# Flap monitoring
Flap monitoring is the postoperative assessment of a transferred tissue flap's perfusion, oxygenation, and viability, by bedside examination or by devices, so that vascular compromise such as pedicle (vessel stalk supplying blood to a transferred tissue flap) thrombosis is detected early enough for salvage surgery. The risk window is short: one review reports that most vascular compromise occurs in the first 24 hours after surgery, another that most flap failures occur within the first 48 hours, and a third that most failures occur within the first 72 hours, with salvage tied to how quickly the vascular insult is treated.[1](https://www.thieme-connect.com/products/ejournals/pdf/10.1055/s-0040-1722182.pdf) • [2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) • [3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) When compromise is detected early and take-back to theater follows promptly, 70% to 80% of compromised flaps are successfully salvaged.[4](https://pubmed.ncbi.nlm.nih.gov/27975015/)
| Key fact | Value |
|---|---|
| What is assessed | Perfusion and oxygenation, via color, capillary refill, turgor, temperature, and flow signals[5](https://www.oaepublish.com/articles/2347-9264.2023.01) • [6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) |
| Highest-risk period | First 24 to 72 hours postoperatively, depending on the review[1](https://www.thieme-connect.com/products/ejournals/pdf/10.1055/s-0040-1722182.pdf) • [2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) • [3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) |
| Salvage with early take-back | 70–80% of compromised flaps[4](https://pubmed.ncbi.nlm.nih.gov/27975015/) |
| Flap success with clinical assessment alone | 85–95%[2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) |
| Implantable Doppler vs clinical monitoring | Flap failure risk ratio 0.40 (95% CI 0.21–0.75); salvage risk ratio 1.73 (95% CI 1.16–2.59)[4](https://pubmed.ncbi.nlm.nih.gov/27975015/) |
| NIRS vs clinical monitoring | Compromise detected a mean 8.1 h earlier; salvage 87.2% vs 50.0% (P<0.01)[7](https://www.sciencedirect.com/science/article/pii/S2096691122000462) |
| Device costs | Cook-Swartz probe about USD 300; NIRS monitor up to USD 30,000 with sensors up to USD 1,200[3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) • [6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) |
## How it works
Monitoring rests on distinguishing arterial inflow failure from venous outflow failure. The clinical exam comprises four components: color, capillary refill, tension or turgor with swelling, and temperature, and it can discern arterial from venous problems.[5](https://www.oaepublish.com/articles/2347-9264.2023.01) Arterial compromise presents as pale color, delayed capillary refill above 3 seconds, decreased pinprick bleeding, cold temperature, and mild late swelling; venous compromise presents as cyanotic or dusky color, shortened capillary refill under 2 seconds, dark venous pinprick outflow, warmth, and increased turgor.[5](https://www.oaepublish.com/articles/2347-9264.2023.01)
Devices measure different quantities. Handheld acoustic Doppler emits 5–8 MHz ultrasound and listens to reflected waves; a shift from a triphasic or biphasic arterial signal to a monophasic one warns of arterial occlusion, while venous occlusion is detected late, with a characteristic water-hammering signal.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) [Near-infrared spectroscopy](https://www.edgepedia.org/near-infrared-spectroscopy) (NIRS) measures the selective absorption of near-infrared light by hemoglobin and computes tissue oxygen saturation (\( StO_{2} \)); one widely used criterion is a drop rate in \( StO_{2} \) of 20% or more per hour lasting more than 30 minutes, although specific thresholds vary across devices.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) • [7](https://www.sciencedirect.com/science/article/pii/S2096691122000462) Surface-temperature recording, described for free flaps by Roger K. Khouri and William W. Shaw in 1992, tracks the cooling that follows loss of blood flow.[8](https://doi.org/10.1097/00006534-199203000-00017)
