Outcomes and Reliability of Flow Coupler in Head and Neck Free Flaps

This retrospective single-institution study, conducted from April 2015 to March 2017, evaluated the accuracy and reliability of the flow coupler compared to the arterial Doppler probe in postoperative monitoring of head and neck free flaps. The study included 120 consecutive cases where both devices were employed. The results indicated that the flow coupler had a sensitivity of 100%, specificity of 86.4%, and a false positive rate (FPR) of 13.6%, particularly effective for venous thrombosis detection, while the arterial probe showed lower sensitivity (9.1%) but higher specificity (97.1%). The salvage rate was 81.8% for the cases that required takeback. The FPR of the flow coupler decreased with increased usage. The study concludes that the flow coupler is a valuable tool for postoperative monitoring of head and neck free flaps, especially for detecting venous thrombosis. The report highlights the importance of supplementing flow coupler monitoring with physical examination to minimize unnecessary interventions and improve patient outcomes. The findings underscore the superior performance of the flow coupler, particularly in identifying venous compromise, and its potential to enhance the management of free flap complications.

The Laryngoscope
VC 2017 The American Laryngological,
Rhinological and Otological Society, Inc.
Outcomes and Reliability of the Flow Coupler in Postoperative
Monitoring of Head and Neck Free Flaps
Rance J. T. Fujiwara, BS ; Jacqueline M. Dibble, APRN; Scott V. Larson, MD; Matthew L. Pierce, MD;
Saral Mehra, MD, MBA
Objectives/Hypothesis:To assess the accuracy and reliability ofthe flow coupler relative to the implantable arterial
Doppler probe in postoperative monitoring of head and neck free flaps.
Study Design: Retrospective single-institution study,April 2015 to March 2017.
Methods: Both the venous flow coupler and arterial Doppler were employed in 120 consecutive head and neck free fl
cases.When Doppler signal loss occurred,flaps were evaluated by physical exam to determine whether signal loss was a true
positive necessitating operating room takeback.Sensitivity,specificity,and false positive rate (FPR) were recorded for each
device.Logistic regression was conducted to identify user trends over time.
Results: Eleven of 120 patients (9.2%) required takeback,10 from venous thrombosis and one from arterial thrombosis.
Permanent signalloss (PSL) occurred in the flow coupler in alltakebacks;PSL occurred in the arterialDoppler only in the
case of arterial thrombosis.Salvage rate was 9/11 (81.8%).For the flow coupler,sensitivity was 100%,specificity 86.4%,and
FPR 13.6%.For the arterialprobe,sensitivity was 9.1%,specificity 97.1%,and FPR 2.9%.A 4.1% decrease in false positives
with each additional flow coupler use was observed.
Conclusions:Monitoring the vein via flow coupler has high sensitivity in identifying vascular compromise compared to
the arterialprobe,especially for venous thrombosis.There is moderate FPR;this decreases with increased usage and,when
supplemented with physicalexamination,does not result in unnecessary takebacks.The flow coupler can be a valuable tool
in postoperative monitoring of head and neck free flaps.
Key Words: Venous flow coupler,cancer of head and neck,microsurgical free flap,sensitivity and specificity.
Level of Evidence: 4.
Laryngoscope,00:000–000,2017
INTRODUCTION
Advancements in free flap reconstruction have fur-
thered our ability to restore function and increased
treatment options for advanced head and neck cancers.
