Factors associated with early and late digital revascularization and replantation failure: a retrospective cohort study

Abstract

Factors associated with failure of digital revascularization and replantation procedures have been well characterized, but studies have not investigated failures occurring beyond the early postoperative period. A single-centre retrospective chart review included 284 patients (434 digits) who underwent digital revascularization or replantation. Patient-, injury- and surgery-related characteristics were compared among successful procedures, digits that failed while in hospital (early failure), and initially viable digits that failed after hospital discharge (late failure). Overall, 202 patients had successful procedures (71%). There were 51 early failures (18%) and 31 late failures (11%). Crush injuries and vein grafting were associated with early failure only. Complete amputations and leeching were strongly associated with both early and late failure. This study revealed that a substantial proportion of initially viable digits fail after discharge from hospital. Patients with signs of venous congestion may benefit from longer observation periods in hospital to avoid late failure.

Level of evidence: IV

Keywords

Failure leeching mechanism replantation revascularization

Introduction

The restoration of amputated digits has a success incidence in the range of 48%–97% reported for replantation procedures (Sebastin and Chung, 2011; Shaterian et al., 2018). Despite this, these procedures carry high incidences of complication, and >50% undergo secondary procedures (Fufa et al., 2013; Shaterian et al., 2019; Yu et al., 2003). With concerns for suboptimal outcomes and costs attributable to complications, revascularization/replantation procedures (replants) are being performed less frequently and becoming more centralized to academic centres (Cho et al., 2018; Friedrich et al., 2011; Ozer et al., 2010; Sears et al., 2014). Controversies surrounding indications for surgery remain.

Replant failure in the early postoperative period has been well studied retrospectively. The majority of replant failures occur within 1 week of surgery and are caused by venous congestion or arterial thrombosis (Breahna et al., 2016; Fufa et al., 2013; Hatchell et al., 2019). The factors most consistently associated with early replant failure are complete amputation (vs. incomplete) (Lee et al., 2019; Tejedor Navarro et al., 2021), inability to perform a vascular anastomosis at the time of surgery (Hatchell et al., 2019; Shaterian et al., 2018), mechanism of injury – with crush, avulsion and degloving mechanisms carrying higher failure rates than sharp mechanisms (Breahna et al., 2016; Waikakul et al., 2000; Zhu et al., 2017), and prolonged warm ischaemia time (Breahna et al., 2016; Zhu et al., 2017).

A few studies have found an association between early failure and younger age (Waikakul et al., 2000; Zhu et al., 2017), cigarette smoking (Breahna et al., 2016; Waikakul et al., 2000; Zhu et al., 2017), having more than three medical co-morbidities (Hustedt et al., 2016) and surgery outside of regular office hours (Breahna et al., 2016). However, these notable findings have not been consistently reproducible. While important for functional outcome after digital replantation, it is also unclear whether the zone of injury is an important predictor of replant failure (Breahna et al., 2016; Zhu et al., 2017). A recent meta-analysis including 32 studies reviewing over 6000 replants found no association between failure and age, sex, zone of injury, digit number, tobacco use or ischaemia time (Shaterian et al., 2018).

Studies have yet to characterize factors predicting failure of initially successful replants after discharge from hospital. It is unclear how frequently delayed presentations of replant failure occur. The aim of this study was to identify patient-, injury- and surgery-related factors predictive of late failure of revascularization or replantation procedures in patients who sustained partial or complete digital amputations.

Methods

Data were collected retrospectively between 2006 and 2016 at a tertiary regional referral centre for hand and wrist injuries. Information from the following three groups of patients was compared: those who underwent successful digital revascularization or replantation; those whose digits failed while in hospital; and those who failed after discharge from hospital with seemingly viable digits. Multinomial regression analysis was used to identify factors associated with early and late failure.

Patients

All data belonged to an institutional database from a regional tertiary referral centre for hand and wrist injuries. Patients diagnosed with absent distal circulation of one or more digits in the emergency department and who underwent emergent revascularization or replantation between 2006 and 2016 were included. Procedures were performed by 11 fellowship-trained hand surgeons with backgrounds in plastic or orthopaedic surgery. The majority of procedures (72%) were performed by highly experienced surgeons (level IV), while 11% of procedures were performed by surgeons with level II expertise and 17% by experienced surgeons with level III expertise (Tang, 2009; Tang and Giddins, 2016). Patient information was collected from the time of presentation with digital amputation to 30 days after discharge from hospital.