A key distinction is between monitoring the pedicle and monitoring the flap tissue. An implantable Doppler probe is a 20-MHz pulsed ultrasonic device mounted on a silicone cuff fixed directly to the pedicle vessels, and it can be placed on the artery or the vein.[4](https://pubmed.ncbi.nlm.nih.gov/27975015/) The placement matters for timing: venous probes detect venous occlusions immediately and arterial occlusions with a mean delay of 6 ± 2.4 minutes, whereas arterial probes detect arterial occlusions immediately but venous occlusions only after 220 ± 40 minutes.[3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) NIRS, thermometry, and imaging instead read the flap tissue itself.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full)
## How it is done
Serial clinical examination supported by handheld acoustic Doppler is the standard of care, with checks of flap color, capillary refill, tissue turgor, and temperature.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) A common schedule is every 30 minutes for the first 3 hours, hourly up to 48 hours after surgery, every 2 hours up to 72 hours, then per departmental guidelines; 75% of surgeons observe at intervals of 3 hours or less.[3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) • [9](https://www.prolekare.cz/en/journals/acta-chirurgiae-plasticae/2019-1-4-1/free-flap-monitoring-review-and-clinical-approach-122018) An abnormal sign or reading triggers immediate take-back for exploration, because salvage rates are higher with early than with delayed reoperation.[2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/)
## Origin
Implantable Doppler monitoring grew from a series of microsurgical papers. John W. Pinnella and colleagues reported direct microvascular monitoring with implantable ultrasonic Doppler probes in [Microsurgery](https://www.edgepedia.org/microsurgery) in 1982.[10](https://doi.org/10.1002/micr.1920030405) Paul M. Parker, Joseph C. Fischer, and William W. Shaw then described an implantable pulsed Doppler cuff for long-term monitoring of free flaps in a preliminary study in Microsurgery in 1984.[11](https://doi.org/10.1002/micr.1920050307) The venous-probe variant was reported by William M. Swartz, Ricardo Izquierdo, and Michael J. Miller as implantable venous Doppler microvascular monitoring in Plastic & Reconstructive Surgery in 1994.[12](https://doi.org/10.1097/00006534-199401000-00024) The physiological basis of NIRS monitoring, noninvasive infrared measurement of tissue oxygen sufficiency, was published by Frans F. Jöbsis in Science in 1977.[13](https://doi.org/10.1126/science.929199) Surface-temperature monitoring of free flaps was reported by Khouri and Shaw in 1992.[8](https://doi.org/10.1097/00006534-199203000-00017)
## Variants
A 2023 systematic review of 22 studies comprising 6,370 flaps shows the spread of adjunctive devices: 1,395 flaps monitored with the Cook-Swartz Doppler (21.90%), 1,417 with [tissue oximetry](https://www.edgepedia.org/tissue-oximetry) (22.24%), 291 with laser Doppler (4.57%), 175 with duplex echography (2.75%), 210 with indocyanine green fluorescence (3.30%), 196 with the Synovis flow coupler (3.08%), and 81 with light spectroscopy (1.27%).[14](https://europepmc.org/article/MED/37844383) NIRS devices are generally not well suited to buried flaps because they require surface access and have limited penetration; fourteen different NIRS devices appear across studies, the most common being the ViOptix T.Ox tissue oximeter.[7](https://www.sciencedirect.com/science/article/pii/S2096691122000462) • [21](https://pmc.ncbi.nlm.nih.gov/articles/PMC8898164/) [Hyperspectral imaging](https://www.edgepedia.org/hyperspectral-imaging) (HSI) records spectra from 500 to 1000 nm and provides \( StO_{2} \), a near-infrared perfusion index, a tissue hemoglobin index, and a tissue water index; the TIVITA Tissue camera acquires a spectrum in 6.4 seconds with \( StO_{2} \) penetration of 1 mm and near-infrared penetration of 4 to 6 mm.[15](https://link.springer.com/article/10.1186/s12893-021-01232-0) A 2025 review classifies techniques as noninvasive (clinical examination, handheld Doppler, thermography, NIRS, laser [Doppler flowmetry](https://www.edgepedia.org/doppler-flowmetry), smartphone-integrated tools) and invasive (implantable Doppler, ICG angiography, microdialysis, bioimpedance sensors), and concludes that clinical monitoring remains the gold standard, with adjunctive tools most useful for buried or ambiguous flaps.[16](https://www.cerrahpasamedj.org/index.php/pub/article/view/984)
Newer variants are automated. FLAPMATE (FLAP Monitoring Through AI Technology and Expertise), an automated smartphone flap-monitoring app, was reported by Jisu Kim and colleagues in JAMA Network Open in 2024; it uses a smartphone camera with FS-Net segmentation and DenseNet121 grading to automate photo capture, segmentation, and perfusion grading.[17](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2821553)
