Is there a need for exploration in pulseless supracondylar fractures of the humerus? A systematic review and individual patient data meta-analysis

Abstract

Background

Different approaches have been proposed to treat patients with pulseless supracondylar humeral fractures (SHF). We aim to analyze the current outcomes of patients who have undergone different treatment options for pulseless SHF, namely watchful waiting versus urgent surgical exploration.

Methods

Electronic databases including PubMed, Embase, and Cochrane Library were explored. Our study included patients from all age groups but only included English-language articles. Single case reports and case studies with adequate description of patient population, injury, and outcomes were included. An individual patient data meta-analysis was done to evaluate key outcomes of surgical exploration versus no surgical exploration in pulseless patients with SHF.

Results

Overall, the data for 1070 individual patients from a total of 48 studies were included.

Patients with pulseless SHF with open fractures have a higher probability of requiring vascular intervention (p < 0.001) as they are more likely to have disrupted arteries (p = 0.050) and more likely to require vascular repairs (p < 0.001) than patients with closed fractures. Similarly, patients with pulseless SHF with pucker sign (p = 0.003) and ecchymosis (p = 0.002) were more likely to undergo surgical exploration. However, the neurological status had no relation to them undergoing surgical exploration (p = 0.382), and also does not affect complication outcomes (p = 0.326) nor the need for vascular intervention (p = 1.00).

However, patients with pale pulseless SHFs were more likely to undergo surgical exploration (p < 0.001), were more likely to have disrupted arteries (p < 0.001), required more vascular intervention (p < 0.001), and had a higher likelihood of complications (p < 0.001).

Notably, there is no significant difference in overall complications between those who underwent exploration of arteries and those who did not among patients with pink pulseless SHF (p = 0.230).

Conclusions

Significant predictors for arterial injury requiring intervention include a pale pulseless limb, open fracture, pucker sign, and ecchymosis; expedient surgical exploration is recommended in these scenarios. However, for patients with pink pulseless SHF, watchful waiting is a viable strategy. Our meta-analysis found no significant difference in overall complication rates between patients with pink pulseless who underwent exploration and those who did not.

Level of evidence

Level II.

Registration

The review process was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines CRD.

Introduction

Supracondylar humeral fractures (SHF) are the most common type of elbow fractures, accounting for about 60% of all elbow fractures [1, 2]. They occur due to an extension or flexion force on the distal humerus, commonly caused by falling onto an outstretched hand, with the majority of cases being extension-type injuries.

Initial trauma and subsequent fracture displacement may damage surrounding neurovascular structures such as the brachial artery and median nerve, leading to reports of associated neurovascular injury at rates as high as 49%, with vascular compromise reported in 3–19% of cases [3–5]. Vascular compromise can be caused by complete transection of the brachial artery, kinking of the artery with reduced flow, thrombosis, intimal tear, arterial contusion or spasm, entrapment of the vessel within the fracture site, or traumatic aneurysm of the brachial artery with subsequent thrombus formation. Such disruptions and injuries of the brachial artery then lead to the presentation of a pulseless hand.

In a subset of patients with pink pulseless supracondylar humeral fractures, the optimal treatment has been highly controversial, with some clinicians recommending that close observation is sufficient after closed reduction and percutaneous pinning of the fractures, while some recommend immediate surgical exploration if the radial pulse is not palpable [6]. Proponents of close observation cite the presence of extensive collateral circulation around the elbow that maintains the viability of the distal extremity despite lack of brachial artery patency [7], and reports of the return of a palpable radial pulse hours or days following reduction and pinning and limited clinical evidence of any long-term complications in patients who did not undergo immediate exploration [8]. Others advocate a more aggressive approach to reestablish the patency of the brachial artery in those with a pulseless pink hand [9, 10]. Pediatric surgeons worry that delaying vascular exploration and intervention could lead to complications such as compartment syndrome [11], Volkmann ischemic contractures [12], and cold intolerance [13]. Blakey et al. reported a high incidence of ischemic contracture in those who did not undergo immediate surgical exploration and advised urgent exploration in patients with absent distal pulses when there is pain and evidence of a deepening nerve lesion to prevent long-term disability [11].

