Dysphagia (2016) 31:721729

DOI 10.1007/s00455-016-9737-3

REVIEW ARTICLE

Diagnostic Accuracy of the Modified Evans Blue Dye Test

in Detecting Aspiration in Patients with Tracheostomy:
A Systematic Review of the Evidence
Sibylle Bechet1 Fiona Hill2 Orla Gilheaney1 Margaret Walshe1

Received: 28 April 2016 / Accepted: 21 July 2016 / Published online: 16 August 2016
 Springer Science+Business Media New York 2016

Abstract Oropharyngeal aspiration (OPA) is a common aspiration in patients with tracheostomy, using a stan-
occurrence in patients with tracheostomy. The modified dardised and reliable procedure. Outcomes from such
Evans blue dye test (MEBDT) is an easily administered studies will update the current level of evidence in relation
bedside procedure for the assessment of tracheostomised to the MEBDT and consequently define best clinical
patients. However, studies evaluating the diagnostic accu- practice.
racy of the MEBDT reach conflicting results. Therefore,
we conducted a systematic review to determine the overall Keywords Deglutition disorders
Oropharyngeal
accuracy of the MEBDT in detecting OPA in adults with aspiration
Modified Evans blue dye test
Diagnostic
tracheostomy. The search strategy incorporated searching accuracy
Tracheostomy
electronic databases, checking reference lists and citations
and retrieving unpublished data. Data of primary studies
were extracted and examined by three independent Introduction
reviewers. The assessment of the methodological quality of
included studies was performed using the QUADAS-2 tool. Within the last decade, the number of tracheostomy tube
Six studies met the inclusion criteria for this systematic placements has drastically increased and the insertion of a
review. The studies presented significant disparities in tracheostomy has become a frequently performed proce-
study design and patient characteristics. Furthermore, high dure in intensive care medicine [13]. Reasons include
discrepancies in the administration of MEBDT across advances in critical care medicine, such as the development
studies were noted. Therefore, a meta-analysis was not of less invasive tracheostomy techniques, and improved
considered appropriate. Sensitivity estimates varied widely management of critically ill patients as well as a growing
across the studies (3895 %), indicating that the MEBDT elderly population [4, 5]. Different manifestations of
is unreliable in detecting OPA. However, the studies swallowing difficulties have been reported in combination
emerge with overall high specificity values, ranging from with the presence of a tracheostomy tube [6, 7]. While it is
79 to 100 %. This true negative rate suggests that the recognised that oropharyngeal aspiration (OPA) frequently
MEBDT correctly identifies patients without OPA. This occurs in patients with long-term artificial airways, cur-
review highlights the need for further research studies rently the literature lacks clinical evidence of a causal
assessing the accuracy of the MEBDT in detecting relationship between tracheostomy and aspiration [8]. An
early diagnosis is therefore crucial to enable immediate
management as well as accurate intervention. The restric-
& Sibylle Bechet ted availability and the need of specialised healthcare
bechets@tcd.ie
practitioners complicate or even prevent the use of instru-
1
Department of Clinical Speech and Language Studies, Trinity mental assessment in evaluating deglutition dysfunction of
College Dublin, 7-9 South Leinster Street, Dublin 2, Ireland tracheostomised patients [9]. Thus, clinical practice would
2
Department of Speech and Language Therapy, Tallaght highly benefit from an easily administered and accurate
Hospital, Dublin 22, Ireland screening tool for aspiration in tracheostomised patients.