Postoperative care and analysis

This cohort study investigated factors associated with the timing of digit failure. We defined revascularization procedures as those performed for partially amputated digits with absent circulation, replantation procedures as those performed for completely amputated digits, and combined procedures as those involving both revascularization and replantation for multiply traumatized hands. Our surgical technique has been described elsewhere (Retrouvey et al., 2020). Our postoperative protocol consists of hourly monitoring in a specialized unit with temperature, colour, and capillary refill. The use of prophylactic and therapeutic anticoagulation was applied on a case-by-case basis when one or more of these parameters was abnormal rather than applying a standard institutional protocol. This is because the benefits of its routine use have not been shown to outweigh the risks (Retrouvey et al., 2019). Leeching or therapeutic anticoagulation in the form of heparin, dextran or heparin with dextran were applied selectively in patients showing evidence of vascular insufficiency. The vascular status of digits was reassessed frequently, and these measures were discontinued shortly after clinical resolution. Many patients underwent surgery for multiple digits at once. In order to ensure independent statistical comparisons, we compared patient outcomes rather than digit-specific outcomes.

There were three groups for comparison:

The successful group included patients with viability of all digits documented upon discharge and throughout the follow-up period. Patients who underwent surgery for multiple digits and had one or more unviable digits were deemed to have failed treatment.

The early failure group consisted of patients with one or more digits that became unviable while in hospital postoperatively. Since almost all of the patients had isolated hand injuries, prolonged hospital stays were rare, and the patients were discharged directly home. Patients with digits showing clear signs of lost viability (dusky colour, coolness, lack of capillary refill, gangrene, etc.) and who opted to be discharged home and returned for outpatient revision amputation were included in this group.

The late failure group consisted of patients demonstrating viability of all digits at the time of discharge from hospital and whose digit(s) subsequently became unviable and required revision amputation during the follow-up period. Patients who met the criteria for both early and late failure were included in the early failure group only.

We investigated patient-, injury- and surgery-related factors. Baseline patient demographic variables included age, sex, body mass index (BMI), American Society of Anesthesiologists (ASA) class, occupation and medical co-morbidities. Smoking was a patient-related variable of interest as it has been associated with early replant failure (Breahna et al., 2016; Zhu et al., 2017). Injury-related variables included blade, crush and avulsion injuries and ischaemia time. Surgery-related variables included type of procedure (revascularization, replantation, combined), surgeon’s level of expertise, duration of surgery, number of digits, vein repair and use of vein grafts during surgery (Chaivanichsiri and Rattanasrithong, 2006). Surgery taking place between 8 a.m. and 6 p.m. was previously associated with replant success and was investigated in our study (Breahna et al., 2016). Finally, postoperative variables included return to the operating room, use of leeches, duration of hospital stay and complications.

Statistical analyses

Univariate analyses were used to compare patient-, injury- and surgery-related characteristics among the successful, early failure and late failure groups, which consisted of one-way analysis of variance for continuous variables, Kruskal–Wallis tests for nonparametric quantitative variables and chi-square tests for categorical variables. Next, we used multinomial logistic regression analysis to determine which factors were most strongly associated with early or late failure. Statistically significant variables from the univariate analyses were used to build this model with the exception of redundant factors (see Results). Age and sex were also included to control for baseline differences. Strength of association was estimated using the odds ratio (OR) and 95% confidence interval (CI). P < 0.05 was considered statistically significant.

Results

Among 284 patients (age range 15–84 years), 202 (71%) underwent successful digital revascularization or replantation, 51 (18%) failed in the early postoperative period and 31 (11%) had initially viable digits that failed during the outpatient follow-up period. Therefore, 31 of 82 failures (38%) occurred late. Among 434 digits, 316 (73%) had successful outcomes.

Patient-related factors

The groups did not differ significantly in age, sex, BMI, medical co-morbidities, ASA class, labour occupation or smoking in packs per day (Table 1).

Table 1.

Comparison of patient, injury and surgical characteristics among patients who underwent digital revascularization or replantation procedures.