Due to these advancements,overall flap success rates
exceed 95% at multiple institutions.1,2 However, free flap
complications,most commonly by vascular thrombosis,
can have devastating effects on morbidity and functional
prognosis.2–4 For vascular thrombosis,early recognition
of vascular compromise is associated with improved sal-
vage rates.5–7 In a single institution study, successful
flap revisions were initiated 46.5 hours after initial
reconstruction,compared to 82.0 hours for failed revi-
sions (P <.05).7
For this reason, there have been efforts to develop
methods to ensure adequate free flap monitoring,espe-
cially in the immediate postoperative period.8 The Cook-
Swartz Doppler (Cook Medical,Bloomington,IN) probe
is used in free flap reconstructions to evaluate the vas-
cular flow of the flap. It is commonly placed on the
artery due to increased specificity and decreased false
positive rates compared to venous placement.9,10 In the
largest series utilizing implantable Doppler probes in
head and neck reconstruction,Wax et al. stopped using
the Cook-Swartz Doppler for venous monitoring due to
unreliability.11 Although forgoing venous monitoring has
resulted in decreased unnecessary operating room (OR)
takebacks,studies have shown that venous thromboses
are the most common cause of vascular compromise,12–14
and arterial Doppler devices can detect arterialflow up
to 6 hours after venous compromise.15
In 2010, Synovis Life Technologies,Inc. (St. Paul,
MN) introduced the flow coupler, combining a venous
anastomoticcoupler and 20 MHz ultrasonic Doppler
probe to assess patency atthe anastomosis.Currently,
From the Department of Surgery (Otolaryngology)(R.J .T.F., S.M.),
Yale University School of Medicine, New Haven, Connecticut;Depart-
ment of Surgery (Otolaryngology) (J .M.D., M.L.P.), Yale–New Haven Hospi-
tal, New Haven, Connecticut; Department of Ear, Nose, and Throat
(S.V.L .), Benefis Hospitals, Great Falls, Montana; and the Yale Cancer
Center (S.M.), New Haven, Connecticut, U.S.A
Editor’s Note: This Manuscript was accepted for publication on
September 5, 2017.
Presented at the 2017 American Head and Neck Society Meeting
at the Combined Otolaryngology Spring Meetings, San Diego, California,
U.S.A., April 26–30, 2017.
R.J .T.F. had full access to alldata in the study and takes responsi-
bility for the integrity of the data and the accuracy of the data analysis.
Author contributions include conception and design (R.J .T.F., J .M.D., S.V.L .,
M.L .P., S.M.), analysis and interpretation of data (R.J .T.F., J .M.D., S.V.L .,
M.L .P., S.M.), drafting of the manuscript (R.J .T.F., S M ), critical revision for
important intellectual content (R.J .T.F., J .M.D., S.V.L ., M.L .P., S.M.), statistical
analysis (R.J .T.F., S.M.), and supervision (J .M.D., S.V.L., M.L .P.).
The authors have no funding, financial relationships,or conflicts
of interest to disclose.
Send correspondence to SaralMehra, MD, MBA, Yale Physicians
Building, 800 Howard Avenue, 4th Floor, New Haven, CT 06519. E-mail:
[e-mail]
DOI: 10.1002/lary.26944
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
1

limited analysis exists on the use of the flow coupler,
especially in head and neck reconstruction.16 In breast
reconstruction,one single institution study showed no
significant difference in false positive rate, false negative
rate, takeback or flap failure rate for the flow coupler
compared to the Cook-Swartz Doppler when monitoring
venous outflow, whereas another noted 36% positive pre-
dictive value and increased vascular thrombotic
events.17,18 The objectives of this study were to measure
the accuracy and reliability of the flow coupler in the
postoperative monitoring ofhead and neck free flaps,
compare its outcomesto arterial monitoring via the
Cook-Swartz Doppler probe, and analyze its performance
over time.
MATERIALS AND METHODS
This retrospective cohort study consisted of120 patients.
This study was approved for exemption by the Yale University
Institutional Review Board. We reviewed the records of all
patients who had undergone head and neck free flap recon-
struction between April 2015 and March 2017 at Yale–New
Haven Hospital by one reconstructive surgeon.Both the Syno-
vis Flow Coupler and Cook-Swartz Doppler were employed in
120 consecutive cases to monitor the patency of the venous and
arterial anastomoses,respectively.All patients underwent flap
checks according to the Yale Head and Neck free flap protocol.
Patient demographics,operative details,flap outcomes,compli-
cations,and postoperative venous and arterialDoppler signals
were obtained.