Thus, the aim of this paper seeks to identify current practices for pulseless supracondylar fractures as well as to determine factors affecting various outcomes among these patients. We also hope to analyze the current practices and outcomes of the subset of patients with pink pulseless supracondylar fractures. Our hypothesis is that urgent surgical exploration is warranted in those with pale pulseless hands to prevent complications such as ischemic contractures and compartment syndrome, whereas watchful waiting is a viable strategy for those with pink pulseless supracondylar fractures.

Methods

The review process was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines CRD42024496335 (Fig. 1) regarding the management of pulseless supracondylar humerus fractures, especially regarding surgical exploration and interventions undertaken as well as postoperative outcomes of the patients.

Fig. 1.

PRISMA flowchart of study records

Electronic databases including PubMed, Embase, and Cochrane Library were explored in June 2023 using the search terms (pulse* OR dysvascular OR arter* OR hypoperfus* OR perfus* OR vascular OR warm OR pink OR pale OR white) AND supracondylar AND (fracture* OR broken). The search was not limited by year of publication, type of journal, or level of evidence. The bibliographies of all papers included were cross-referenced for further relevant studies. The study included patients from all age groups but only included English-language articles. Single case reports and studies with adequate description of patient population, injury, and outcomes were included. We excluded papers with no original data (e.g., opinion pieces) and those with no mention of pulse status.

Two authors reviewed the abstracts of all papers that corresponded to the search terms and assessed them for inclusion, with disagreement being resolved by a third author. The abstract sieve identified 1140 papers in total that fit our inclusion criteria and full texts were obtained for more detailed evaluation. After our full-text sieve, we eventually decided on 48 papers, including 5 additional papers suitable for inclusion that were found in the full-text sieve.

In total, we analyzed the data of 1070 patients who presented with pulseless supracondylar humeral fractures. We included papers with patients who were pulseless on palpation but in whom there was a positive flow using Doppler. The review was conducted in accordance with the PRISMA guidelines.

For the current practices, we looked at factors such as perfusion status, presence of neurological deficits, presence of open fracture, pucker or dimple sign (skin indentation due to fracture fragment tenting the skin), and ecchymosis to determine their relationship with surgical exploration and intervention of the arteries. For the outcomes, we looked at whether pulses returned post-fixation or postoperation, the need for subsequent vascular exploration, eventual fracture reduction, any cubitus varus or valgus deformities, range of motion, neurological deficits, the presence of compartment syndrome, and Volkmann’s ischemic contractures.

Coding of the data was carried out by two independent authors in a blinded, pairwise fashion to ensure accuracy and any discrepancies were resolved by consensus in consultation with a third independent author. As for data analysis, we used SPSS model v22 for analysis and two-tailed χ² test to analyze for statistical significance in the differences in outcomes between different groups. Tests of statistical significance such as Pearson’s, Fisher’s exact, and χ² tests were used.

ROBINS-I tool was utilized to assess the risk of bias in nonrandomized studies of intervention, assessing seven domains of bias—bias due to confounding, bias due to selection of participants, bias in classification of intervention, bias due to deviation from intended outcome, bias due to missing data, bias in measurement of outcome, and bias in selection of reported result.

Results

This systematic review is the first Individual Patient Data Meta Analysis of surgical exploration on pulseless supracondylar humeral fractures, synthesizing data from 48 studies comprising 1070 patients, of whom 255 underwent vascular intervention. All of the studies were case reports, case series, retrospective cohort studies, and prospective cohort studies that originally examined patients diagnosed with SHF. Unfortunately, most studies did not provide the detailed demographics of the population with pulseless SHF. As a demographical approximation based on all patients with SHF regardless of pulselessness, 58.3% were male (2294) and 41.7% were female (1640), with a mean age of 7.10 years and age range of 1–16 years (Table 1). Of the reported perfusion status, 134 pale hands and 436 pink hands were reported before any procedures.

Table 1.