Parameter	Successful (n = 202)	Early failure (n = 51)	Late failure (n = 31)	P-value^a
Patient characteristics
Age (years)^b	42 (17–77)	42 (15–84)	42 (17–79)	NS
Male gender (%)	189 (94)	46 (90)	27 (87)	NS
BMI (kg/m²)^b	27 (17–40)	27 (19–36)	28 (20–36)	NS
Smoking, ppd^b	0.3 (0–2.5)	0.2 (0–1.0)	0.2 (0–1.0)	NS
ASA class^b	2 (1–3)	2 (1–3)	2 (1–4)	NS
Injury characteristics
Mechanism of injury				0.001
Blade	164 (81)	33 (65)	20 (67)
Crush	20 (10)	11 (22)	3 (10)
Avulsion	18 (9)	7 (14)	8 (27)
Ischaemia time (h:min)	7:53	8:30	6:52	NS
Surgical characteristics
Type of procedure				<0.001
Revascularization only	137 (88)	14 (9)	5 (3)
Replantation only	53 (50)	30 (28)	23 (22)
Combined	12 (55)	7 (31)	3 (14)
Surgeon level of expertise^c	4	4	4	NS
Daytime surgery^d	59 (29)	15 (29)	6 (19)	NS
Duration of surgery (min)	307 (SEM 10)	416 (SEM 26)	391 (SEM 30)	<0.001
Number of digits^c	3	3	3	NS
Vein graft	16 (8)	14 (28)	7 (23)	0.028
Vein repair	78 (39)	41 (80)	25 (81)	<0.001
Postoperative characteristics
In-hospital infection	5 (3)	3 (6)	2 (7)	NS
In-hospital return to OR	7 (3)	14 (27)	1 (3)	<0.001
Leeching	42 (21)	32 (63)	24 (77)	<0.001
Length of stay (days)^c	5 (2)	7 (2)	7 (3)	<0.001

Values are given as n (%) unless otherwise indicated.

P < 0.05 represents significant differences in covariates between groups.

Values are given as mean (range).

Values are given as median or median (IQR).

Surgery started between 8 a.m. and 6 p.m.

ASA: American Society of Anesthesiologists score; BMI: body mass index; IQR: interquartile range; NS: not significant; ppd: pack per day; SEM: standard error of the mean.

Injury-related factors

The three groups differed significantly in terms of mechanism of injury (P = 0.001) (Table 1). In the successful group, patients were more likely to have sustained an injury by blade mechanism than in the failure groups. By contrast, crush mechanism accounted for a higher proportion of injuries in the early failure group than the other groups, and avulsion mechanism accounted for a higher proportion of injuries in the late failure group. The groups did not differ significantly in ischaemia time or transport distance.

Surgery-related factors

Type of surgery (revascularization, replantation, combined) differed significantly among groups (P < 0.001) (Table 1). Both early and late failure were more likely in patients who had undergone replantation or combined procedures. Patients in both failure groups underwent longer surgeries (P < 0.001), and higher proportions of patients requiring intraoperative vein grafts (P = 0.028) or vein repair (P < 0.001) were in the failure groups. Surgeon level of expertise, surgery during daytime hours, number of digits involved and intraoperative blood loss did not differ significantly among groups.

The groups also differed in terms of postoperative course. Relatively fewer patients in the successful group underwent leeching compared to the early and late failure groups (P < 0.001). Length of hospital stay differed significantly among groups (P < 0.001). Patients with successful procedures had a shorter average hospital stay. The incidence of return to the operating room also differed significantly among groups (P < 0.001), owing to revision amputations in the early failure group. Postoperative infections and minor cardiac or respiratory complications were uncommon and did not differ significantly across groups. The incidence of return to the operating room and complications did not differ among different levels of surgeon expertise.

Factors predictive of failure

Significant covariates were used to construct a multinomial logistic regression model, with the exception of duration of surgery and length of stay. The former was felt to be non-specific and redundant with other factors, while differences in length of stay were felt to be an effect rather than a cause of failure. Relative to the successful group (control), the model produced sets of factors predictive of early and late failure (Table 2).

Table 2.

Odds ratio (95% confidence interval) of factors predictive of revascularization or replantation failure.