Venous and arterial Doppler signals were monitored inde-
pendently. When signal was lost, the flap was immediately eval-
uated by physical exam; dark or no bleeding on needle prick
was considered a positive event suggestive of vascular compro-
mise necessitating OR takeback. Sensitivity, specificity,and
false positive rate (FPR) were recorded independently for the
flow coupler and arterial Doppler to assess the accuracy of each
probe (Fig. 1). Six cases were excluded due to accidental
removal of the arterial and venous probes by medicalstaff or
the patient.The quality of venous and arterial Doppler signals,
OR findings, and salvage rates were recorded for cases necessi-
tating OR takeback. False positives,accidentalremovals,and
intentional probe removalwere trended over time for the flow
coupler.Variables associated with a false positive were analyzed
using v2 or univariate logistic regression analysis.Variables with
P <.20 in univariate analysis were subsequently included in bino-
mial logistic regression.Because a large fraction of flaps (59.2%)
had two venous anastomoses by flow coupler, subset analysis was
conducted for patients with one venous anastomosis.Statistical
analysis was performed using Stata 14 (StataCorp,College Sta-
tion, TX).
RESULTS
Patient and Flap Demographics
Patient characteristics for 120 patients are detailed
in Table I. Mean age was 62.9 years (range, 21–94
years), and 66.7% were male.The majority of patients
were former or active smokers (68.3%), and 22.5% had a
history of radiation therapy. Free flap reconstruction
was indicated primarily following composite resections
or resection of oral cavity malignancies (65.8%).Average
length of stay was 10.1 days (range, 5–29 days).
Flap characteristics are detailed in Table II. Sixty
(50.0%) had radial forearm free flaps; 28 (23.3%)
received fibula free flaps.Two venous anastomoses were
monitored by flow coupler in 71 (59.2%) cases. The facial
artery was used as a recipient artery in 80 patients
(66.7%);the externaljugular vein (42.7%) was the most
common recipient vein.The median venous flow coupler
size was 2.5 mm (range, 1.5-4.0 mm).
Flow Coupler Versus Cook-Swartz Doppler
Eleven patients (9.2%) required OR takeback due to
vascular compromise (Table III).Each of these cases is
detailed in Table IV. Flap checks in 10 cases demon-
strated findings positive for vascular compromise neces-
sitating takeback. One patient had a buried flap and
was taken to the OR based only on change in the venous
Doppler signal. Venous thrombosis was discovered intra-
operatively in 10 flaps and arterial thrombosis in one
Fig. 1. Classification ofDopplersignals afterphysicalexam and
calculations ofstatisticalmetrics for each Doppler device.A true
positive (TP)was defined as a physicalexam positive for vascular
compromise following Dopplersignalloss, whereas a false posi-
tive (FP)was defined as a normalphysicalexam with Doppler sig-
nal loss. A true negative (TN)occurred with a normalphysical
exam and strong Dopplersignals,whereas a false negative (FN)
occurred when a Doppler signal remained strong despite a
remarkable physicalexam. FPR 5 false positive rate.
TABLE I.
Patient Characteristics
Characteristic Value
No. of patients 120
Age, yr, mean (SD) 62.9 (12.0)
Male, no. (%) 80 (66.7)
Smoking history, no. (%)
Never 38 (31.7)
Former 65 (54.2)
Active 17 (14.2)
History of radiation, no. (%) 27 (22.5)
Length of stay, d, mean (SD) 10.1 (4.7)
Site of defect, no. (%)
Composite 30 (25.0)
Oralcavity 49 (40.8)
Oropharynx 10 (8.3)
Hypopharynx/larynx 3 (2.5)
Mandibular osteoradionecrosis 4 (3.3)
Scalp 5 (4.2)
Maxilla 5 (4.2)
Parotid 3 (3.4)
Other 11 (9.2)
SD 5 standard deviation
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
2

flap. The salvage rate was 81.8%.Total flap failure rate
was 0.8%. Among 49 patients with one flow coupler,
seven cases required OR takeback (14.3%);the salvage
rate was 71.4%.
Permanent signalloss of the flow coupler occurred
in all 11 cases requiring takeback.In the flap compli-
cated by arterial thrombosis,a permanent signal loss
also occurred in the arterial Doppler. In the 10 cases
complicated by venous thrombosis,the arterial Doppler
signal remained strong at the time when signal loss
occurred in the flow coupler, during the time of OR
venous exploration,and after flap salvage.The sensitiv-
ity of the arterial Doppler was 9.1%,specificity 97.1%,
and FPR 2.9%. Among patients with one flow coupler,
the sensitivity of the arterial Doppler was 14.3%,specif-
icity 97.4%, and FPR 2.6%.