Summary table of the included studies

References	Sample size with pulseless supracondylar humeral fractures	Age (years) Mean	Male	Female	Follow-up (months) Mean	Neurologic deficit before procedure	Perfusion status before procedure	Arterial exploration performed	Complication reported
Aksakal et al. [20]	15	7.0	46	19	25.9	8	NR	15	Nil
Ay et al. [21]	7	8.7	35	26	42.0	4	NR	7	Nil
Basbug et al. [22]	1	9.0	1	0	NR	0	1 pale	1	Nil
Benedetti et al. [23]	11	8.0	30	18	34.0	1	3 pale 8 pink	7	2 paresthesia
Brahmamdan et al. [24]	8	7.4	7	5	3.6	5	NR	2	2 neurological deficit 1 ischemia
Cambon-Binder [25]	27	7.0	19	8	3.5	5	18 pink 9 pale	3	3 neurological deficit 2 paresthesia 2 secondary displacement 1 superficial skin infection 1 pulseless on 6th week
Chaturvedi [26]	21	7.6	16	5	20.0	NR	NR	1	8 poor fracture union
Cheng [27]	9	6.0 (median)	253	150	NA	NR	8 pink 1 pale	2	Nil
Cheng [28]	8	6.0	119	61	42.0	NR	NR	1	NR
Clark [29]	1	7.0	0	1	NA	NR	1 pale	1	1 neurological deficit
Clement [30]	9	6.5	NR	NR	36.0	9	9 pale	9	1 neurological deficit
Delniotis [31]	23	6.5	124	86	12.0	5	20 pink 3 pale	2	NR
Erhart [32]	1	3.0	0	1	NR	0	1 pink	1	Nil
Ermutlu [33]	3	3.7	1	2	NR	NR	3 pink	3	Nil
Ghasemzadeh [34]	21	NR	50	11	NR	1	NR	4	NR
Gosens [10]	4	7.7	302	315	5.4	2	4 pink	0	NR
Grant [35]	5	NR	NR	NR	NR	NR	NR	5	1 neurological deficit
Harris [36]	71	6.9	40	31	8.2	71	NR	19	1 compartment syndrome
Karlsson [37]	4	6.5	1	3	5.8	3	2 pink 1 pale	4	1 neurological deficit 1 10° loss of extension
Khan [38]	8	7.0	45	20	NR	2	6 pink	2	NR
Korompilias[39]	5	8.5	38	28	34.0	1	5 pink	4	Nil
Kow [40]	1	6.0	1	0	NR	2	1 pink	1	Nil
Louahem [41]	68	7.5	275	129	100.8	32	63 pink 5 pale	8	NR
Luria [42]	17	6.0	15	9	NR	2	NR	11	2 paresthesia 3 pain 1 osteomyelitis
Mahdy [43]	18	5.5	6	12	NA	0	18 pink	4	NR
Mangat [44]	19	3.8	14	5	NR	10	19 pink	12	NR
Matuszewski [45]	67	9.5	46	21	18.0	NR	67 pale	35	3 restricted range of motion 1 Volkmann ischemic contracture
Rabee [46]	6	5.5	59	27	NR	NR	5 pink 1 pale	6	1 compartment syndrome
Ramesh [47]	15	7.9	61	44	36.6	5	15 pink	0	1 cubitus varus
Salvati [9]	4	11.0	4	0	NR	NR	NR	4	Nil
Schoenecker [48]	6	5.7	2	5	2.1	NR	NR	7	Nil
Soh [49]	37	NR	NR	NR	NR	7	26 pink	4	Nil
Štichhauer [50]	125	NA	62	63	NR	55	125 pink	11	26 neurological deficit
Tomaszewski [51]	20	7.2	140	189	NR	NR	NR	3	1 intolerance to cold temperature
Towler [52]	5	NR	NR	NR	NR	NR	5 pink	4	1 Volkmann ischemic contracture
Tunku-Naziha [53]	7	8.5	5	2	NR	2	7 pink	7	Nil
Vasli [54]	4	9.0	4	0	4.0	1	NR	2	NR
Wegmann [55]	40	6.5	274	225	3.5	NR	1 pink	6	NR
Weller [19]	53	NR	NR	NR	NR	17	4 pale	5	NR
Wong [56]	1	6.0	1	0	1.0	0	1 pink	1	Nil
Xie [18]	13	6.0	8	5	4.5	8	13 pink	0	Nil
Shaw [4]	17	5.7	11	6	16.4	8	17 pale	3	NR
Sabharwal [57]	13	6.4	7	6	31.0	8	13 pink	2	NR
Louahem [58]	29	7.5	140	70	NR	NR	26 pink 3 pale	6	1 cubitus varus
Lyons [59]	12	5.7	NR	NR	NR	6	NR	4	NR
Saglam [60]	58	8.3	NR	NR	42.0	NR	NR	9	1 neurological deficit
Choi [8]	33	5.3	14	19	2.0	10	24 pink 9 pale	5	2 compartment syndrome
Noaman [61]	120	7.0	18	13	26.0	22	NR	31	1 Volkmann's ischemic contracture

The cumulative risk of bias was deemed moderate across domains (Table 2). Overall, bias in selection of participants into the study; bias in classification of interventions; bias due to deviations from intended interventions; bias due to missing data; and bias in measurement of outcome was low. However, bias due to confounding and bias in selection of the reported result was high. Risk of bias was not expected to compromise pooled results, as it appeared similar across treatment arms.