Logistic regression	Early	Late
Patient characteristics
Age	1.0 (0.98–1.03)	1.0 (0.97–1.0)
Male gender	0.8 (0.2–2.8)	0.7 (0.2–3.1)
Injury characteristics^a
Crush	4.4 (1.4–13.3)^b	2.0 (0.4–9.7)
Avulsion	0.9 (0.3–2.9)	2.4 (0.7–8.4)
Surgical characteristics
Replant	4.1 (1.5–11.4)^b	6.3 (1.6–24.6)^b
Leeching	3.4 (1.6–7.3)^b	7.5 (2.8–20.2)^b
Vein repair	2.3 (0.8–6.9)	0.98 (0.3–3.6)
Vein graft	3.5 (1.4–8.9)^b	1.9 (0.5–6.7)

Compared to blade mechanism.

P < 0.01, no significant differences in all other comparisons.

Demographic variables were not significantly associated with early failure. Crush mechanism was significantly associated with early failure (OR 4.4, 95% CI: 1.4 to 13.3; P = 0.009), whereas avulsion mechanism was not (P = 0.892). Procedures involving replantation, rather than revascularization only, were also predictive of early failure (OR 4.1, 95% CI: 1.5 to 11.4; P = 0.007). Leeching (OR 3.4, 95% CI: 1.6 to 7.3 P < 0.001) and use of a vein graft (OR 3.5, 95% CI: 1.4 to 8.9; P = 0.009) were predictive of early failure, while vein repair was not (P = 0.1).

Demographic variables and injury mechanism were not significantly associated with late failure. Procedures involving replantation (OR 6.3, 95% CI: 1.6 to 24.6; P = 0.008) and leeching (OR 7.5, 95% CI: 2.8 to 20.2; P < 0.001) were significant predictors of late failure. Vein grafting and vein repair were not significant.

Discussion

In our patients, late failures were much more common than in another study that reported on delayed replant failure (Hatchell et al., 2019). Crush mechanism was predictive of early failure, whereas avulsion mechanism was not predictive of failure despite being most common in the late failure group. Complete amputations requiring replantation, rather than partial amputations salvaged by revascularization, were more likely to fail both early and late. Postoperative leeching was predictive of both early and late failure. Use of a vein graft was predictive of early but not late failure.

In our patients, patient-related variables were not significantly associated with failure. These findings are in agreement with a recent meta-analysis (Shaterian et al., 2018). We did not find an association between medical co-morbidities and failure as suggested by Hustedt et al (2016). This may be attributable to the fact that our patients were generally healthy based on ASA scores. We also did not find an association between smoking and failure. Zhu et al. (2017) found that only heavy cigarette consumption (>20 cigarettes/day) was associated with replant failure, but the proportion of heavy smokers in our cohort was relatively low. Regional trends toward decreased cigarette consumption may have rendered this variable less relevant in our study.

Consistent with previous publications, the mechanism of injury was a predictor of replant failure (Sebastin and Chung, 2011; Shaterian et al., 2018; Waikakul et al., 2000; Zhu et al., 2017). Although the traditional wisdom has been that avulsion mechanisms carry worse outcomes than sharp or crush mechanisms (Wolfe and Wang, 2015), a meta-analysis showed slightly higher failure incidence after crush injuries than after avulsion injuries (Shaterian et al., 2018). The overall picture may reflect a lack of emphasis on timing of failure in previous studies. In our study, the only significant association between mechanism and outcome was between crush injury and early failure. Preliminary analyses showed that avulsion injuries were more common in the late failure group despite failing to reach significance in the regression analyses. The zone of injury may demarcate over a longer period after avulsion injuries, potentially warranting longer observation periods after surgery.

Ischaemia time has been inconsistently associated with replant failure (Breahna et al., 2016; Cavadas et al., 2018; Shaterian et al., 2018; Zhu et al., 2017), and the association was not significant in our study. This was likely aided by strict local protocols regarding the preservation of amputated parts together with timely transport to hospital. Cold ischaemia time did not exceed 24 hours for any of the replanted digits.