In contrast, the sensitivity of the flow coupler was
100.0%, specificity 86.4%, and FPR 13.6%. Among patients
with one flow coupler,the sensitivity for the flow coupler
was 100%,specificity 82.1%,and FPR 17.9%.In contrast,
among patients with two flow couplers,sensitivity was
100%, specificity 89.1%, and FPR 10.9%.
Venous Flow Coupler Trends
The accuracy,FPR, and rate of accidentalremoval
over time are presented in Figure 2.Eight patients had
at least one venous Doppler wire accidentally removed
by medical staff; among these, six patients also had their
arterial Doppler wire accidentally removed. Five of these
six accidental removals occurred during the first 4
months of usage.
No patient or flap factors were associated with
increased false positives from the flow coupler on univar-
iate analysis, including use of two vein anastomoses
(results not shown).However,on multivariate analysis,
there was a statistically significant 4.1% decrease in
likelihood of a false positive signalloss with each addi-
tional case of flow coupler use (odds ratio: 0.96, 95% con-
fidence interval [CI]:0.94-0.98, P 5.001).This trend was
no longer significant (odds ratio: 0.97, 95% CI: 0.94-1.00,
P 5.09) on subset analysis of patients with one flow
coupler.
DISCUSSION
In this study, we utilized the flow coupler and
implantable Doppler to monitor venous outflow and arte-
rial inflow in 120 patients undergoing head and neck
reconstruction.We found that postoperativevascular
compromise occurred in 11/120 (9.2%) flaps,10 of which
were venous.Permanent signal loss occurred in the flow
coupler in all 11 cases, whereas the arterial Doppler sig-
nal remained strong in all cases with venous thrombosis.
The sensitivity and specificity for the venous flow cou-
pler were 100% and 86.4%,respectively.For the arterial
Doppler, the sensitivity and specificity were 9.1% and
97.1%,respectively.We also found a statistically signifi-
cant decrease in frequency of false positives with
increased usage of the flow coupler.
To our knowledge,this is the first study to analyze
the flow coupler in which each flap was simultaneously
monitored at the vein and artery by a flow coupler and
arterial Doppler probe.Previous studies have compared
groups of patients receiving one of the two devices,with
only one vesselmonitored.In 2014 Um et al. analyzed
TABLE II.
Flap Characteristics
Flap Value
Radialforearm 60 (50.0)
Fibula 28 (23.3)
Anterolateralthigh 15 (12.5)
Scapula 7 (5.8)
Latissimus dorsi 6 (5.0)
Other (rectus, scapula 1 latissimus dorsi) 4 (3.3)
Two venous anastomoses 71 (59.2)
Flow coupler size, median (IQR) 2.5 (2.5–3.0)
Artery
Facial 80 (66.7)
Superior thyroid 21 (17.4)
Lingual 6 (5.0)
Externalcarotid (or branch) 8 (6.7)
Superficialtemporal 3 (2.5)
Transverse cervical 2 (1.7)
Vein
Externaljugular 85 (42.7)
Common facial 53 (26.6)
Branch of common facial 16 (8.0)
Facial 16 (8.0)
Superior thyroid 10 (5.0)
Internaljugular (or branch) 10 (5.0)
Anterior jugular 6 (3.0)
Other 3 (1.5)
Data are presented as number (%) unless otherwise indicated.
IQR 5 interquartile range.
TABLE III.
SignalLoss and Operating Room Takebacks for Venous Flow Coupler Versus Implantable ArterialDoppler Probe, Excluding Patients With
AccidentalRemovalof Both Devices.
