Current controversies in management of fracture neck femur in children: A review

Abstract

Paediatric femur neck fracture is an uncommon injury and is known for high rate of complications. This in turn, can have significant impact on development of hip and overall function of the child. Controversy prevails in many areas of management and it still remains a difficult injury to manage. Through this paper, an attempt has been made to summarize the current concepts in management and suggest current recommendations regarding the controversies so that these injuries can be managed judiciously. A systematic review was done as per PRISMA guidelines using pre-defined inclusion and exclusion criteria. 18 studies with better scientific evidence after quality assessment were included in the systematic review. Current trends and Controversial issues in management were then identified and discussed. It was observed that existing literature is inconclusive regarding several aspects of management of this injury with no clear guideline available. However, certain recommendations useful for decision making could be made. These injuries should be managed aggressively with operative fixation at the earliest and one should not hesitate to open reduce if acceptable alignment is not obtained after one or two gentle closed manipulations. Choice of implant and their configuration in neck can vary depending on age and weight of patient, type of injury and surgeon’s preference. Initial fracture displacement could be considered to be most predictive for development of osteonecrosis of the femoral head, whereas, role of surgical decompression, type of fixation and timing of surgery is still debatable. Functional outcome is primarily affected by osteonecrosis of the femoral head, nonunion and severe coxa vara, whereas mild coxa vara, shortening, and premature closure of physis when considered alone, don’t have significant influence on functional outcome in short term. Treatment of fracture neck femur in children is still controversial in many aspects and needs further research. It should be understood that complications can occur regardless of the method of treatment and might reveal their full impact many years after injury. Hence, a guarded prognosis should be explained to the parents and care-givers at the time of injury.

1. Introduction

Fracture neck of femur is uncommon in paediatric population and accounts for only 1% of all fractures in children.¹ In contrast to elderly where majority of these injuries occur due to low energy trauma, about 80–90% of these fractures in children are due to high velocity trauma and commonly present with other associated injuries.^2,3 The most common classification system used is that of Delbet (1907)⁴ which was later popularized by Colonna in 1927.² It helps in planning of treatment and has prognostic importance. These fractures are known for their high rate of complications like avascular necrosis, coxa vara and non-union due to compromised vascularity of femoral head and biomechanical challenges in operative fixation.^3,5,6 Controversies of management still prevail regarding surgical timing, need for capsulotomy, crossing of physis for stabilization, treatment of late presenters and choice of implant with support of literature in either way.^7,8 Thus, in absence of established guidelines, this injury remain a challenge to manage. With available literature, we attempted to discuss existing controversies and suggest recommendations in management of these challenging fractures.

2. Material and methods

This study was conducted as per PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines.⁹ A Medline database was searched using PubMed (http://www.ncbi.nlm.nih.gov/pubmed), EMBASE (http://www.elsevier.com/online-tools/embase) and Google Scholar with different combinations of MeSH terms like “femur neck fracture, children, controversies, management, complications” over a period of six months from June 2019 to November 2019. Related Cochrane reviews and systemic reviews were searched (http://www.cochrane.org) for additional references. References of all eligible studies were explored further to get additional studies. An attempt was made to eliminate risk of bias in selecting studies through factors based on reliability, internal consistency and empirical evidence if available.

The selection criteria were determined before initiating literature search.

2.1. Inclusion criteria

Studies published in English language.

Studies on paediatric femoral neck fractures (age < 16 years).

Studies with sample size more than 20.

Studies with a minimum average follow up of 2 years.

This included all original case series, randomized clinical trials and their abstracts. The limit of 20 patients was applied for significant statistical inference.

2.2. Exclusion criteria

Case reports, reviews, letter to the editor and unpublished data.

Studies in patients with previous hip anomalies.

Studies on fracture dislocation of hip.

Differences if any were discussed and resolved among both authors. Full text of retrieved articles were reviewed and controversial issues in management were identified. The quality assessment of selected studies was done with modified Jadad scale¹⁰ for randomized control trials and GRADE ¹¹ (Grading of Recommendations, Assessment, Development and Evaluation) approach for case series. Studies with modified Jadad score ≥6 or High quality according to GRADE approach were then selected to write the manuscript.