## Applications
Comparative data come mainly from breast reconstruction and from DIEP (deep inferior epigastric perforator) flaps, the most common flap type (85.9%) in the NIRS review.[7](https://www.sciencedirect.com/science/article/pii/S2096691122000462) A meta-analysis of six comparative studies found the implantable Doppler significantly lowered flap failure (risk ratio 0.40; 95% CI 0.21–0.75) and raised the salvage rate (risk ratio 1.73; 95% CI 1.16–2.59) versus clinical assessment; pooled sensitivity was 1.00 versus 0.98, and the false-positive rate 0.01 versus 0.[4](https://pubmed.ncbi.nlm.nih.gov/27975015/) For NIRS, detection preceded clinical recognition of compromise by an average of 8.1 hours (range 0.5 to 32.0 hours); salvage was 87.2% with clinical monitoring plus NIRS versus 50.0% with clinical monitoring alone (P<0.01), and flap survival 98.1% versus 96.3% (P = 0.02).[7](https://www.sciencedirect.com/science/article/pii/S2096691122000462) Across all adjunctive technologies in the 2023 review, the overall true-positive rate for microvascular compromise in taken-back flaps was 70.18%; the Cook-Swartz Doppler showed a true-positive rate of 80.17%, an 83.63% salvage rate, and a 2.60% flap failure rate, while tissue oximetry showed a 74.76% true-positive rate and an 88.62% salvage rate.[14](https://europepmc.org/article/MED/37844383)
## Limitations and alternatives
The main trade-off of device monitoring is false positives, which drive unnecessary take-backs. Meta-analyses report an implantable Doppler false-positive rate of 8 to 17%, up to 31% in one case series of pharyngoesophageal and tracheal reconstructions, while the pooled comparative figure was 0.01 versus 0 for clinical assessment; the discrepancy reflects different study designs and denominators, and both figures are cited in the literature.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) • [2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) • [4](https://pubmed.ncbi.nlm.nih.gov/27975015/) Probe placement changes the failure mode: arterial monitoring was associated with a 74% reduced risk of false-positives (RR 0.26, 95% CI 0.12–0.55) and a 63% reduced risk of signal loss (RR 0.37, 95% CI 0.24–0.59) compared with venous monitoring, although sensitivities, specificities, salvage rates, and flap failure rates did not differ significantly between the two.[18](https://onlinelibrary.wiley.com/doi/10.1002/micr.30564) The delayed detection of venous occlusion by arterial probes, 220 ± 40 minutes, is the practical expression of signal persisting after venous problems develop.[3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/)
Cost-effectiveness remains debated. NIRS hardware can reach USD 30,000 with sensors up to USD 1,200, and one analysis found clinical examination alone more cost-effective than adding NIRS; the Cook-Swartz probe costs about USD 300, and clinical monitoring is the least costly technique overall.[6](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full) • [3](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/) • [2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) Against this, some NIRS series report no false positives or negatives at all: Repez and colleagues monitored 50 flaps continuously and detected every flow failure before clinical observation.[15](https://link.springer.com/article/10.1186/s12893-021-01232-0) The nearest alternative to any device remains structured clinical observation alone, which achieves 85 to 95% flap success; devices are best justified where flaps are buried or findings are ambiguous.[2](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/) • [16](https://www.cerrahpasamedj.org/index.php/pub/article/view/984)
The automated systems carry their own error profiles. FLAPMATE achieved a sensitivity of 92.9% for arterial insufficiency and 97.5% for venous insufficiency, but most of its 457 false positives arose from bruising-related discoloration and most of its 45 false negatives from segmentation errors in small finger or toe flaps.[17](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2821553) The deep-learning remote monitor reached an AUC of 0.92 in internal and 0.93 in external validation, with flap survival of 98.2% versus 91.1% under conventional monitoring (p = 0.11), but it could not detect arterial insufficiency, a gap its authors note is shared by many existing AI tools.[19](https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0347343) [Smartphone](https://www.edgepedia.org/smartphone) thermal imaging uses a flap-to-adjacent-skin difference above 2 °C as the failing-flap threshold, with an acute 3 °C drop at the skin-island center suggesting arterial thrombosis and a 1 to 2 °C uniform drop suggesting venous compromise.[20](https://journals.lww.com/jcsm/fulltext/2024/10010/clinical_utility_of_smartphone_based_digital.17.aspx)