123
722 S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting

The modified Evans blue dye test (MEBDT) is a recom- Search Strategy
mended bedside procedure for the evaluation of deglutition
and aspiration in tracheostomised patients [10, 11] and The search strategy for identification of studies incorporated
involves mixing blue dye with water or semisolid food. The searching eight different electronic databases (PubMed,
presence of blue dye in or around the tracheostomy tube CINAHL, Embase, ScienceDirect, Web of Science, Psy-
indicates a possible aspiration. cINFO, LILACS, ProQuest Nursing and Allied Health
However, primary studies assessing the diagnostic accu- Source, ProQuest Dissertation & Theses). Databases were
racy (DA) of the MEBDT reach conflicting results. Current sought from October 2014 to April 2016. A preliminary
research lacks clear and reliable results, which could be used search in the databases PubMed and CINAHL was con-
as a basis for clinical practice. Despite a lack of strong evi- ducted to determine the relevant indexing terms. Because of
dence of its accuracy, the MEBDT is used in clinical practice the paucity of studies assessing the DA of the blue dye test,
to diagnose deglutition disorders underlying various dis- the electronic database search was performed using only the
eases. A survey on clinical consistency among SLTs in the index test as key search term. After this initial search, the
management of tracheostomies indicates that out of the 64 reference standards and target condition served as inclusion
participants, 45 % of the SLTs reported having used the criteria in the screening process of the identified records.
MEBDT in their clinical practice [12]. To date, no systematic Furthermore, the search strategy incorporated scanning
review identifying and summarising the findings of all rel- reference lists and citations of relevant studies as well as
evant primary studies and thereby making the available retrieving unpublished data for the purposes of minimising
evidence accessible to health care practitioners has been publication bias [16]. In addition, a search for ongoing studies
conducted. The presence of conflicting evidence and the lack assessing the diagnostic accuracy of the MEBDT was carried
of studies making a strong impact on clinical practice out. The search process was documented and reported using a
therefore justify the need for an objective analysis of all flow diagram, as recommended by the PRISMA statement [17].
available evidence on the accuracy of the MEBDT.
The aims of this systematic review are to investigate the Data Collection and Quality Assessment
overall DA of the MEBDT in assessing the occurrences of
OPA in adult patients with tracheostomy and consequently An initial screening on the basis of title and abstract of all
define the implications of the findings for research and articles generated by the literature search allowed to
clinical practice. exclude studies that were clearly not meeting the inclusion
criteria. Studies that appeared to meet the inclusion criteria
were assessed for eligibility, and data extraction and
Methods analysis was undertaken by four independent reviewers
(SB, FH, OG and MW). Disagreements were discussed and
Study Selection resolved by consensus. Data for each included study were
recorded on a standard data extraction form. The following
The index test assessed within the scope of this review was items were extracted: (1) publication details (title, authors,
the MEBDT as described by Thompson-Henry and Brad- country where the study was conducted; (2) study design
dock [13]. The chosen reference standard tests were the (prospective or retrospective); (3) patient characteristics (n
videofluoroscopy (VFSS) and fiberoptic endoscopic eval- included; population type, mean age); (4) TP, FP, TN and
uation of swallowing (FEES), as both assessment methods FN from the individual studies; (5) index test (protocol and
are widely used in clinical practice and show good con- type of blue dye used); (6) reference standard tests and time
cordance regarding laryngeal aspiration [14, 15]. All types interval between the latter and the index test; (7) follow-up
of studies assessing the DA of MEBDT against VFSS or duration and reported adverse effects of the blue dye. In
FEES were included, regardless of their publication status. addition, the Standards for the Reporting of Diagnostic
No restrictions were made regarding the data-collection Accuracy Studies (STARD) was used as a checklist for the
design, including both retrospective and prospective stud- reporting of DTA studies [18].
ies. Inclusion criteria comprised tracheostomised adults The risk of bias was evaluated using the revised quality
([18 years) with diagnosed or suspected dysphagia, with- assessment of diagnostic accuracy studies-2 tool [19, 20].
out restrictions concerning the underlying medical cause of Each record was evaluated both in terms of the risk of bias
the tracheostomy. Studies were excluded if (1) they did not and of concerns regarding the applicability of the findings.
provide sufficient data for calculating a 2 9 2 contingency Results of the methodological quality assessment were
table; (2) it was impossible to retrieve data concerning the managed on the Cochrane Review writing RevMan soft-
relevant participant group in case of a mixed population ware [21], and a methodological quality graph was con-
(adults and paediatrics). structed for all the included studies.