Table 2.

ROBINS-I risk of bias table

First author	Co	P	Ca	D	Mi	Me	R	O
Aksakal et al., 2013	Moderate	Low	Low	Low	Moderate	Moderate	Moderate	Moderate
Ay et al., 2005	Moderate	Moderate	Low	Moderate	Moderate	Low	Moderate	Moderate
Basbug et al., 2017	Serious	Serious	Low	Low	Moderate	Moderate	Moderate	Serious
Benedetti et al., 2013	Moderate	Low	Low	Low	Low	Low	Moderate	Moderate
Brahmamdan et al., 2011	Moderate	Moderate	Low	Low	Low	Moderate	Low	Moderate
Cambon-Binder, 2018	Moderate	Low	Low	Moderate	Moderate	Low	Low	Moderate
Chaturvedi, 2018	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Cheng, 2001	Moderate	Low	Moderate	Moderate	Moderate	Low	Moderate	Moderate
Cheng, 1995	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Moderate	Moderate
Clark, 2009	Moderate	Moderate	Low	Moderate	Low	Moderate	Moderate	Moderate
Clement, 1990	Moderate	Low	Low	Low	Moderate	Low	Low	Moderate
Delnoitis, 2018	Moderate	Moderate	Low	Low	Moderate	Low	Moderate	Moderate
Erhart, 2013	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Moderate	Moderate
Ermutlu, 2022	Moderate	Low	Low	Low	Moderate	Moderate	Moderate	Moderate
Ghasemzadeh, 2002	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Gosens, 2003	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Grant, 2018	Serious	Moderate	Moderate	Moderate	Serious	Moderate	Moderate	Serious
Harris, 2019	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Karlsson, 1986	Moderate	Moderate	Moderate	Moderate	Serious	Moderate	Serious	Serious
Khan, 2021	Moderate	Low	Low	Low	Low	Low	Moderate	Moderate
Korompilias, 2009	Moderate	Moderate	Low	Moderate	Moderate	Low	Moderate	Moderate
Kow, 2018	Moderate	Moderate	Moderate	Moderate	Serious	Moderate	Moderate	Moderate
Louahem, 2016	Moderate	Low	Low	Low	Moderate	Low	Low	Moderate
Luria, 2007	Moderate	Low	Moderate	Low	Low	Moderate	Moderate	Moderate
Mahdy, 2021	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Mangat, 2009	Moderate	Moderate	Low	Low	Low	Low	Moderate	Moderate
Matuszewski, 2014	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Low	Moderate
Rabee, 2001	Moderate	Low	Moderate	Low	Low	Moderate	Moderate	Moderate
Ramesh, 2011	Moderate	Moderate	Low	Low	Low	Moderate	Moderate	Moderate
Salvati, 2017	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Moderate	Moderate
Schoenecker, 1996	Moderate	Low	Low	Low	Low	Low	Moderate	Moderate
Soh, 2013	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Štichhauer, 2022	Moderate	Low	Low	Low	Low	Low	Moderate	Moderate
Tomaszewski, 2022	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Towler, 2010	Serious	Serious	Moderate	Moderate	Serious	Moderate	Serious	Serious
Tunku-Naziha, 2017	Moderate	Moderate	Low	Low	Low	Moderate	Moderate	Moderate
Vasli, 1998	Moderate	Moderate	Moderate	Moderate	Moderate	Moderate	Moderate	Moderate
Wegmann, 2014	Moderate	Low	Low	Low	Low	Low	Moderate	Moderate
Weller, 2013	Moderate	Moderate	Low	Low	Moderate	Low	Low	Moderate
Wong, 2020	Moderate	Low	Low	Moderate	Low	Moderate	Low	Moderate
Xie, 2021	Moderate	Low	Low	Low	Low	Moderate	Moderate	Moderate
Shaw, 1990	Moderate	Moderate	Low	Low	Low	Moderate	Moderate	Moderate
Sabharwal, 1997	Serious	Serious	Moderate	Moderate	Serious	Moderate	Moderate	Serious
Louahem, 2006	Moderate	Low	Moderate	Moderate	Low	Low	Moderate	Moderate
Lyons, 2000	Moderate	Moderate	Moderate	Low	Moderate	Moderate	Moderate	Moderate
Saglam, 2014	Moderate	Moderate	Low	Low	Moderate	Moderate	Moderate	Moderate
Choi, 2010	Moderate	Moderate	Low	Low	Low	Low	Low	Moderate
Noaman, 2006	Moderate	Low	Moderate	Moderate	Moderate	Moderate	Low	Moderate

