Page 130 - Read Online

Page 4 of 14 Foppiani et al. Plast Aesthet Res 2023;10:53 https://dx.doi.org/10.20517/2347-9264.2022.137

Outcomes
The primary outcomes were detection of complications, identification of threatened flaps, patient return to
the operating room (flap “take-back”), flap salvage, flap loss, and time to identification of complications.

Quality assessment
To assess the risk of bias, we utilized the National Institute of Health (NIH) quality assessment tool. Each
article was categorized as follows: “low risk,” “moderate risk,” or “high risk” of bias.

Statistical analysis
A comprehensive qualitative analysis was made. For the quantitative analysis, the binomial data was
analyzed. Each complication rate's pooled prevalence was estimated using a proportion meta-analysis with
[17]
Stata statistical software (STATA Corp., College Station, TX version 16.1) . Due to the heterogeneity
among studies, a logistic-normal-random-effect model was conducted. Ninety-five percent exact confidence
interval (CIs) and 95% Walds CIs were performed for study-specific and overall pooled prevalence,
respectively. Additionally, the Freeman-Tukey double arcsine transformation was used. The percentage of
2
weight and effect size of each individual study were presented [17,18] . To assess heterogeny, I statistics were
used. Significant heterogeneity was considered if pp-value < 0.05 or I > 50%.
2
RESULTS
Study selection and characteristics
A total of 614 studies were initially retrieved following the removal of duplicates. Of those, 18 met all
inclusion criteria. However, 4 of the 18 articles contained duplicate or already published patient information
and were removed. Therefore, 14 articles were ultimately included for qualitative and quantitative analysis
[Figure 1] [19-32] . Of the 14 articles, 11 were focused on oximetry, and 3 were focused on thermography. When
using the NIH quality assessment tool, 7 were found to be at low risk of bias, 6 at moderate risk, and 1 at
high risk based on the NIH quality assessment tool [Table 1] [19-32] . The Prisma Flow diagram is seen in
Figure 1.

Patient and flap characteristics [Table 1]
From all 14 included studies, a total of 2,529 female patients who underwent microvascular breast
reconstruction were included in this analysis, which ultimately totaled 3,289 flaps overall [19-32] . The mean age
for the cohorts included in this study ranged from 48.9 to 57 years of age. The most common flap used in
the patient population was the deep inferior epigastric (DIEP) flap with 2,372 total flaps, followed by 96
transverse rectus abdominis (TRAM) flaps, 43 superior gluteal artery perforator (SGAP) flaps, 17 superficial
inferior epigastric artery (SIEA) flaps, 8 profunda artery perforator (PAP), 6 diagonal/transverse upper
gracilis (DUG/TUG) and 1 latissimus dorsi (LD) flap. The remainder of the flaps included 746 flaps
described only as “abdominal-based flaps” and stacked flaps, which can be seen in Table 1 [19-32] .

Diagnostic tools and monitoring protocols
The studies included in this review used a variety of diagnostic tools for thermography and oximetry, each
with its own nuances in terms of application and protocols.

In the realm of thermography, the study by Saxena et al. employed the FLIR A320 IR thermal camera, a
specialized device designed for thermal imaging, while the research conducted by Phillips et al. utilized the
FLIR One device, which is connected to a mobile smartphone for ease of use [20,21] . On the other hand, the
study by Thiessen et al. did not explicitly indicate which tool was employed for dynamic infrared
thermography . Notably, the approaches to measurement in these studies showed some variation. Both
[19]
Thiessen et al. and Saxena et al. conducted two measurements within the initial 1-2 days post-procedure,