3. Results

1148 studies were screened initially and after further screening, 158 studies of Paediatric hip injuries were identified (Fig. 1). 62 studies met the inclusion criteria and after quality assessment, 18 were selected to write the final manuscript (Table 1).^3,5,6,12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26 The largest series was by Ratliff et al.³ (n = 71). Average age of children at time of surgery varied from 8.2 to 13.2 years (mean of 10.5 years). All studies used Delbet’s classification⁴ for determining type of injury. Most common was Delbet type II (52.12%) followed by Delbet type III (37.94%). Most common mechanism of injury were road traffic accident and fall from height while most common mode of treatment was closed reduction and osteosynthesis (319/654 patients) followed by open reduction and osteosynthesis (217/654 patients). The average time duration between injury and surgical procedure was 4.03 days (range from 4 hours–21 days). For internal fixation, lateral approach to proximal femur was used by all and the same was extended anteriorly (Watson Jones) for an open reduction when required. Most common implant used were cannulated screws followed by smooth k wires. Decompression of fracture hematoma was done by Qi et al.²³ and Ju et al.¹⁵ routinely in all patients, whereas Bukwa et al.²⁴ did it in 21 out of 28 patients. Decompression happened spontaneously with open reduction of fractures. 11 out of 16 studies used postoperative immobilization. In 7 studies, a hip spica was used for children under 10 years of age while 2 studies used it routinely for all patients. Qi et al.²³ used skin traction whereas Ju et al.¹⁵ used a brace for immobilization.

Fig. 1.

PRISMA format of literature search.

Table 1.

Detailed characteristics of the included studies. C – Conservative, ORIF – Open Reduction & Internal Fixation, CRIF – Closed Reduction & Internal Fixation, AVN – Avascular Necrosis, SCFE – Slipped Capital Femoral Epiphysis, LLD – Limb Length Discrepancy, IF – Implant Failure, FAI – Femoro-Acetabular Impingement, SA – Septic Arthritis, CV – Coxa Vara, VO – Valgus Osteotomy, NU – Non Union, DU – Delayed Union, PPC – Premature Physeal Closure, C Val – Coxa Valga, HHS – Harris Hip Score, VGTPF – Vascularized Greater Trochanter Periosteal Flap, NL – Neck Lengthening, NS – Neck Shortening.

