and non-randomised
trials (Downs and Black, 1998). A modified version of the checklist was then proposed,
which nowadays appears more largely adopted (Hooper, et al., 2008; Trac, et al., 2016). It
consists of 27 specific questions, divided into five subscales: Reporting; External validity;
Internal validity - bias; Internal validity - confounding; Power. Answers may be: “Yes”, “No”,
“Partially” (only for question 5) and “Unable to determine” (only for questions 11-26). The
highest possible score for the checklist is 28. Four studies were assessed as Fair quality
and five studies were considered Good quality. The quality index for each subscale was
measured in all nine studies (Figure 2a) and the proportion of studies with the high, medium
and low-quality index for each subscale is presented (Figure 2b).
Power subscale ranges 0-1: it is not shown on the histogram if statistical significance (p<0.05) was not reached from the study.
The quality index for each subscale was measured, and studies were considered of:
• HIGH quality index for subscale (if score ≥80%)
• MEDIUM quality index for subscale (if 40%<score<80%)
• LOW quality index for subscale (if score ≤40%).
All studies present with a research protocol linked with the research question of this review
and overall, their quality was satisfactory.
A summary of trial characteristics and outcomes is reproduced in table 9 below.
The nine primary research studies, all single centre trials, were undertaken between 2002
and 2019, with their publications dated from 2010 to 2020. Eight studies adopted a
prospective methodology and presented the following study designs: four observational
studies; one cohort study; one randomised trial; one randomised controlled trial; one pilot
study. Only one study adopted a retrospective cohort data analysis from a previously
conducted observational trial. Sample size of the studies ranges from 11 to 536
participants, whose characteristics are presented in table 10.
In five studies (Kohonen, et al., 2010; Yadava, et al., 2015; Yadava, et al., 2016; Altinsoy,
et al., 2017; Vukovic, et al., 2017) participants were adults patients undergoing elective
CABG surgery, with their RAs being assessed as potential graft conduits for coronary
revascularisation treatment. In the remaining four studies (Al-metwalli, 2014; Gokhroo, et
al., 2016; Elwali and Moussavi, 2020; Zhang, et al., 2020) participants were adult
individuals, undergoing RA assessment, in the event of future CABG and RA harvest
procedure. Baseline demographic characteristics, including gender and mean age of
participants, were specified in the majority of the studies, except for one study (Yadava, et
al., 2015) in which gender of participants was not disclosed. Seven studies also reported
detailed information regarding prevalence of major coronary risk factors among their study
Different RA assessment techniques were compared in the studies. Modified Allen test
and ultrasonography resulted in the most commonly discussed, being investigated in eight
and six trials respectively. Pulse-oximetry and plethysmography were also examined in
two studies each. Moreover, six other assessment techniques were documented within
five of the nine trials, however, with the aim to establish validity and reliability of the four
previously mentioned screening techniques. These further tests include: intraoperative
pressure measurement, histopathological examination, angiography arteriogram of
forearm arteries, manual morphometry, histopathology based morphometry and
multidetector computed tomography angiography.
The study from Yadava, et al. (2016) evaluated the morphological characteristics of RAs
with three assessment techniques: pre-operative Doppler ultrasonography, intra-operative
manual morphometry and post-operative histopathology based morphometry. A correlation
was established among morphometric findings of tunica intima and tunica media
thicknesses. The study revealed the accuracy of Doppler ultrasonography, validated
against histopathology examination, for providing reliable data on wall thickness and
luminal diameter measures. Nonetheless, confounding factors were encountered, first
among all, the histopathological confirmation of Doppler findings are only available for the
distal and proximal segments of RAs. Moreover, this is the first study validating Doppler
against histopathological examination of RAs during CABG surgery, and therefore results
are not confirmed by other research. Nonetheless, a correlation between the two
examinations, while assessing RA wall thickness, was performed within the context of
haemodialysis patients for the formation of A-V fistula (Kim, et al., 2004).
Similarly, the study performed by Vukovic, et al. (2017) confirmed the findings from
Yadava, et al. (2016), identifying ultrasound as an accurate technique in providing preoperative quality assessment of RAs. It enables the examiner to select disease-free
sections of RAs, thus promoting the surgical harvesting of best morphological segments of
the vessel, thus improving CABG outcomes. Although this research adopted the same
design as Yadava, et al. (2016), conducting a prospective observational study, the review
question was addressed on a significantly larger population (N=536). Nonetheless, this
trial did not conduct angiographic or morphometric follow-up examinations to prove the
hypothesis that RAs with better morphological characteristics result in superior long-term
graft patency. As a consequence, predictive value of the examination is not confirmed,
neither the statistical significance documented.
Conversely, in 2015 Yadava and his team presented a study with discordant results.