Venous Synovis Flow Coupler ArterialCook-Swartz Doppler
Required Takeback No Takeback Total Required Takeback No Takeback Total
Signalloss 11 14 25 Signalloss 1 3 4
No signalloss 0 89 89 No signalloss 10 100 110
Totals 11 103 114 Totals 11 103 114
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
3

220 free flap breast reconstructions in which 109
received the venous Cook-Swartz probe and 110 received
the flow coupler.18 Sensitivity was 100% for both groups;
FPR for the Cook-Swartz Doppler and flow coupler was
1.0% and 1.9%, respectively,and no statistical difference
was found in the rate of OR takebacks and flap failure
rates. The authors concluded that the flow coupler was
as effective as the Cook-Swartz Doppler in monitoring
the venous system.In a similar 2015 study in free flap
breast reconstructions,Kempton et al. demonstrated
100% sensitivity for the flow coupler in the intra- and
postoperative settings and 36% positive predictive value,
which was significantly increased compared to the
venous Cook-Swartz Doppler.17 However,they also dem-
onstrated an increased rate of vessel thrombosis in
patients with the flow coupler (17.6% vs. 2.9%, P 5.038).
Our study has similarly demonstrated 100% sensitivity
for the venous flow coupler, with an 8.3% venous throm-
bosis rate. These rates are comparable to the reconstruc-
tive surgeon’s venous thrombosis rates prior to use of
the flow coupler.
Recent literature in head and neck reconstruction
has described a trend toward arterial monitoring due to
decreased false positives and unnecessary OR re-explo-
rations.8,11 Doppler signal fluctuations in the head and
neck have been attributed to position changes resulting
in flow alterations, Doppler probe wire movement,or
dislodgement.8,19 In a retrospective review of 439
patients undergoing head and neck,breast,or extremity
free flap reconstruction,Chang et al. found that arterial
implantable Doppler device monitoring was associated
with a significant increase in sensitivity (77.8% vs.
66.7%) and specificity (97.2% vs. 73.7%) compared to
monitoring the vein.9 In 2014, Wax found a sensitivity of
87% and specificity of 99% for 1,142 patients when moni-
toring the artery; however, there were 10 false negatives
not associated with changes in Doppler signal.11 Guille-
maud et al. had a sensitivity of 65.8% and specificity of
98.2% for detecting flap compromise using primarily
TABLE IV.
Cases Requiring Operating Room Takeback
Patient Defect Flap Type Recipient Vein(s)
Coupler
Size (mm)
ArterialSignal
When Venous
SignalLost
POD of
Takeback Complication Salvaged
1 Oraltongue,
FOM
RFFF EJ, CF 2, 2.5 Strong 1 VT Yes
2 Oraltongue,
FOM RFFF CF 2 Strong 4 VT No
3 Radical
parotidectomy
ALT AJ (branch) 2.5 Absent (lost on
POD 0)
1 AT No
4 Oraltongue,
FOM, mandible
FFF (buried) CF 3 Strong 2 VT Yes
5 Buccal
mucosa, FOM,
RMT, mandible
FFF EJ 4 Strong 1 VT Yes
6 RMT, mandible FOCFF CF, CF (branch) 2.5, 3.5 Strong 0 VT Yes
7 Soft palate RFFF IJ (branch), EJ 2.0, 2.5 Strong 0 VT Yes
8 FOM, AR FFF EJ 3.5 Strong 0 VT Yes
9 FOM, oral
tongue, tongue
base
FFF C, IJ (branch) 2.5, 2.5 Strong 3 VT Yes
10 ORN FFF IJ 3.5 Strong 1 VT Yes
11 FOM, oral
tongue RFFF EJ 2.5 Strong 1 VT Yes
AJ 5 anteriorjugular;ALT 5anterolateralthigh free flap;AR 5 alveolarridge; AT 5arterialthrombosis;C 5 cephalic;CF 5 common facial;EJ 5 external
jugular; FFF 5 fibula free flap; FOCFF 5 fibula osteocutaneous free flap; FOM 5 floor of mouth; IJ 5 internal jugular; ORN 5 osteoradionecrosis;
POD 5 postoperative day; RFFF 5 radialforearm free flap; RMT 5retromolar trigone; VT 5 venous thrombosis.