123
S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting 723

Statistical Analysis and Data Synthesis identified through the searching process, 6 studies met the
eligibility criteria. Table 1 lists the characteristics of the
The statistical analysis of the data was carried out using the included studies. An overview of the methodological
RevMan software [21]. The type of data was binary, with quality of the included studies is presented graphically in
either positive or negative reporting of the target condition. Figs. 2 and 3. The highest risk of bias involves the patient
The following statistical measures provided in the primary selection, followed by the domain flow and timing. In
study were re-calculated using the contingency table: sen- general, all studies presented low applicability concerns in
sitivity, specificity, predictive values, likelihood ratios and the patient selection and their choice of reference standard.
diagnostic odds ratios. The data of each reference standard However, the applicability concerns were greater with
were subsequently analysed and presented separately. Two regard to the index test, as the lack of information on the
forms of graphical display were used to provide an illustrated MEBDT protocol in some studies limited the transparency
overview of the estimates of sensitivity and specificity of the and consequently their applicability.
primary studies: summary ROC plots and forest plots. The analysis of the methodological quality by means of
the QUADAS-2 tool shows that the methodology used in the
study by Thompson-Henry and Braddock [13] is open to
Results question. The small sample size (n = 5), the poorly articu-
lated inclusion and exclusion criteria as well as the lack of
Characteristics and Methodological Quality information on the subject selection largely reduce the power
of Studies of the studys findings. In addition, the use of two different
reference standard procedures and the lack of consistency in
Figure 1 shows the flow of the literature through the search their use limit the trustworthiness of these findings. Infer-
and screening process. Out of 39 different records ences made during the study were based on two events, with a
Fig. 1 Flow diagram showing
the process of selection of
records and studies for this
review

123
724 S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting

Table 1 Study characteristics

Study n Data Sampling Reference Consistencies trialled Time interval
collection method standard between RST and IT

Brady et al. [23] 21 Prospective Convenience FEES Pureed and solid Simultaneous procedures
consistencies, ice chips
Donzelli et al. [24] 15 Prospective Consecutive FEES Not specified Simultaneous procedures
ONeill-Pirozzi et al. 37 Prospective Unreported MBS Liquid consistencies: thin Simultaneous procedures
[22] liquid; thick nectar;
saliva
Pureed solid consistencies
Peruzzi et al. [8] 20 Prospective Consecutive VFSS Liquid consistencies Consecutive procedures
(water) (immediate)
Thompson-Henry 5 Retrospective Case study VFSS & FEES Liquid and semisolid Consecutive procedures
and Braddock [13] consistencies (422 days)
Winkl-maier et al. 30 Prospective Consecutive FEES Liquid consistencies (water, Consecutive procedures
[25] artificial saliva) (immediate)

Fig. 2 Risk of bias and applicability concerns across included studies as per review authors judgement for each domain presented as
percentages

delay reaching up to 22 days. This methodological flaw

introduces a high risk for misclassification due to recovery or
deterioration of the condition of the patients. Furthermore,
the lack of information on the analysis and interpretation of
the reference standard raises potential for bias and concerns
regarding the investigator objectivity. Owing to these
methodological flaws, the study by Thompson-Henry and
Braddock [13] was excluded from the analysis, graphs and
tables, as its data would distort the results of this review.

Estimates of Diagnostic Accuracy

Pooling of the studies in a meta-analysis was felt to be

inadvisable due to the high diversity of the studies, par-
ticipant samples, and index test protocols of the included
studies.
Both studies comparing the MEBDT to VFSS reported
estimates of sensitivity and specificity [8, 22]. However,
Fig. 3 Summary of the risks of bias and applicability concerns for the sensitivity and specificity values in ONeill-Pirozzi
included studies as judged by the review authors for each domain et al. [22] are incorrectly inversed. Table 2 provides an

123
S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting 725

0.480331 (0.2736

0.615385 (0.4004
overview of the statistical measures of the MEBDT com-

LR- (95 % CI)

pared with VFSS. The prevalence of OPA in the study
samples ranged from 42 to 65 %. However, the high
variability patient characteristics and the overall small

0.8433)

0.9458)
sample sizes of the primary studies reduce the representa-
tivity of the calculated prevalence. Sensitivity and speci-
ficity estimates of the primary studies were highly variable

2.992059 (1.3612
and ranged from 38 to 62 % and from 79 to 100 %,
LR? (95 % CI)

respectively. No procedure emerged with both high sensi-

6.5770) tivity and specificity estimates, although high levels of
Infinity specificity were more commonly reported than high levels
of sensitivity. Comparative analyses of sensitivity and
specificity estimates of the MEBDT versus VFSS are dis-
0.741935 (0.5507

0.466667 (0.2228

played in a coupled forest plot in Fig. 4 and in a SROC plot

NPV (95 % CI)

in Fig. 5.
The above three studies investigated and compared the
0.8746)

0.7258)

estimates of accuracy of the MEBDT versus FEES.