Co bias due to confounding, P bias in selection of participants into the study, Ca bias in classification of interventions, D bias due to deviations from intended interventions, Mi bias due to missing data, Me bias in measurement of outcomes, R bias in selection of the reported result, O overall risk of bias

Overall supracondylar fractures

Overall, on the basis of current literature, it is evident that patients with pale pulseless limbs (75/82 versus 73/299, p < 0.001) (i.e., 75 pale who underwent surgical exploration, out of the total 75 + 7 pale; 73 pink who underwent surgical exploration, out of the total 73 + 226 pink), open fracture (20/23 versus 104/280, p < 0.001), pucker sign (26/26 versus 54/74, p = 0.003), and ecchymosis (20/21 versus 55/95, p = 0.002) were more likely to undergo surgical exploration. Patients with open fractures (15/24 versus 87/335, p < 0.001) and pucker signs (18/26 versus 33/129, p < 0.001) were also more likely to undergo open reduction.

The presence of neurological deficits in patients had no relation to them undergoing surgical exploration (104/234 versus 65/131, p = 0.382) or the type of reduction (73/291 versus 40/169, p = 0.738) (Table 3).

Table 3.

Overall current practices for patients with SHF

Overall current practices	Underwent surgical exploration	No surgical exploration	p-Value	Underwent open reduction	Underwent closed reduction	p-Value
Perfusion status	75 pale versus 73 pink	7 pale versus 226 pink	< 0.001	41 pale versus 16 pink	341 pale versus 76 pink	0.106
Fracture status	20 open versus 104 closed	3 open versus 176 closed	< 0.001	15 open versus 87 closed	9 open versus 248 closed	< 0.001
Pucker sign	26 positive versus 54 negative	0 positive versus 20 negative	0.003	18 positive versus 33 negative	8 positive versus 96 negative	< 0.001
Ecchymosis	20 positive versus 55 negative	1 positive versus 40 negative	0.002	16 positive versus 48 negative	5 positive versus 47 negative	0.051
Neurological deficits	104 positive versus 65 negative	130 positive versus 66 negative	0.382	73 positive versus 40 negative	218 positive versus 129 negative	0.738
Ecchymosis	20 positive versus 55 negative	1 positive versus 40 negative	0.002	16 positive versus 48 negative	5 positive versus 47 negative	0.051

Patients with pale pulseless limbs are likely to require more vascular intervention (63/70 versus 64/138, p < 0.001) and have a higher likelihood of complications (9/20 versus 7/90, p < 0.001). Patients with open fractures have a higher probability of requiring vascular intervention (19/22 versus 87/243, p < 0.001) (Table 4).

Table 4.

Overall current outcomes for patients with SHF

Overall outcomes	Underwent vascular intervention	No vascular intervention	p-Value	Complications	No complications	p-Value
Perfusion status	63 pale versus 64 pink	7 pale versus 74 pink	< 0.001	9 pale versus 7 pink	11 pale versus 83 pink	< 0.001
Neurological status	76 deficit versus 58 normal	82 deficit versus 61 normal	1.00	11 deficit versus 8 normal	44 deficit versus 54 normal	0.326
Fracture status	19 open versus 87 closed	3 open versus 156 closed	< 0.001	3 open versus 15 closed	4 open versus 92 closed	0.077

The presence or absence of neurological deficits does not affect complication outcomes (11/55 versus 8/62, p = 0.326) nor the need for vascular intervention (76/158 versus 58/119, p = 1.00) (Table 4).