Replantation, rather than revascularization, was predictive of both early and late failure. This finding is consistent with previous literature reporting improved survival incidence associated with revascularization and is attributed to preservation of venous drainage in revascularized digits (Kamarul et al., 2018; Lee et al., 2019; Soucacos et al., 1995; Tejedor Navarro et al., 2021). Amputations may be accompanied by more devastating soft-tissue injury and require longer, more complex salvage operations. This raises the concern that revascularization and replantation are quite different procedures and should be treated as such both surgically and when reporting results.

Arterial insufficiency accounts for the majority of replant failures within the first 24 hours after surgery (Breahna et al., 2016; Hatchell et al., 2019; Sammer, 2015). A recent study showed that interpositional vein grafting to mitigate poor arterial flow did not affect survival of replanted digits (Lee et al., 2019). In our study, direct arterial repair was attempted for well-preserved vessels whenever a tension-free repair could be achieved. If large segments were damaged or the ends of the vessel could not be approximated, interpositional vein grafting was performed. Vein grafting was associated with early failure, suggesting that digits with complex arterial injuries that could not be directly repaired were more likely to fail early.

Beyond 24 hours postoperatively, venous congestion was the most common cause of replant failure (Breahna et al., 2016; Hatchell et al., 2019; Sammer, 2015). We performed vein repairs for partial and complete amputations whenever possible. Contrasting with previous studies showing that vein repair is important for success (Shaterian et al., 2018; Waikakul et al., 2000), this was not associated with outcomes in our study. This may be explained by preserved venous outflow in partial amputations and by use of simpler techniques such as controlled nailbed bleeding, which have been shown to have good outcomes without vein repair (Erken et al., 2013).

Therapeutic intravenous anticoagulation administered postoperatively was not shown to be predictive of failure in our cohort in a previously published study (Retrouvey et al., 2019). In the current study, leeching was an important predictor of both early and late failure, implicating venous congestion as a common cause of failure. In a recent study, salvage of failing digits by nonoperative methods had a 44% success rate in restoring viability (Hatchell et al., 2019). While leeching is commonly used, there are no evidence-based recommendations for duration of leeching. We typically keep patients in hospital for 1–2 days after the discontinuation of leeching to monitor for signs of ongoing venous congestion. However, recurrent symptoms may have only become apparent after discharge. Such patients would be overrepresented in the late failure group, even though their digits began failing before the time of discharge. Our findings beg the question of whether patients should have been kept in hospital longer after discontinuation of leeching.

A limitation of our study is its retrospective nature. Retrospective data are particularly susceptible to the influence of selection bias as indications for replantation differ among surgeons and centres. We cannot rule out the possibility that younger, healthier patients with blade mechanism of injury were disproportionately selected for digit salvage. It is also impossible to guarantee the accuracy of all patient charts. Nevertheless, the use of a large database from a single centre ensured that the data were collected and documented in a consistent manner. Finally, we chose to analyse patient outcomes rather than digit-specific outcomes as we felt that survival of all traumatized digits was an important, clinically relevant parameter for gauging treatment success. While this limited our ability to draw conclusions about the importance of digit-specific variables (i.e. level of amputation), we found that patient- and digit-specific incidences of failure were similar and not influenced by the number of traumatized digits.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Ethical approval for this study was obtained from the University Health Network Research Ethics Board.

Informed consent

Informed consent was not sought for the present study because it was a retrospective chart review of an anonymized database.

References

Breahna

Siddiqui

Fitzgerald O’Connor

Iwuagwu

. Replantation of digits: A review of predictive factors for survival. J Hand Surg Eur. 2016, 41: 753–7.

Cavadas

Rubi

Thione

Perez-Espadero

. Immediate versus overnight-delayed digital replantation: Comparative retrospective cohort study of survival outcomes. J Hand Surg Am. 2018, 43: 625–30.

Chaivanichsiri

Rattanasrithong

. Type of injury and number of anastomosed vessels: Impact on digital replantation. Microsurgery. 2006, 26: 151–4.

Cho

Zhong

Kotsis

Chung

. Finger replantation optimization study (front): Update on national trends. J Hand Surg Am. 2018, 43: 903–12.e1.

Erken

Takka

Akmaz

. Artery-only fingertip replantations using a controlled nailbed bleeding protocol. J Hand Surg Am. 2013, 38: 2173–9.