S no	Series	Sample size (n)	Gender (M/F)	Mean Age at surgery (Years)	Mean delay from injury to surgery (Days)	Classification (Delbet) I/II/III/IV	Mode of treatment	Average follow up (Months)	Complications	Functional outcome (Ratliff) Good/Fair/Poor
1	Ratliff et al3 (1962)	71	Not specified	11–12	Not specified	2/38/26/4	C 40, ORIF 1, CRIF 21, VO 4	48–60	30 AVN, 7 NU, 14 CV, 11 PPC	31/21/20
2	Spence et al5 (2016)	70	41, 29	13.2	48 within 24 Hours	5/31/29/5	C 1, ORIF 38, CRIF 31	33.5	20 AVN, 2 SCFE, 2 LLD, 1 IF, 1 FAI,1 SA	59/5/6
3	Dendane et al12 (2010)	21	14, 7	12.1	4.8	0/9/10/2	ORIF 13 CRIF 8	26.4	6 AVN, 5 CV, 4 LLD, 1 NU, 1 SA	14/1/6
4	Morsy et al6 (2001)	53	25, 28	10.2	Not specified	1/28/21/3	C 9, IF 38, VO 6	112	21 AVN, 19 NU, 19 CV, 20 PPC, 29 LLD, 4 SA, 5 C Val	26/13/14
5	Varshney et al13 (2009)	21	14, 7	11.8	8.1	0/14/7/0	ORIF 8, CRIF 13	81	3 AVN, 2 LLD, 1 PPC	11/7/3
6	Azam et al14 (2009)	22	14, 8	8.9	11.2	0/13/9/0	ORIF 8 CRIF 14	70.8	8 AVN, 4 NU	14/2/6
7	Ju et al15 (2016)	58	40, 18	9.1	4.6	0/30/21/7	ORIF 37 CRIF 21	35.1	11 AVN, 3 CV, 1 NU	29/16/13
8	Bombaci et al16 (2006)	22	11, 11	10	7	2/11/9/0	ORIF 2 CRIF 20	73	12 AVN, 14 PPC, 3 CV, 1NU, 1 DU, 1SA	12/8/2
9	Eberl et al17 (2010)	22	10, 12	12	Not specified	0/7/15/0	ORIF 10 CRIF 12	48	2 AVN, 12 CV	HHS Not Specified
10	Song et al18 (2010)	27	11, 16	9.7	24 within 24 Hours	0/15/12/0	ORIF 15 CRIF 12	33.7	2 AVN, 2 NU, 2 CV	21/3/3
11	Stone et al19 (2015)	22	13, 9	11.5	18 within 24 Hours	0/13/8/1	ORIF 6 CRIF 16	25.2	8 AVN, 2 NU, 2 CV	17/1/4
12	Astur et al20 (2010)	29	18, 11	8.6	15 within 24 Hours	2/19/5/3	C 4, ORIF 24 CRIF 1	42	5 AVN, 8 CV, 6 LLD	17/4/8
13	Inan et al21 (2009)	39	22, 17	11.1	1.5	0/21/14/4	C 2, ORIF 5, CRIF 32	40.8	11 AVN, 1 SA	28/4/7
14	Bali et al22 (2011)	36	20, 16	10	Not specified	0/16/11/9	C 13 ORIF 18 CRIF 13	38.4	7 AVN, 3 CV, 3 NU, 1 PPC, 1 SA, 1 Screw perforation	27/7/2
15	Qi et al23 (2008)	23	15, 8	10.2	1.2	2/12/9/0	OR + VGTPF 23	36	3 AVN, 1 PPC	17/3/3
16	Bukwa et al24 (2015)	28	16, 12	10.75	16 within 12 Hours	1/8/16/3	C 5, ORIF 5, CRIF 18	108	11 AVN, 6 CV, 8 PPC	14/4/10 (HF Grade)
17	Kuo et al25 (2016)	30	20, 10	8.2	Not specified	5/17/8/0	ORIF 1 CRIF 29	58.8	11 AVN, 12 NL, 6 PPC, 5 NS, 3 CV, 2 C Val	16/11/3
18	Togrul et al26 (2005)	62	35, 27	10.2	57 within 48 Hours	1/22/25/14	C 9 In situ fixation 30 ORIF 3 CRIF 20	168	9 AVN, 2 C Val, 1 NU, 5 PPC, 5 CV, 7 LLD	Charnley Hip scores Pain 6, Motion 5.6, Walking 5.2

Average follow up of patients in this study varied between 25.2 and 168 months (mean 60.26 months). Avascular necrosis of femoral head was the most common complication and was seen in 180 cases (27.52%). Furthermore, as per Ratliff classification for AVN of femoral head, type 1 AVN was most commonly seen followed by type 2. This was followed by coxa vara (13.51%) and premature physeal closure (10.47%). Nonunion was seen in 41 cases (6.26%) whereas 50 (7.64%) had limb length discrepancy in the range of 1–5 cm. Other uncommon complications reported were; septic arthritis,^5,6,12 slipped capital femoral epiphysis,⁵ femoroacetabular impingement,⁵ neck shortening,²⁵ coxa valga^6,25,26 and overgrowth of femoral neck.²⁵ Ratliff’s method was used for evaluation of functional outcome by all except Eberl et al.,¹⁷ Bukwa et al.,²⁴ and Togrul et al.²⁶ According to Ratliff’s criteria, 62.54% (339 out of 542) patients had good, 19.5% fair and 18.4% poor functional outcome respectively.

4. Discussion

Fracture neck femur in children needs aggressive management with anatomical reduction and stable internal fixation.22, 23, 24 This approach offers best opportunity for preserving femoral head and avoiding complications like nonunion, coxa vara and avascular necrosis of femoral head (AVN). However, conservative treatment can be considered for incomplete and undisplaced fractures in very small children (<2 years) and for those unfit for surgical intervention.^3,6,26 Similarly, there is consensus for attempting closed reduction of fracture by gentle manipulation and proceeding to an open reduction if anatomic reduction (<2 mm of displacement and <5° of angulation) is not obtained after one or two gentle closed manipulations.^5,12,18 An attempt to achieve reduction percutaneously using k-wires as joystick may be successful.²⁷ Other aspects of management like surgical timing, role of capsulotomy and post-operative care remains debatable. Recent reviews by Palocaren et al.⁷ and Patterson et al.⁸ have discussed some of these controversies while issues like implant configuration, need for implant removal and management of late presenters were untouched.