Comparing MAT with ultrasonography and histopathological examination of RAs
extremities, the researchers analysed the predictive value, sensitivity and specificity of
ultrasonography. They suggested that the use of ultrasonography was not justified prior to
RA harvest, with the MAT recognised as a technique easy to perform, interpret and
reproduce, yet emphasising that it is a safe isolated screening technique. In this study only
one reviewer was involved in the ultrasonography assessment of RAs, while it is not
declared who performed the MAT, thus weakening the validity of the methodology
adopted. Furthermore, the authors propose that the use of ultrasound should be limited to
those patients presenting with positive MAT, which would suggest a non-harvestable RA,
in order to exclude false-positive results. However, in this study ultrasonography was only
implemented on patients presenting with negative MAT, thus preventing from revealing
false-positive results. Another finding worth highlighting is that they advise ultrasound RA
assessment not necessary in all cases, especially in countries with limited resources. This
does not justify that people living in developing countries should receive inferior care.
In earlier years, Kohonen, et al. (2010) compared, through a prospective cohort study, the
pre-operative Allen test and ultrasonography findings, with plethysmography of the second
and fourth digits and invasive intra-operative blood pressure measurement. The authors
revealed contradictory findings between Allen test and ultrasonography, highlighting the
reduced specificity of the Allen test, which presents with significant false-positive results.
Therefore, Kohonen, et al. (2010) suggest the use of intraoperative pressure
measurement when Allen test is positive and no other objective metric techniques are
available. Nonetheless, statistical significance was not reached to confirm this hypothesis.
Another limitation of the study is represented by the relatively limited sample size (N=90).
Follow up assessment was considered; however, this was conducted only in 19 patients.
Moreover, a low quality index was measured for the external validity and internal validityconfounding subscales of the modified Downs and Black checklist (Figure 2a).
More recently, Gokhroo, et al. (2016) conducted a prospective observational study,
relating MAT results to the forearm and hand angiography arteriograms, obtained from
their study population (N=302). They recommended that the use of isolated MAT is not
justifiable for documenting collateral hand circulation, prior to any RA radical procedures
including surgical harvesting. This trial presented with the second larger study population,
and detailed information was also provided for the operator performing the procedures and
the examiners analysing their results. However, no evidence of statistical significance or
predictive value of results was documented. Additionally, a low quality index for internal
validity-confounding subscale was calculated (Figure 2a).
The study from Altinsoy, et al. (2017) adopted a similar research protocol, comparing, in a
prospective randomised trial, the Allen test and the MAT with the multidetector computed
tomography (MDCT) angiography of upper limbs. The study determined the presence of
RA occlusive disease, secondary to calcification, in six patients, confirmed at the MDCT
angiography, however not detectable from an abnormal Allen test/MAT result. Therefore,
validity and effectiveness of these screening tests were revealed to be limited and their
use questionable. Nevertheless, computed tomography angiography was only performed
in patients presenting with negative, hence normal Allen/MAT, thus preventing from
considering false-positive results, yet limiting the use of arterial grafting. Although the
study establishes the subjectivity and reduced sensitivity of Allen test and MAT, specificity
of those screening techniques was not considered, highlighting a limitation of the study.
Among all reviewed research, the only trial presenting with a retrospective design is the
cohort study recently conducted by Zhang, et al. (2020). This research was performed by
reviewing the medical records of 164 patients who underwent endoscopic thoracic
sympathectomy (ETS) treatment. The study aimed to compare the results obtained from
Doppler ultrasonography and MAT, completed before and after ETS, to provide insight into
RA status, in the event of future CABG and RA surgical harvesting. Endoscopic thoracic
sympathectomy resulted in dilating RAs, determining a transition to a negative MAT result
in patients who presented with positive MAT prior to ETS. This may potentially lead to
false-negative MAT results, as concluded by the researchers. The main limitation of the study is represented by the adopted retrospective design, which limits researchers’ control
on the exposure to confounders (Polit and Beck, 2017). Moreover, monitoring of long-term
changes in RAs diameter was omitted, thus precluding the possibility to evaluate the
temporary/permanent variation of MAT results.
Al-metwalli (2014) presented a prospective randomised controlled trial (RCT), evaluating
sensitivity and specificity of Doppler ultrasonography, MAT and pulse-oximetry in
assessing collateral hand circulation (CHC). The results highlighted the reduced specificity
of MAT, when findings are compared with Doppler ultrasonography. However, the author
suggested the use of a perfusion index (PI) test, obtained through the combination of MAT
and pulse-oximetry, identified as an accurate objective test with high sensitivity, specificity
and positive predictive value. These findings, validating the safe use of pulse-oximetry
during RA harvesting, are supported by a previously conducted study (Johnson, et al.,
1998). Nevertheless, this trial presented with some limitations. Firstly, participants in this
research were all healthy volunteers, whom RA characteristics may be well distinctive and
respond differently than the vessel of ill patients requiring CABG and RA surgical
procedures. Therefore, despite the high quality evidence of an RCT (Murad, et al., 2016;
Thoma, et al., 2019), a more representative study population is required to give more
convincing statistical results in the safety of adopting PI. Additionally, with PI being a
dynamic parameter, consideration of internal and external stress factors would have been
Another strategy to overcome MAT limitations was recently proposed by Elwali and
Moussavi (2020). They indicated the use of photo-plethysmography in combination with
MAT to provide information on the CHC and recommendation on the safe use of RA grafts.