Fig. 2. Time trends in venous flow coupler outcomes.Of the 120
patients who received both arterialand venous Dopplerprobes,
11 had true permanent signalloss resulting in operating takeback
(true positive [TP]),89 were removed attime of discharge with
strong Doppler signaland were considered true negative (TN),14
had permanentsignal loss without concerning changes on flap
checks and were considered false positives (FP),and six had
accidentalremovalof allvenous flow couplers (accidentalremoval
[AR]).
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
4

arterial Doppler probes, but 13/38 operative explorations
with flap compromise were not associated with Doppler
signal loss.10 Four of those cases were due to venous
obstruction. There was also a significant increase in
mean time to operative exploration (4.99 vs.7.67 hours)
and decrease in flap salvage rate (92.0% vs.61.5%)for
cases withoutchange in Doppler signal.Both Wax and
Guillemaud found that the Cook-Swartz implantable Dopp-
ler placed on the vein in head and neck reconstruction was
unreliable and, as described by Wax, “frustrating.”10,11
Despite FPR at 13.6%,our analysis has shown that
monitoring venous flow via the flow coupler has
improved sensitivity compared to monitoring the artery
both for single- and double-vein monitoring, especially in
cases of venous thrombosis. Moreover, because each
patient had both a venous and arterial Doppler, each
patient effectively functioned as his/her own control,
allowing a direct comparison between the venous and
arterial signals at the time of a permanent signal loss or
positive event.Regarding the moderate FPR associated
with venous monitoring,a physical exam with pin prick
was utilized following Doppler signal loss to ensure that
false positives were not re-explored in the OR unneces-
sarily. The purpose ofa screening test in any setting is
to identify all positive cases with high sensitivity and
low false negative rate, such that a positive screening
test prompts a more specific test to rule out false posi-
tives. In the case of vascular compromise,it is crucial
that true positives be identified quickly,and that false
negatives be minimized to optimize salvage rates.7,20
Paydar in 2010 similarly argued that accepting
decreased sensitivity from arterialDoppler monitoring
fails to detect compromised flaps early enough to per-
form revisions.12 Although arterial Doppler signal loss
may eventually have occurred in our observed cases of
venous thrombosis,the ability of the flow coupler to
detect these cases more quickly ensured prompt return
to the OR for flap revision. In fact, even in the case of
venous thrombosis thatwas not salvaged,the patient
was brought to the OR in a timely fashion, and the
venous anastomosis was opened with resultant flow re-
established;there was subsequentvenous thrombosis
and the decision was made to remove the free flap and
place a supraclavicular artery island regional flap
instead.In all 10 cases of flap compromise with venous
flap thrombosis in this series, return to the OR was
accomplished prior to any visible external changes to the
skin paddle; inability to salvage was not becauseof
delayed return to the OR, but rather lack of suitable
recipient vein or another not easily modifiable pedicle
problem.
In the present series, the FPR for the venous and
arterial Doppler probes were 13.6% and 2.9%,respec-
tively. However,the FPR rate for the venous flow cou-
pler decreased from 26.0% in the first half of the series
to 1.9% in the latter half.A learning curve likely exists
for the reconstructive surgeon,OR staff, and medical
staff in using the venous flow coupler both intraopera-
tively and postoperative, possibly contributing to the ele-
vated FPR. Our analysis confirmed improved accuracy,
decreased false positives, and decreased accidental
removals of the flow coupler Doppler probes over time. A
statistically significant 4.1% decrease in the likelihood of
a false positive with each additional usage was observed,
though this trend was not statistically significant on
subset analysis ofpatients with one flow coupler,likely
due to sample size. With increased usage,we found a
number of technical improvements that helped to
decrease accidentalremovals and facilitate usage of the
flow coupler in head and neck reconstruction.First, the
flow coupler wire can cause kinking of the vein,and we
found that secure and stable wire positioning in the
neck is important to reduce twisting of the vein related
to wire position. Second, securing the wires and external
apparatus to the mastoid region with Tegaderm (3M, St.
Paul, MN) dressings and skin staples has been the most
reliable way to prevent accidentalremovals by patient
head movementand nursing staff. Third, reliable and
consistent nurse and provider education is necessary to
interpret the flow coupler sound, which can be subjective
and not as clear in some cases as the arterial Doppler
signal.