4.1. Timing of surgery

Mitchell et al.²⁸ in 1936, stressed for early fixation of proximal femoral fractures in children. They suggested that early reduction and fixation takes care of kinked blood vessels especially in a displaced fracture which in turn, decrease incidence of AVN and improve functional outcome. A similar significant correlation for fixation within 24 h of injury was reported by seven studies in current review.12, 13, 14^,16,17,24,26 Azam et al.¹⁴ further observed a progressive increase in rate of AVN with delay from first to third week and believed it to be due to added manipulation required to achieve reduction. However, others considered direct trauma to vessels at the time of injury, as main determining factor for occurrence of AVN which could not be avoided even with early treatment.^3,20,21 Furthermore, Spence et al.⁵ reported greater incidence of AVN in those undergoing surgery within 24 h. Thus, usefulness of early definitive fixation remains debatable. Based on current evidence, we suggest intervention, as early as logistically feasible in the hospital setting. The trends have been to avoid operating in the middle of the night, especially where there is no definite advantage as suggested by Ratliff et al.³ and Astur et al.²⁰ in their observations. Displaced fracture neck of femur however must be dealt with as an emergency within 12 h if possible to reduce the chances of AVN (Bukwa et al., 2015).

4.2. Role of capsulotomy/joint aspiration

The role of joint aspiration and open decompression to prevent osteonecrosis is unclear. Only one study reported significant association between decompression done within 12 h of injury and reduced rate of AVN.²⁴ Moreover, better outcome was observed in those undergoing open drainage when compared to aspiration. Four studies reported better outcome and less AVN in patients undergoing open reduction when compared to closed reduction due to simultaneous decompression occurring with open reduction.^15,18,19,23 Others didn’t find any difference in rate of osteonecrosis with or without capsulotomy.^5,13,21 Furthermore, displaced fractures may already have lower intracapsular pressure due to possible disruption of hip capsule. Hence, in absence of consensus, urgent decompression as a routine is not advised. However, percutaneous aspiration of hemarthrosis at the time of fixation can be considered for undisplaced and minimally displaced fractures for the possible advantage of reducing subsequent AVN.²⁴ It is less invasive, safe and effective in decreasing intra-articular pressure.²⁹

4.3. Choice of implant

A variety of implants were used including smooth pin, Knowles pin, screw, Paediatric DHS and plate.12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26 Choice of implant was determined by age of the child, Body Mass Index (BMI), type of fracture and operating surgeon’s preference. None of the studies found a significant correlation between type of osteosynthesis and functional outcome. As per current evidence, use of screw should be avoided wherever possible in younger children with Delbet type 1 fracture as it can further jeopardize an already injured physis thereby, producing significant shortening and deformity.^17,25 Multiple smooth k-wires or pins supplemented with a hip spica is ideal in this situation. Use of screws can be considered in older children around the age of maturity, those with high body mass index and when stability seems unsatisfactory. For Delbet type 2 and type 3, use of three threaded wires supplemented with hip spica is preferred in pre-school age group.^15,17 A combination of 4.0 mm cannulated screw and wire can also be used.¹⁸ Three 4.0 mm cannulated screws and three 6.5 mm screws are suggested for children over 6 years and at puberty respectively.^17,20,21 Delbet type 4 injuries can be fixed effectively with locking compression paediatric hip plate or a paediatric DHS.^5,15,19 Use of washer with screw is recommended as it prevents the screw head from penetrating cortex and improve purchase in femoral head.³⁰ Titanium is more biocompatible, less stiff and has lower infection rate than stainless steel. Moreover, the size of artefact produced in Magnetic Resonance imaging is significantly small.³¹ Hence, use of titanium screw is recommended as it promotes fracture healing and improves quality of post-operative assessment and detection of avascular necrosis of femoral head.