[13, 2325]. The estimates of the accuracy of the MEBDT
compared with FEES are presented in Table 3. The
prevalence of OPA in the study samples ranged from 53 to
0.684211 (0.43498

71 %. Estimates of sensitivity varied across studies and

PPV (95 % CI)

1.00 (0.4629

ranged from 40 to 95 %. All the three studies agree on the

specificity of the MEBDT compared with FEES, reporting
0.8644)

an estimate of 100 %. These results suggest that the

1.00)

MEBDT is highly accurate in diagnosing aspiration when it

is present, rather than excluding it. Comparative analyses
Prevalence (95 % CI)

of sensitivity and specificity estimates of the MEBDT

versus FEES are displayed in a coupled forest plot in Fig. 6
0.42 (0.28489

and in a SROC plot in Fig. 7.

0.65 (0.4094

Overall, the reported low sensitivity of the MEBDT

0.5673)

0.8369)

indicates its inaccuracy in unfailingly identifying patient

with OPA. The observed high true negative rate, on the
other hand, seems to suggest that MEBDT is reliable in
identifying patients without OPA. Extrapolation of these
Table 2 Estimates of accuracy of the MEBDT compared with VFSS

1.00 (0.56091.00)
0.793103 (0.5974
Spec (95 % CI)

findings into clinical settings adverts that a clinician can

rely on the positive results of the MEBDT, indicating the
patient most certainly has OPA.
0.9129)

Discussion
0.619048 (0.3869

0.384615 (0.1513
Sens (95 % CI)

No definite conclusions could be drawn regarding the

accuracy of the MEBDT in detecting OPA in adult patients
0.8105)

0.6772)

with tracheostomy. The completeness and applicability of

evidence was limited due to the paucity of primary studies
investigating the DA of the MEBDT. Moreover, the
50

20

strength of results yielded in this review is limited by the

n

discrepancies in MEBDT protocols used across the inclu-

Peruzzi et al. [8]

ded studies.
ONeill-Pirozzi

The lack of a universally accepted standard MEBDT

et al. [22]

protocol led to high variations across the studies tech-

Study

niques in the administration of the index test. This review

shows significant differences in the amount and

123
726 S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting

Fig. 4 Coupled forest plot of the estimates of sensitivity and specificity in VFSS studies. TP true positive, FP false positive, FN false negative,
TN true negative

Fig. 5 Summary ROC plot of

estimates of the accuracy of
MEBDT compared with VFSS

consistency of the administered boluses as well as in the Preliminary evidence emerging from this review sug-
nature and amount of blue dye used to tint the test material. gests that the MEBDT was generally better at excluding
These discrepancies may account for the high variability in OPA, with all studies reporting higher specificity values.
the reported accuracy measures, which consequently makes OPA-positive findings as assessed by the MEBDT-
a comparison of the studies findings difficult. Considering screening procedure thus provide an accurate basis for
the paucity of primary studies assessing the MEBDT and diagnosis, as the patient passed the exclusion test. Nev-
the lack of consensus regarding their sensitivity and ertheless, it is important to stress that all studies reported
specificity values, there is a definite need for further low sensitivity values, indicating that the MEBDT is
research in that area. It is therefore essential for further associated with potential false negative results. As a result,
research to provide large and well-designed primary diag- clinicians are not advised to use the MEBDT in order to
nostic accuracy studies using a standardised MEBDT rule-in OPA. However, the paucity of primary studies
protocol in order to yield generalizable DA measures. investigating the MEBDTs accuracy and their conflicting

123
S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting 727

0.047619 (0.00700.32245)
findings fail to make a definite statement about the relia-