Out of 63 patients with pale pulseless limbs, upon surgical exploration, there were 3 intact arteries, 7 trapped arteries, 51 disrupted arteries, and 2 unknown statuses. Out of 64 patients with pink pulseless limbs, these included 4 intact arteries, 38 trapped arteries, 14 disrupted arteries, and 8 unknown statuses. Out of 70 patients with pale pulseless limbs, 63 underwent vascular intervention, for which there were 45 vascular repairs, 0 stenting, 4 bypasses, 9 releases, and 7 thrombectomies. Out of 138 patients with pink pulseless limbs, 64 underwent vascular intervention, for which there were 25 vascular repairs, 9 stentings, 2 bypasses, 32 releases, and 3 thrombectomies. In summary, patients with pale pulseless limbs are more likely to have disrupted arteries (p < 0.001) and require more vascular repairs (p < 0.001) (Table 5).

Table 5.

Breakdown on the status of the arteries explored and the type of vascular intervention performed (perfusion status, fracture status)

	Pale pulseless limbs	Pink pulseless limbs	p-Value	Open fracture	Closed fracture	p-value
Artery status on exploration	63	64		19	87
- Intact arteries	3	4		1	4
- Trapped arteries	7	38		2	37
- Disrupted arteries	51	14	< 0.001	15	45	0.050
- Unknown status	2	8		1	1
Vascular intervention	63	64		19	87
- Vascular repairs	45	25	< 0.001	10	37	< 0.001
- Stenting	0	9		0	9
- Bypasses	4	2		4	2
- Releases	9	32		2	34
- Thrombectomies	7	3		3	5

Out of 19 patients with open fractures, upon surgical exploration, there were 1 intact artery, 2 trapped arteries, 15 disrupted arteries, and 1 unknown status. Out of 87 patients with closed fractures, these included 4 intact arteries, 37 trapped arteries, 45 disrupted arteries, and 1 unknown status. Out of 22 patients with open fractures, 19 underwent vascular intervention, for which there were 10 vascular repairs, 0 stenting, 4 bypasses, 2 releases, and 3 thrombectomies. Out of 243 patients with closed fractures, 87 underwent vascular intervention, for which there were 37 vascular repairs, 9 stentings, 2 bypasses, 34 releases, and 5 thrombectomies. In summary, patients with open fractures are more likely to have disrupted arteries (p = 0.050) and require more vascular repairs (p < 0.001) (Table 5).

Out of 26 patients with either a pucker or dimple sign, upon surgical exploration, there were 3 intact arteries, 19 trapped arteries, and 4 disrupted arteries. Out of 46 patients without either a pucker or dimple sign, upon surgical exploration, there were 0 intact arteries, 8 trapped arteries, and 38 disrupted arteries. Out of 21 patients with either a pucker or dimple sign who underwent vascular intervention, there were 5 vascular repairs, 15 releases, and 1 thrombectomy. Out of 48 patients without either a pucker or dimple sign who underwent vascular intervention, there were 36 vascular repairs, 1 bypass, 8 releases, and 3 thrombectomies. Patients with pucker or dimple signs are more likely to have trapped arteries (p < 0.001) and also more likely to undergo surgical release of their trapped arteries (p < 0.001) (Table 6).

Table 6.

Breakdown on the status of the arteries explored and the type of vascular intervention performed (pucker sign, ecchymosis)

	Positive pucker sign	Negative pucker sign	p-Value	Positive ecchymosis	Negative ecchymosis	p-Value
Artery status on exploration	26	46		20	47
Intact arteries	3	0		3	0
Trapped arteries	19	8	< 0.001	13	13	< 0.001
Disrupted arteries	4	38		4	34
Unknown status	0	0		0	0
Vascular intervention	21	48		15	49
Vascular repairs	5	36		5	32	0.107
Stenting	0	0		0	0
Bypasses	0	1		0	1
Releases	15	8	< 0.001	9	13
Thrombectomies	1	3		1	3

Out of 20 patients with ecchymosis present, upon surgical exploration, there were 3 intact arteries, 13 trapped arteries, and 4 disrupted arteries. Out of 47 patients without ecchymosis present, upon surgical exploration, there were 0 intact arteries, 13 trapped arteries, and 34 disrupted arteries. Out of 14 patients with ecchymosis who underwent vascular intervention, there were 5 vascular repairs, 0 stenting, 0 bypasses, 9 releases, and 1 thrombectomy. Out of 48 patients without ecchymosis who underwent vascular intervention, there were 32 vascular repairs, 0 stenting, 1 bypass, 13 releases, and 3 thrombectomies. Patients with ecchymosis are more likely to have trapped arteries (p < 0.001) (Table 6).