Furthermore,with increased usage of the flow cou-
pler, our institution was able to undergo a de-escalation
of our resident flap check protocol.Over the course of
the study,the information provided by the flow coupler
proved significantly more reliable and valuable, as nurs-
ing staff, who do not perform pin prick exams on free
flaps, conducted routine flap checks via the flow coupler
signal and alerted residents if the Doppler signal was
lost. With decreased frequency ofresident flap checks,
patients were subjectto less free flap pin pricks, and
residents could more efficiently conductother hospital
responsibilities.This technology thus may serve wellin
the context of limited resident duty hours and also may
aid institutions in which residents are unavailable to
conduct flap checks.
There are several limitations to this study. First,
our sample included 120 patients,11 of whom necessi-
tated free flap revisions for vascular compromise. Due to
the limited sample size,these values in our study may
not accurately reflect true values for the flow coupler or
implantable Doppler arterialmonitoring.Second,use of
Doppler monitoring becomes very valuable in cases of
buried flaps in which proper flap checks are not possible.
Our analysis included only one buried flap, which was
taken to the OR based only on change in the venous
Doppler signal; based on our study, it is difficult to
assess the true utility of the flow coupler in such cases
in which there may be increased rates of unnecessary
OR re-explorations given the high false positive rate
early in the study.Third, a large percentage of patients
had two venous anastomoses,and in cases of one signal
loss without flap compromise,it is difficult to conclude
whether the signal loss correlated with a venous throm-
bosis. We defined true positives such that our institu-
tion’s use ofthe flow coupler accurately correlated with
the clinical exam of the free flap, rather than the physio-
logic occurrence ofa venous thrombosis that does not
compromise the free flap. Therefore, in cases of two
flow couplers, it is possible that signal loss by one
flow coupler could have correlated with a clinically
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
5

nonsignificant venous thrombosis (unrecorded true posi-
tive), or that one vein thrombosed despite strong Dopp-
ler signal and normalexam (unrecorded false negative).
To control for this potential confounder,we conducted
subset analysis of free flaps with only one venous anas-
tomosis and observed identicalsensitivity (100.0%)and
improved specificity and FPR. Fourth, though there
were no cases of flow coupler false negatives in this
study, the authors make note of one case not captured in
this study protocolin which there was venous Doppler
signal but no arterial flow identified intraoperatively
soon after completion of the anastomosis (possibly
related to venous backflow across the anastomosis or
from the second vein); the arterial anastomosiswas
revised and there were no postoperative flap or Doppler
issues. Fifth, a true comparison of the two devices would
have required that the implantable Doppler probe be
placed at the venous anastomosis;our analysis also is
comparing arterial and venous monitoring. However,
several studies have already shown poor reliability of
the implantable Doppler at the venous anastomosis.10,11
Thus, the primary objective of this study was to analyze
and revisit the reliability of venous monitoring via the
flow coupler and compare it to a more commonly used
technique (arterial monitoring alone).
CONCLUSION
The flow coupler is an adequate tool to monitor free
flap viability postoperatively in head and neck recon-
struction. There is an elevated false positive rate when
monitoring the vein, which necessitates physicalexam
checks following Doppler signal loss to minimize the
number of unnecessary re-explorations in the OR.How-
ever,monitoring the vein via flow coupler has high sen-
sitivity compared to implantable arterial Doppler
monitoring,particularly in cases complicated by venous
thrombosis, possibly leading to earlier OR takebacks
and improved salvage rates. In addition, a learning
curve exists in optimizing usage ofthe flow coupler to
minimize accidentalremovals and false positive signal
losses.
BIBLIOGRAPHY
1. Spiegel JH, Polat JK. Microvascular flap reconstruction by otolaryngolo-
gists: prevalence,postoperative care, and monitoring techniques.Laryn-
goscope 2007;117:485–490.
2. Wu CC, Lin PY, Chew KY, Kuo YR. Free tissue transfers in head and neck
reconstruction:complications,outcomes and strategies for management
of flap failure: analysis of 2019 flaps in single institute. Microsurgery
2014;34:339–344.