0.60 (0.396920.90699)
0.50 (0.2500490.9998)
bility of the MEBDT as a screening tool. Nevertheless, the
use of a colouring agent might be useful as part of a

LR- (95 % CI)

comprehensive bedside swallowing assessment of patients
with artificial airways, or in conjunction with other
instrumental assessment tools. If a patient demonstrates a
positive MEBDT, the timing of further instrumental
LR? (95 % CI)
swallowing evaluations may be postponed until the patient
is able to pass the MEBDT [23]. This can provide SLTs
with baseline information, and give early indications on the
Infinity
Infinity
Infinity patients deglutition ability. This procedure may be of
value in clinical settings, where instrumental assessment is
not readily available. However, the MEBDT should be
0.40 (0.17460.6711)
0.636364 (0.31610.8763)
0.90 (0.54120.9948)

used with caution, as there is insufficient evidence to

support or reject the MEBDT as a sole screening tool.
NPV (95 % CI)

Furthermore, it is important to acknowledge that the use

of FD&C blue no.1. has become highly controversial. The
FDA alerted healthcare professionals of several clinical
cases of toxicity, including death, associated with the use
of FD&C blue no.1 in enteral feeding solutions [26]. None
1.00 (0.51681.00)
1.00 (0.39581.00)
1.00 (0.79951.00)

of the included studies address the possible adverse effects

PPV (95 % CI)

of the blue dye. The different protocols of the studies do

not include a long-term follow-up after the MEDBT pro-
cedure, which precluded an analysis of possible compli-
cations due to blue dye absorption within this review. It
0.533333 (0.274230.7772)
0.714286 (0.47690.8781)

0.70 (0.5044210.845872)

would therefore be of interest for future research studies to

monitor the subjects undergoing the MEBDT with FD&C
Prevalence (95 % CI)

blue no.1 on a long-term basis, in order to provide infor-

mation on the possible adverse effect of the blue dye.
Outcomes from such studies will update the current level of
evidence in relation to the MEBDT and consequently
define best clinical practice.
1.00 (0.51681.00)
1.00 (0.56091.00)
1.00 (0.62881.00)
Spec (95 % CI)

Limitations of the Review

Table 3 Estimates of accuracy of the MEBDT compared with FEES

Because of the foreseen paucity of primary studies inves-

tigating the diagnostic accuracy of the MEBDT in tra-
cheostomised adults, wide inclusion criteria were set for
0.40 (0.17460.6711)
0.5 (0.17450.8255)
0.95238095 (0.74130.9975)

this review. As a consequence, the included studies were

subject to heterogeneity and low overall methodology [27].
Therefore, a major limitation of this review process con-
Sens (95 % CI)

stitutes the impossibility of pooling the studies together in a

meta-analysis in order to yield one overall effect estimate.
In addition, the use of multiple reference standards to
determine the diagnostic accuracy of the MEBDT consti-
tutes a limitation of the review, as it may result in biased
21
15
30
n

estimates of the accuracy results if the findings are treated

Winkl-maier et al. [25]

as interchangeable [28]. Practical or ethical constraints

Donzelli et al. [24]

may lead researchers to use an alternative and less accurate

Brady et al. [23]

reference standard instead of the preferred reference stan-

dard to assess the disease status of a specific population.
This may consequently lead to differential verification
Study

[29, 30].

123
728 S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting

Fig. 6 Coupled forest plot of the estimates of sensitivity and specificity in FEES studies. TP true positive, FP false positive, FN false negative,
TN true negative

Fig. 7 Summary ROC plot of

the estimates of accuracy of
MEBDT compared with FEES

Conclusions limited by the discrepancies in MEBDT protocols used

across studies. As a consequence, empirical evidence for or
This review critically examined evidence pertaining to the against the use of the MEBDT is not strong enough to
effectiveness of the MEBDT in detecting OPA in tra- direct clinical practice. Therefore, there is a strong neces-
cheostomised patients. The results yielded in this review, sity for further well-designed diagnostic accuracy studies
however, indicate that the MEBDT generally presents with using a standardised protocol to bridge the gap between
higher specificity values, and thus seems better at exclud- clinical practice and documented evidence.
ing OPA. Furthermore, it is noteworthy to stress the low
Compliance with Ethical Standards
sensitivity of the MEBDT. Clinicians should therefore use
the MEBDT screening procedure with caution, as it is Conflict of Interest There are no relevant disclosures or conflict of
associated with high risk of false negative results. How- interest to declare for the content of this review.
ever, the strength of the findings reported in this review is