4.4. Screw configuration in neck

Different configurations were reported with use of one, two or three screws in different situations without citing advantage of one over other.^12,13,17 An inverted triangle configuration was most commonly used with use of three screws whereas a vertical configuration was used with two screws. Basile et al. ³² compared use of two verses three screws in adults and found no difference in healing or failure rate. In a biomechanical study on synthetic bone, Selvan et al.³³ observed a significantly higher peak and ultimate vertical load with triangular configuration when compared to two screws but no significant difference was found between four different triangle configurations. Li et al.³⁴ found inverted triangle configuration to have mechanical advantage and less cut out rate. This is supported further by AO-ASIF group which recommends an inverted triangle configuration as the most stable option.³⁵

Booth et al.³⁶ compared biomechanics of central and calcar position of inferior screw and found significantly better stability, load, stiffness, and displacement in all tested parameters for the calcar screw. Lindequist et al.³⁷ compared the stability offered by posterior superior screw in two different anatomical positions and reported that a posterior position with cortical support increases the stability when compared to a more central screw. Most authors have advocated that the screws should be placed parallel to each other. However, Spangler et al.³⁸ didn’t find it a predictable risk factor for nonunion.

Based on these evidences, we suggest using three screws in inverted triangle configuration for older children. Inferior screw should be placed in the center of neck along the calcar. The posterior superior and anterior superior screws should be placed as peripheral as possible in lateral plane and centrally in antero-posterior plane. Two parallel screws placed centrally in lateral plane and one above another in AP plane supplemented with a spica cast should be sufficient for younger children.

4.5. Crossing of physis

Extension of fixation beyond physis particularly in Delbet type 2 injuries is debatable. Most studies avoid doing so in children less than 10 years of age due to fear of physeal arrest.^{12,13,15,16,23} Smooth pins were preferred over screws, when crossing the physis was inevitable and fixation was further supplemented with a hip spica .^12,15 However there is consensus for choosing stability of fixation over preservation of physis and for using screws across the physis if required to get a stable fixation.^7,8,17,39This is based on the observation that amount of shortening due to physeal arrest is not significant in majority of cases (average 1.5–2 cm) and does not affect functional outcome. An unstable fixation may lead to more severe complications like AVN, nonunion and coxa vara. Combination of physis sparing screw and smooth wires across the physis could be another viable alternative.^5,18,20,21

4.6. Post-operative care and need for implant removal

Additional stabilization after fracture fixation is important in younger, noncompliant children and in those where stability of fixation is inadequate.^13,14,16,21 Most studies suggested a single hip spica for children younger than 10 years irrespective of the type of osteosynthesis.^13,14,16,21 Some authors used brace while others used traction to keep children in bed.^15,23 This protocol ensures compliance about non-weight bearing and thereby avoids loss of reduction and subsequent coxa vara.¹⁷ Current practice is to initiate partial weight bearing at 6 weeks and full weight bearing by 12 weeks after surgery in the presence of satisfactory fracture healing.^14,18

Necessity for implant removal and correct time to do so was not discussed by any study in the current review. Literature suggests implant removal at an average of 12 months (9–22 months) after fixation in presence of satisfactory union.^23,40,41 This is due to the fact that screw removal becomes difficult with time and it is difficult to manage a secondary fracture with implant in situ. However, it is not considered a necessity by others as burden of additional surgery increases the cost and morbidity to patient. It also carries risk of fracture through screw removal site.^41,42 Kovara et al.⁴¹ classified the indications for implant removal as medical (infection, mechanical problems and implant failure) and non-medical (patient’s demand, metal sensitivity and foreign body sensation) and suggested to avoid removal in non-medical indications due to high rate of complications (28%). Based on challenges of removing titanium screws, we suggest implant removal once the fracture has healed and child resumes normal activity. Once removed, protected weight bearing is advised for four weeks for the screw tracks in the lateral cortex to heal.