## References
1. [Free Flap Monitoring, Salvage, and Failure Timing (Thieme)](https://www.thieme-connect.com/products/ejournals/pdf/10.1055/s-0040-1722182.pdf)
2. [Postoperative Free-Flap Monitoring Techniques](https://pmc.ncbi.nlm.nih.gov/articles/PMC6408240/)
3. [Flap Monitoring Techniques: A Review (2024)](https://pmc.ncbi.nlm.nih.gov/articles/PMC11432309/)
4. [Implantable Doppler Probes for Postoperatively Monitoring Free Flaps: Efficacy. A Systematic Review and Meta-analysis (Chang et al., PRS Global Open 2016)](https://pubmed.ncbi.nlm.nih.gov/27975015/)
5. [Detecting flap compromise: an updated review of techniques to monitor microsurgical flaps post-operatively in breast reconstruction](https://www.oaepublish.com/articles/2347-9264.2023.01)
6. [Postoperative free flap monitoring in reconstructive surgery, man or machine?](https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2023.1130566/full)
7. [Current evidence on the use of near-infrared spectroscopy for postoperative free flap monitoring: A systematic review](https://www.sciencedirect.com/science/article/pii/S2096691122000462)
8. [Roger K. Khouri, William W. Shaw (1992). Monitoring of Free Flaps with Surface-Temperature Recordings. Plastic & Reconstructive Surgery.](https://doi.org/10.1097/00006534-199203000-00017)
9. [Free-flap monitoring: review and clinical approach (Acta Chirurgiae Plasticae)](https://www.prolekare.cz/en/journals/acta-chirurgiae-plasticae/2019-1-4-1/free-flap-monitoring-review-and-clinical-approach-122018)
10. [John W. Pinnella and colleagues (1982). Direct microvascular monitoring with implantable ultrasonic doppler probes. Microsurgery.](https://doi.org/10.1002/micr.1920030405)
11. [Paul M. Parker, Joseph C. Fischer, William W. Shaw (1984). Implantable pulsed doppler cuff for long‐term monitoring of free flaps: A preliminary study. Microsurgery.](https://doi.org/10.1002/micr.1920050307)
12. [William M. Swartz, Ricardo Izquierdo, Michael J. Miller (1994). Implantable Venous Doppler Microvascular Monitoring. Plastic & Reconstructive Surgery.](https://doi.org/10.1097/00006534-199401000-00024)
13. [Frans F. Jöbsis (1977). Noninvasive, Infrared Monitoring of Cerebral and Myocardial Oxygen Sufficiency and Circulatory Parameters. Science.](https://doi.org/10.1126/science.929199)
14. [Adjunctive technologies in postoperative free-flap monitoring: a systematic review (JPRAS, 2023)](https://europepmc.org/article/MED/37844383)
15. [Hyperspectral Imaging (HSI) as a new diagnostic tool in free flap monitoring for soft tissue reconstruction: a proof of concept study](https://link.springer.com/article/10.1186/s12893-021-01232-0)
16. [Flap Monitoring: What We Know and What Is To Come (Cerrahpaşa Medical Journal, 2025)](https://www.cerrahpasamedj.org/index.php/pub/article/view/984)
17. [Development of an Automated Free Flap Monitoring System Based on Artificial Intelligence (JAMA Network Open, 2024)](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2821553)
18. [Comparison of arterial and venous implantable Doppler postoperative monitoring of free flaps: Systematic review and meta-analysis of diagnostic test accuracy (Microsurgery)](https://onlinelibrary.wiley.com/doi/10.1002/micr.30564)
19. [A remote monitoring system based on deep learning for real-time assessment of free flaps (PLOS One)](https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0347343)
20. [Clinical utility of smartphone-based digital infrared thermal imaging in predicting vascular compromise in free flaps (2024)](https://journals.lww.com/jcsm/fulltext/2024/10010/clinical_utility_of_smartphone_based_digital.17.aspx)
21. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC8898164/)
---
*Topic: Encyclopedia › Life and health › Human health and medicine › Clinical assessment and procedures › Surgery and surgical specialties › Plastic, reconstructive, and oncologic surgery procedures*
*Initially written Sep 29, 2026 · Edited: Oct 15, 2026*
*Copyright 2026 EdgeChat AI, a subsidiary of Biostate AI.*
License: Edgepedia Community License 1.0, https://www.edgepedia.org/edgepedia/license