123
S. Bechet et al.: Diagnostic Accuracy of the Modified Evans Blue Dye Test in Detecting 729

References the PRISMA statement. Ann Intern Med. 2009;151:2649. doi:10.

7326/0003-4819-151-4-200908180-00135.
1. Duann CW, Hsieh MS, Chen PT, Chou HP, Huang CS. Suc- 18. Bossuyt MMP, Reitsma BJ, Bruns ED, et al. Towards complete
cessful percutaneous tracheostomy via puncture through the and accurate reporting of studies of diagnostic accuracy: the
STARD initiative. Clin Chem. 2003;49:16. doi:10.1373/49.1.1.
thyroid isthmus. Respirol Case Rep. 2004;2:5760. doi:10.1002/
rcr2.48. 19. Whiting P, Rutjes WSA, Reitsma BJ, Bossuyt MMP, Kleijnen J.
2. Kluge S, Baumann HJ, Maier C, et al. Tracheostomy in the inten- The development of QUADAS: a tool for the quality assessment
sive care unit: a nationwide survey. Anesth Analg. 2008;107: of studies of diagnostic accuracy included in systematic reviews.
BMC Med Res Methodol. 2003;10(3):25. doi:10.1186/1471-
163943. doi:10.1213/ane.0b013e318188b818.
3. Young D, Harrison DA, Cuthbertson BH, Rowan K, Collabora- 2288-3-25.
tors TracMan. Effect of early vs late tracheostomy placement on 20. Whiting P, Rutjes WSA, Westwood EM, et al. QUADAS-2: a
survival in patients receiving mechanical ventilation: the tracman revised tool for the quality assessment of diagnostic accuracy
randomized trial. JAMA. 2013;309:21219. doi:10.1001/jama. studies. Ann Intern Med. 2011;155:52936. doi:10.7326/0003-
2013.5154. 4819-155-8-201110180-00009.
4. Baskin JZ, Panagopoulos G, Parks C, Rothstein S, Komisar A. 21. Review Manager (RevMan) (computer program) Version 5.3.
Clinical outcomes for the elderly patient receiving a tracheotomy. Copenhagen: The Nordic Cochrane Centre, The Cochrane Col-
Head Neck. 2004;26:715. laboration; 2014.
5. Durbin CG. Tracheostomy: why, when, and how? Respir Care. 22. ONeill-Pirozzi MT, Lisiecki JD, Momose KJ, Connors JJ, Mil-
2010;55:105668. liner PM. Simultaneous modified barium swallow and blue dye
tests: a determination of the accuracy of blue dye test aspiration
6. Garuti G, Reverberi C, Briganti A, Massobrio M, Lombardi F,
Lusuardi M. Swallowing disorders in tracheostomised patients: a findings. Dysphagia. 2003;18:328. doi:10.1007/s00455-002-
multidisciplinary/multiprofessional approach in decannulation 0081-4.
protocols. Multidiscip Respir Med. 2014;9:110. doi:10.1186/ 23. Brady LS, Krieger R, Wesling M, Scott K, Donzelli J, Pietrantoni
2049-6958-9-36. M. Sensitivity of the blue dye food test for detecting aspiration in
7. Jung SJ, Kim DY, Kim YW, Koh YW, Joo SY, Kim ES. Effect of patients with a tracheotomy. BJAST. 2015;7:42935. doi:10.
decannulation on pharyngeal and laryngeal movement in post- 9734/BJAST/2015/11652.
stroke tracheostomized patients. Ann Rehabil Med. 2012;36: 24. Donzelli J, Brady LS, Wesling M, Craney M. Simultaneous
35664. doi:10.5535/arm.2012.36.3.356. modified Evans blue dye procedure and video nasal endoscopic
8. Peruzzi WT, Logemann JA, Currie D, Moen SG. Assessment of evaluation of the swallow. Laryngoscope. 2001;111:174650.
aspiration in patients with tracheostomies: comparison of the 25. Winklmaier U, Wust K, Plinkert PK, Wallner F. The accuracy of