Pink pulseless supracondylar fractures

When we perform subgroup analysis focusing specifically on patients with pink pulseless hands, those with median nerve deficit (12/38 versus 7/70, p = 0.008) and those with ulnar nerve deficit (2/3 versus 9/92, p = 0.035) are more likely to undergo surgical exploration. Patients with open fractures (1/10 versus 121/369, p = 0.014) were also more likely to undergo open reduction (Table 7).

Table 7.

Current practices for pink pulseless patients with SHF

Pink pulseless current practices	Underwent surgical exploration	No surgical exploration	p-Value	Underwent open reduction	Underwent closed reduction	p-Value
Fracture status	1 open versus 33 closed	3 open versus 97 closed	1.00	1 open versus 121 closed	9 open versus 248 closed	0.014
Pucker sign	6 positive versus 8 negative	0 positive versus 0 negative		3 positive versus 5 negative	6 positive versus 60 negative	0.052
Ecchymosis	11 positive versus 11 negative	20 positive versus 20 negative		18 positive versus 18 negative	13 positive versus 13 negative
Neurological deficits	21 positive versus 22 negative	76 positive versus 61 negative	0.486	16 positive versus 16 negative	121 positive versus 102 negative	0.707
Median nerve status	12 injury versus 7 without	26 injury versus 63 without	0.008	15 injury versus 17 without	65 injury versus 53 without	0.431
Ulnar nerve status	2 injury versus 9 without	1 injury versus 83 without	0.035	0 injury versus 24 without	6 injury versus 68 without	0.331

Subgroup analysis of outcomes for patients with pink pulseless hands also had several significant outcomes, especially for patients with and without concomitant neurological injuries. Pink pulseless patients with neurological deficits did not require more vascular intervention (15/31 versus 22/72, p = 0.117), and there is also no difference in complications suffered (2/26 versus 4/40, p = 1.00) The fracture status does not affect complication outcomes (0/1 versus 7/74, p = 1.00) nor the need for vascular intervention (1/4 versus 24/88, p = 1.00) (Table 8).

Table 8.

Current outcomes for pink pulseless patients with SHF

Pink pulseless outcomes	Underwent vascular intervention	No vascular intervention	p-Value	Complications	No complications	p-Value
Neurological status	15 deficit versus 22 normal	16 deficit versus 50 normal	0.117	2 deficit versus 4 normal	24 deficit versus 36 normal	1.00
Fracture status	1 open versus 24 closed	3 open versus 64 closed	1.00	0 open versus 7 closed	1 open versus 67 closed	1.00

Out of the 15 patients with neurological deficits, upon surgical exploration, there were 1 intact artery, 10 trapped arteries, 2 disrupted arteries, and 2 unknown statuses. Out of the 22 patients without neurological deficits, upon surgical exploration, there were 0 intact arteries, 13 trapped arteries, and 9 disrupted arteries.

There is no significant difference in overall complications between those who underwent exploration of arteries and those who did not (p = 0.230). Complications analyzed included eventual fracture reduction, cubitus varus or valgus deformities, restricted range of motion, neurological deficits, presence of compartment syndrome, and Volkmann ischemic contractures.

All of the 73 patients who underwent exploration and 226 patients who did not undergo exploration did not require subsequent vascular intervention. All of the 30 patients whose pulses returned postoperatively immediately (fixation and arterial exploration) and the 3 patients whose pulses did not return postoperatively immediately had no subsequent vascular exploration. All of the 17 patients whose pulses returned postoperatively immediately (fixation and arterial exploration) did not have any overall complications.

Discussion

In this study, our systematic review and meta-analysis analyzed the indications for and outcomes of interventions performed across 48 studies in 1070 individuals with pulseless supracondylar fracture of the humerus. Several significant clinical predictors were found to be associated with a higher incidence of clinical decisions for surgical exploration, including the presence of pale pulseless limb, open fracture, pucker sign, and ecchymosis. Expedient surgical exploration and vascular reconstruction may be required in pale pulseless hands due to a greater concern for the development of compartment syndrome, ischemic contracture, and delayed growth of the limb should arterial reconstruction be delayed [9, 14]. Skin puckering and antecubital ecchymosis, as a result of fracture fragments piercing the brachialis, may similarly heighten the suspicion for impending compartment syndrome or vascular compromise, warranting immediate surgical exploration [15].