Despite the attentive consideration of different variables, this research has an extremely
small sample size (N=11): the smallest among all nine studies. Furthermore, participants coronary risk factors were not documented, neither morphometric RA characteristics
assessed. All of the above limitations resulted in limiting the generalisability and validity of
the trial, which also presents with a low quality index for internal validity-confounding
subscale (Figure 2a).
Across the nine studies, different conclusions were drawn.
Two studies highlight the validity of ultrasonography in RA assessment (Yadava, et al.,
2016; Vukovic, et al., 2017). Accuracy of ultrasound, in evaluating wall thickness and
luminal diameter of RAs was validated by intra-operative physical measurement
(Yadava, et al., 2016). Ultrasonography allows the detection of calcification and
atherosclerotic plaques, providing an accurate and reliable assessment of RA
morphology, avoiding the harvesting of diseased RA sections (Vukovic, et al., 2017).
One study identifies the MAT as an efficient and safe isolated test, remarking on the
non-justifiable routine use of ultrasound (Yadava, et al., 2015).
The validity and reliability of the MAT are being questioned by four studies. Using the
MAT, approximately 1 in 9 patients resulted in having a false negative MAT (Altinsoy,
et al., 2017), increasing to 1 in 3 patients according to Gokhroo, et al. (2016) (poor
sensitivity) (Gokhroo, et al., 2016; Altinsoy, et al., 2017; Zhang, et al., 2020). Similarly,
approximatively 6% of patients reported a false positive MAT (poor specificity)
(Kohonen, et al., 2010). Furthermore, Altinsoy, et al. (2017) reveal that the MAT is a
subjective (operator/patient dependent) test.
Two further studies suggest a more objective alternative to the MAT, represented by the
combination of this with pulse-oximetry (Al-metwalli, 2014) or plethysmography (Elwali
and Moussavi, 2020). One study reported a higher sensitivity (100%) and specificity
(98.8%) than the isolated MAT (Al-metwalli, 2014).
A correlation between findings and quality level of the studies is synthesised and
presented in table 11.
The overall finding of the quantitative synthesis, based on the outcomes and the quality
assessment of the nine studies demonstrates a reduced validity and reliability of the MAT.
The view of using pulse-oximetry and plethysmography to overcome the limitations of MAT, and obtain a more objective RA assessment is limited and would require further
investigations. Ultrasonography technique obtained a consensus of a better screening test
for the morphological assessment of RAs (see background literature).
A detailed pre-operative vessel examination is extremely important in avoiding
unnecessary surgical exploration of RAs, yet minimising ischaemic complication risks in
those patients not suitable to undergo their harvesting procedure (Dick, et al., 2011;
Altinsoy, et al., 2017). Furthermore, an effective and comprehensive RA quality
assessment plays a fundamental role in the prevention of harvesting diseased RA
sections, which may cause graft failure and cardiac events after CABG (Chowdhury, et al.,
2004; Hosono, et al., 2019). Since Starnes, et al. (1999) acknowledged the limitations of
the most commonly adopted MAT, and revealed the risks carried out by its use, different
research studies have been conducted to investigate the most valid and reliable RA
assessment technique. However, their results demonstrated conflicting evidence, as
revealed by this review. Ultrasonography was recognised as a good screening technique,
to assess morphological characteristics and anatomical variations in the RA course
(Yadava, et al., 2016; Vukovic, et al., 2017). Conversely, the MAT results in a more cost-effective and time-efficient test to evaluate CHC prior to RA procedures (Yadava, et al.,
2015), however, it presents with some limitations (Kohonen, et al., 2010; Gokroo, et al.,
2016; Altinsoy, et al., 2017; Zhang, et al., 2020). These are overcome by adopting pulse oximetry or plethysmography to obtain more objective results (Al-metallic, 2014; Elwali
and Moussavi, 2020).