the modified Evans blue dye test in detecting aspiration in head and
bedside colored dye assessment with videofluoroscopic exami-
nation. Respir Care. 2001;46:2437. neck cancer patients. Eur Arch Otorhinolaryngol. 2007;264:
9. Singh S, Hamdy S. Dysphagia in stroke patients. Postgrad Med J. 105964. doi:10.1007/s00405-007-0299-8.
2006;82:38391. doi:10.1136/pgmj.2005.043281. 26. Acheson D (2003) Reports of blue doscoloration and death in
patients receiving enteral feedings tinted with the dye FD&C blue
10. Deutsche Gesellschaft fur Neurologie. Neurogene Dysphagien.
In: Diener H-C, Weimar C, editors. Leitlinien fur diagnostik und no. 1. FDA Public Health Advisory. http://www.fda.gov/ForIndu
therapie in der neurologie. Stuttgart: Thieme Verlag; 2012. stry/ColorAdditives/ColorAdditivesinSpecificProducts/InMedical
11. NHS Health Scotland. Caring for the patient with a tracheostomy: Devices/ucm142395.htm. Accessed 27 May 2015.
best practice statement. Scotland: NHS quality improvement. 27. Kleijnen J, Westwood M, Whiting P. Applicability of diagnostic
2007. http://www.nhshealthquality.org/nhsqis/files/TRACHEO studies: statistics, bias and estimates of diagnostic accuracy.
REV_BPS_MAR07.pdf. Z Evid Fortbild Qual Gesundheitswes. 2011;105:498503.
12. Downes LS. Clinical consistency among speech and language doi:10.1016/j.zefq.2011.07.025.
therapists in the management of tracheostomies in the Republic 28. Naaktgeboren AC, de Groot J, van Smeden M, Moons K, Reitsma
of Ireland. Dublin: Department of Clinical Speech and Language BJ. Evaluating diagnostic accuracy in the face of multiple ref-
Studies, Trinity College Dublin; 2012. erence standards. Ann Intern Med. 2013;159:195202. doi:10.
7326/0003-4819-159-3-201308060-00009.
13. Thompson-Henry S, Braddock B. The modified Evans blue dye
procedure fails to detect aspiration in the tracheostomized patient: five 29. Lijmer JG, Mol BW, Heisterkamp S, et al. Empirical evidence of
case reports. Dysphagia. 1005;10:1724. doi:10.1007/BF00260973. design-related bias in studies of diagnostic tests. JAMA. 1999;
14. Brady S, Donzelli J. The modified barium swallow and the 282:10616. doi:10.1001/jama.2015.7649.
functional endoscopic evaluation of swallowing. Otolaryngol 30. Rutjes AW, Reitsma JB, Di Nisio M, Smidt N, van Rijn JC, Bossuyt
Clin North Am. 2013;46:100922. doi:10.1016/j.otc.2013.08. PM. Evidence of bias and variation in diagnostic accuracy studies.
001. CMAJ. 2006;174:46976. doi:10.1503/cmaj.050090.
15. Lee W-S, Lee S-H, Park YH, et al. Effectiveness of fiberoptic
endoscopic evaluation of swallowing (FEES) in patients with
Sibylle Bechet MSc
swallowing disorders: a systematic review. Brain Neurorehabil.
2008;6:916. Fiona Hill MSc
16. de Vet HCW, Eisinga A, Riphagen I, Aertgeertes B, Pewsner D.
Chapter 7: Searching for studies. Cochrane handbook for sys- Orla Gilheaney MSc
tematic reviews of diagnostic test accuracy version 0.4. Oxford: Margaret Walshe PhD
The Cochrane Collaboration; 2008.
17. Moher D, Liberati A, Tetzlaff J, Altman GD, The PRISMA Group.
Preferred reporting items for systematic reviews and meta-analyses:

123