Previously, a similar individual patient data meta-analysis by White et al. concluded that absence of pulse after supracondylar humerus fracture is a strong indicator of vascular injury [16]. Thus, the authors recommended an aggressive approach to a pulseless hand by exploration and repair, given the fact that the success rate for revascularization is sufficiently high.

However, our overall suggestion would be that following closed reduction for the treatment of pink pulseless SHFs, watchful observation instead of aggressive vascular exploration is recommended as the next step in clinical management [17]. Similar to the opinions of Xie et al. [18], Weller et al. [19] and Griffin et al. [7], our results demonstrated that close observation after urgent closed reduction and percutaneous pinning is a sufficient approach, and that surgical exploration is not necessary as long as the hand is warm and well perfused. From our analysis, there was no difference in outcomes with or without exploration of the arteries, and the majority of the patients had their pulses returned within 48 h. The pink perfusion status likely suggests that collateral circulation will suffice for subsequent recovery without further adverse outcomes. Unnecessary surgical procedures might also increase the risk of adverse outcomes including bleeding, thrombosis, infections and pose additional healthcare costs for both patients and healthcare providers.

Patients and surgeons should maintain close careful observation on clinical predictors of vascular damage such as pale perfusion, which will be an indication for surgical exploration due to the higher likelihood of arterial disruption. Besides the color of the fractured hand, other red flags or clinical predictors must also be noted carefully, such as the presence of open fracture, pucker sign, and ecchymosis mentioned above. Surgical exploration may be suggested when these concomitant injuries appear due to a higher likelihood of brachial injury.

Adjunctive measures include colored-flow Doppler ultrasound, which may provide additional clinical parameters for deciding whether to proceed with vascular exploration should be considered. These first-line investigations are of lower cost and may be a suitable wait-and-watch approach while providing greater assurance to patients. These assessments were only reported in a small number of studies and a limitation of our paper is that we cannot determine how useful they are in guiding decision-making.

In addition, among patients with pink pulseless SHF, when there were neurological deficits, the most common being median nerve injury, clinicians associated these with a higher likelihood of adverse outcomes and were more likely to propose aggressive vascular exploration. One reason could be the close relationship between the natural course of arteries and nerves in the arm, and it was assumed that an injury to one would likely affect the other on the basis of their close proximity. However, on the basis of our results, we disagree with this approach and would rather recommend watchful waiting as well. Neurological deficits in fact did not significantly predict greater incidence of arterial injuries, more vascular intervention, or complications suffered. It can be observed that neurological deficit was not significant in the management regime or the recovery outcomes, thus surgical exploration is not necessarily needed in isolated cases of nerve injuries without any of the red flags above. Further study could be conducted to determine the recommended wait and watch time before the risks of arterial damage become greater.

Our study has several limitations. Firstly, the quality of included studies may vary, due to the inclusion of case studies, absence of randomized controlled trials in the analysis, and bias toward publishing positive results. Secondly, there is variability in study designs that contributes to heterogeneity of results. However, we have used the ROBINS-I tool to assess the risk of bias and it is deemed to be moderate. Lastly, it would be ideal for studies to report long-term outcomes following treatment, so as to fully understand the efficacy of treatment over time.

Conclusions

The presence of a pale pulseless limb, open fracture, pucker sign, or ecchymosis are significant clinical predictors of arterial disruption and necessitate expedient surgical exploration. Conversely, for patients with pink pulseless supracondylar humeral fracture, watchful waiting is a safe and appropriate strategy. Our data support conservative management for this subset, as there was no significant difference in complication rates nor outcomes between patients who underwent immediate exploration and those who were observed. Thus, in the absence of specific red flag predictors, patients with pink pulseless SHF should be monitored closely for neurovascular status without the need for immediate surgical exploration.

Abbreviation

SHF

Supracondylar humeral fractures

Author contributions

CHY, CJG, and LHX were involved with data curation, formal analysis, and writing of the manuscript. CHY and CJG are joint first authors in this manuscript. TSHS, LAKS, and HHPJ were involved with the conceptualization, supervision, and review and editing of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Availability of data and materials

The dataset supporting the conclusions of this article is available upon request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

No conflicts of interest are reported by the authors.