This systematic review compared the findings of nine studies, addressing the review
question after inclusion and exclusion criteria were applied to the search results. One
prospective randomised trial and one RCT were identified, however methods
of randomisation were not stated in either of the two studies. In two studies (Al-metwalli,
2014; Elwali and Moussavi, 2020) participants’ young age and absence of coronary risk
factors revealed the questionable validity and generalisability of their methodology
adopted. Similarly, insufficient documentation of who performed the RA assessment
techniques was found in three studies (Kohonen, et al., 2010; Yadava, et al., 2015;
Altinsoy, et al., 2017), leading to reduced validity of their methodology. All articles were
published in peer-reviewed journals. However, the majority of the trials presented with a
small sample size (N≤100), except for one study (Vukovic, et al., 2017) recruiting 536
patients. However, this research was conducted over a longer period of time (12 years:
2004-2016) when compared to other trials.
The findings of the review were impacted by the quality of the studies: ultrasound was considered
a more effective technique in higher quality studies, and therefore they received major
consideration when generating the main outcome of this SR.
Although ultrasonography was evaluated as the best screening technique, in current
clinical practice it is not commonly used as considered a more costly and time-consuming
technique, requiring examiners’ advanced skills. In contrast, MAT is widely used due to
historical practice and convenience (Habib, Baetz and Satriani, 2012; Sivaharini, Babu and
Mohanraj, 2018). Seven of the nine studies evaluating the validity and reliability of MAT
reported negative outcomes in comparison to other screening techniques. This finding is
consistent with previously conducted research, highlighting the reduced sensitivity and
specificity of the MAT and the subjectivity of its results (Starnes, et al., 1999; Agrifoglio, et
al., 2005; Asif and Sarkar, 2008; Vukovic, et al., 2008). An attempt was made to obtain
more objective results through the combination of MAT with pulse-oximetry or digital
plethysmography, yet avoiding time-consuming and expensive techniques (Al-metallic,
2014; Elwali and Moussavi, 2020). Although the Al-metallic (2014) study showed promising results and was of good quality, significant confounding variables were identified
in both primary studies. The main confounder was the recruitment of healthy volunteers,
as they may possess different characteristics from patients who require coronary
revascularisation. However, results are promising and will hopefully inspire further
research in this area.
An early literature review by Habib, Baetz and Satiani (2012) and more recently the study
by Chakravarthy, et al. (2020) focused on the specific field of cardiac surgery and
presented consistent findings in supporting the use of ultrasonography in RA assessment,
thus revealing benefits to patients undergoing CABG. This SR was designed with the aim
of determining the validity and reliability of the most commonly adopted RA screening
techniques, thus promoting evidence-based practice in the correct RA assessment for
patients requiring coronary revascularization.
The review process was performed by a single individual, with limited experience in
research and data analysis: this may generate bias, thus resulting in the main limitation of
this SR. A study should be carried out by a co-ordinated specialist research team, in order
to minimise bias and subjectivity of results (Polit and Beck, 2017). Only one RCT, relevant
to the research question, was included in the SR. This may potentially reduce the
statistical power of the review, as the credibility of its results largely depends on the quality
of the studies. Moreover, all trials were single-centre studies, limiting the generalisability of
their findings. Finally, only English articles published between 2010 and 2020 were
considered. Restraints of language and time frame may not only exclude relevant
international or earlier studies but also introduce selection bias.
Limited studies related to the RA assessment within the context of cardiac surgery were
identified, notwithstanding the extensive literature search undertaken. To overcome
selection bias, caused by database searches, screening of reference lists from all selected
studies was conducted, leading to the inclusion of any additional records. A quantitative
synthesis was produced by incorporating all relevant evidence from the current available
All reviewed studies emphasise the importance of an accurate and reliable RA
assessment, to prevent harvesting unsuitable vessels which would compromise long-term
outcomes after CABG surgery.
Modified Allen test and ultrasonography resulted in the most commonly investigated
screening techniques, with the MAT being considered the standard examination in current
clinical practice prior to RA harvest. Nonetheless, this SR revealed the reduced validity
and reliability of this technique as a general consensus in most of the studies. The use of
pulse-oximetry and plethysmography, in combination with the MAT, was documented in
two articles and established to be more objective alternatives to the isolated MAT in the
RA assessment. Ultrasonography has shown to be superior in the evaluation of
morphological characteristics of RAs, including detection of physiological variations and
pathological aspects (atherosclerosis and calcification) of the vessel. Ultrasonography
provides objective and consistent results in the pre-operative insight of RAs, enabling
selection of RA segments with favourable morphological features. However, these findings
are yet to be confirmed by high quality evidence, such as RCTs. The systematic review suggests ultrasonography to be the most accurate and reliable RA assessment technique
among the screening tests compared. Nonetheless, confirmation of ultrasound superiority
by an RCT represents an essential prerequisite to recommend its routine use in CABG
surgery. Conversely, the MAT should no longer be recommended prior to RA harvest, as it
presents with limited diagnostic validity and reliability. The need for a large RCT,
comparing all four RA assessment techniques, is required to support the findings of this