3. Chang EI, Zhang H, Liu J, Yu P, Skoracki RJ, Hanasono MM. Analysis of
risk factors for flap loss and salvage in free flap head and neck recon-
struction. Head Neck 2016;38(suppl 1):E771–E775.
4. Genden EM, Rinaldo A, Suarez C, Wei WI, Bradley PJ, Ferlito A. Compli-
cations of free flap transfers for head and neck reconstruction following
cancer resection. Oral Oncol 2004;40:979–984.
5. May JW Jr, Chait LA, O’Brien BM, Hurley JV. The no-reflow phenomenon
in experimental free flaps. Plast Reconstr Surg 1978;61:256–267.
6. Chen KT, Mardini S, Chuang DC, et al. Timing of presentation of the first
signs of vascular compromise dictates the salvage outcome offree flap
transfers. Plast Reconstr Surg 2007;120:187–195.
7. Smit JM, Acosta R, Zeebregts CJ, Liss AG, Anniko M, Hartman EH. Early
reintervention of compromised free flaps improves success rate.Micro-
surgery 2007;27:612–616.
8. Cervenka B, Bewley AF. Free flap monitoring: a review of the recent liter-
ature. Curr Opin Otolaryngol Head Neck Surg 2015;23:393–398.
9. Chang EI, Ibrahim A, Zhang H, et al. Deciphering the sensitivity and spe-
cificity of the implantable doppler probe in free flap monitoring.Plast
Reconstr Surg 2016;137:971–976.
10. Guillemaud JP, Seikaly H, Cote D, Allen H, Harris JR. The implantable
Cook-Swartz Doppler probe for postoperative monitoring in head and
neck free flap reconstruction.Arch Otolaryngol Head Neck Surg 2008;
134:729–734.
11. Wax MK. The role of the implantable Doppler probe in free flap surgery.
Laryngoscope 2014;124(suppl 1):S1–S12.
12. Paydar KZ, Hansen SL, Chang DS, Hoffman WY, Leon P. Implantable
venous Doppler monitoring in head and neck free flap reconstruction
increases the salvage rate. Plast Reconstr Surg 2010;125:1129–1134.
13. Riot S, Herlin C, Mojallal A, et al. A systematic review and meta-analysis
of double venous anastomosis in free flaps.Plast Reconstr Surg 2015;
136:1299–1311.
14. Tran NV, Buchel EW, Convery PA. Microvascular complications ofDIEP
flaps. Plast Reconstr Surg 2007;119:1397–1405; discussion 1406–1398.
15. Swartz WM, Izquierdo R, Miller MJ. Implantable venous Doppler micro-
vascular monitoring:laboratory investigation and clinicalresults. Plast
Reconstr Surg 1994;93:152–163.
16. Zhang T, Dyalram-Silverberg D,Bui T, Caccamese JF Jr, Lubek JE. Anal-
ysis of an implantable venous anastomotic flow coupler:experience in
head and neck free flap reconstruction. Int J Oral Maxillofac Surg
2012;41:751–755.
17. Kempton SJ, Poore SO, Chen JT, Afifi AM. Free flap monitoring using an
implantable anastomotic venous flow coupler:analysis of 119 consecu-
tive abdominal-based free flaps for breast reconstruction.Microsurgery
2015;35:337–344.
18. Um GT, Chang J, Louie O, et al. Implantable Cook-Swartz Doppler probe
versus Synovis Flow Coupler for the post-operative monitoring offree
flap breast reconstruction.J Plast Reconstr Aesthet Surg 2014;67:960–
966.
19. Salgado CJ, Chim H, Schoenoff S, Mardini S. Postoperative care and mon-
itoring of the reconstructed head and neck patient.Semin Plast Surg
2010;24:281–287.
20. Brown JS, Devine JC, Magennis P, Sillifant P, Rogers SN, Vaughan ED.
Factors that influence the outcome of salvage in free tissue transfer.Br
J Oral Maxillofac Surg 2003;41:16–20.
Laryngoscope 00: Month 2017 Fujiwara et al.: Flow Coupler for Head and Neck Free Flaps
6