Friedrich

Poppler

Mack

Rivara

Levin

Klein

. Epidemiology of upper extremity replantation surgery in the United States. J Hand Surg Am. 2011, 36: 1835–40.

Fufa

Calfee

Wall

Zeng

Goldfarb

. Digit replantation: Experience of two U.S. Academic level-I trauma centers. J Bone Joint Surg Am. 2013, 95: 2127–34.

Hatchell

Sandre

McRae

Farrokhyar

Avram

. The success of salvage procedures for failing digital replants: A retrospective cohort study. Microsurgery. 2019, 39: 200–6.

Hustedt

Chung

Bohl

Olmscheid

Edwards

. Evaluating the effect of comorbidities on the success, risk, and cost of digital replantation. J Hand Surg Am. 2016, 41: 1145–52.e1.

10.

Kamarul

Mansor

Robson

Albusaidi

Suhaeb

Samsudin

. Replantation and revascularization of amputated upper limb appendages outcome and predicting the factors influencing the success rates of these procedures in a tertiary hospital: An 8-year retrospective, cross-sectional study. J Orthop Surg (Hong Kong). 2018, 26: 1.

11.

Lee

Klifto

Milone

, et al.Survival after digit replantation and revascularization is not affected by the use of interpositional grafts during arterial repair. Plast Reconstr Surg. 2019, 143: 551e–7e.

12.

Ozer

Kramer

Gillani

Williams

Smith

. Replantation versus revision of amputated fingers in patients air-transported to a level 1 trauma center. J Hand Surg Am. 2010, 35: 936–40.

13.

Retrouvey

Makerewich

Solaja

Giuliano

Niazi

Baltzer

. Effect of vasopressor use on digit survival after replantation and revascularization-a large retrospective cohort study. Microsurgery. 2020, 40: 5–11.

14.

Retrouvey

Solaja

Baltzer

. Role of postoperative anticoagulation in predicting digit replantation and revascularization failure: A propensity-matched cohort study. Ann Plast Surg. 2019, 83: 542–7.

15.

Sammer

. Management of complications with flap procedures and replantation. Hand Clin. 2015, 31: 339–44.

16.

Sears

Shin

Prosser

Chung

. Economic analysis of revision amputation and replantation treatment of finger amputation injuries. Plast Reconstr Surg. 2014, 133: 827–40.

17.

Sebastin

Chung

. A systematic review of the outcomes of replantation of distal digital amputation. Plast Reconstr Surg. 2011, 128: 723–37.

18.

Shaterian

Rajaii

Kanack

Evans

GRD

Leis

. Predictors of digit survival following replantation: Quantitative review and meta-analysis. J Hand Microsurg. 2018, 10: 66–73.

19.

Shaterian

Sayadi

Anderson

WKY

Evans

GRD

Leis

. Characteristics of secondary procedures following digit and hand replantation. J Hand Microsurg. 2019, 11: 127–33.

20.

Soucacos

Beris

Touliatos

Korobilias

Gelalis

Sakas

. Complete versus incomplete nonviable amputations of the thumb. Comparison of the survival rate and functional results. Acta Orthop Scand Suppl. 1995, 264: 16–8.

21.

Tang

. Re: Levels of experience of surgeons in clinical studies. J Hand Surg Eur. 2009, 34: 137–8.

22.

Tang

Giddins

. Why and how to report surgeons' levels of expertise. J Hand Surg Eur. 2016, 41: 365–6.

23.

Tejedor Navarro

Vendrell Jorda

Puente Alonso

. Digital replantation/revascularization: Predictive factors to microsurgery success-a single-center study. Eur J Trauma Emerg Surg. 2021, 47: 225–32.

24.

Waikakul

Sakkarnkosol

Vanadurongwan

Un-nanuntana

. Results of 1018 digital replantations in 552 patients. Injury. 2000, 31: 33–40.

25.

Wolfe

Wang

. Replantation of the upper extremity: Current concepts. J Am Acad Orthop Surg. 2015, 23: 373–81.

26.

Shieh

Lee

Hsu

Chiu

. Secondary procedures following digital replantation and revascularisation. Br J Plast Surg. 2003, 56: 125–8.

27.

Zhu

Zhang

Zheng

. Pre-operative predictive factors for the survival of replanted digits. Int Orthop. 2017, 41: 1623–6.