4.7. Treatment of delayed presentation

Inadvertent delay in fixation is common in developing countries for a variety of reasons and a general agreement about their management is lacking. Depending on duration of delay, they can be termed as late presenters (24 hours–3 week), neglected (3 weeks–6 months) or nonunion (>6 months).^39,43, 44, 45, 46, 47 For cases presenting well up to 3 weeks of injury, most authors prefer an attempt at closed reduction and proceed to open reduction and internal fixation if closed reduction fails.14, 15, 16

For delay of over 3 weeks, internal fixation alone is likely to have high failure rate.^39,44 The existing literature suggests different techniques to augment fixation and improve biology and mechanics. It includes vascularized muscle pedicle grafts, vascularized bone graft (iliac crest or fibula), non-vascularized (tibia or fibula) strut graft and various types of valgus osteotomies done alone or in combination. Vascularized bone grafts, though confirming excellent results, is technically demanding.²³ Valgus osteotomy and fibular strut graft are the two most commonly done procedures and there is argument over which one to prefer. Another conflict is regarding open or closed reduction of pseudarthrosis. Excellent results of closed reduction, valgus osteotomy and fixation without any bone graft has been reported in literature.^43,44,46,47 It is based on Powel’s concept that converting the shearing forces of pseudarthrosis into compression forces helps in healing of pseudarthrosis. Nagi et al.⁴⁵ and Algeidi et al.³⁹ combined screw osteosynthesis and fibular graft without valgus osteotomy and reported 100% union rate. Addition of valgus osteotomy was advised if neck-shaft angle was less than 110°. Fibula takes care of neck resorption, provides additional stability and augments bone healing. Additionally, it is known to take care of early osteonecrosis. Management of AVN accompanying nonunion is further controversial. Based on current evidence, we suggest guidelines for management of delayed presentation of fracture neck femur in children (Fig. 2).

Fig. 2.

Algorithm for management of delayed presentation of fracture neck femur in children. CRIF – Closed Reduction and Internal Fixation, ORIF – Open Reduction and Internal Fixation. ∗Added in the presence of neck resorption and/or early osteonecrosis.

4.8. Factors determining complications and functional outcome

Fracture neck femur in children is known to have a high incidence of complications like AVN, coxa vara, nonunion, delayed union, premature physeal closure, limb length discrepancy and arthritis of hip joint.^5,6,12,16 A conclusive understanding of associated factors for these is lacking.

4.8.1. Avascular necrosis

AVN is the most common complication following fracture neck femur in children with a reported incidence of 0–92% in literature.^{3,16,18,48,49} Average incidence of AVN in current review was 28.88% (7.4%–54.5%). A significant association between osteonecrosis and poor functional outcome was observed in most studies.^3,6,13,22,25 Ratliff type 1 AVN (total involvement of epiphysis) was the most common and had worst functional outcome.^3,21,24 Many predictive factors for occurrence of AVN have been suggested.^3,5,6,16 These can be grouped into non-modifiable and modifiable factors. The non-modifiable factors include age of child, type of fracture (Delbet classification) and initial displacement of fragments. Modifiable factors include timing of surgery, type of reduction (open/close), quality of reduction (anatomical/non anatomical), type of fixation (implant, spica), and decompression (drainage of fracture hematoma). Moon et al.⁵⁰ in their meta-analysis suggested type of injury (Delbet classification) and age of child as the most significant predictors. However, we found only 2 out of 18 studies showing statistical significance for each.^5,12,20,25 Others either didn’t find them significant or didn’t analyze them. Rather, initial displacement of fracture at the time of injury was found to be statistically significant in 5 studies.^3,5,6,20,22 On the other hand, similar studies with displaced fractures showed good results and low rates of AVN.^12,23 Furthermore, cases of AVN have been reported in undisplaced fractures as well.3, 22 Similarly, consensus was not observed about modifiable factors with studies both supporting and opposing them being available. Thus, the most definite determining factors causing AVN remains controversial. Treatment options for osteonecrosis includes observation, prolonged non-weight bearing or surgical intervention in the form of osteotomy, arthrodesis or Arthroplasty after maturity.^51,52 However, studies comparing the functional outcome of these treatment modalities is lacking and hence their superiority over each other cannot be commented upon.

4.8.2. Coxa vara

Coxa vara is the next common complication and current review shows an incidence of 13.54%. It is unanimously agreed that incidence of coxa vara increases with conservative management and can be reduced with internal fixation.^6,26 However, cases have been reported even in patients treated surgically thus, suggesting other causes like improper reduction and/or fixation, physeal injury, osteonecrosis, nonunion and implant failure.^6,12,15, 16, 17, 18 Ju et al.¹⁵ noticed all coxa vara in closed reduction group and related it to poor reduction. Morsy et al.⁶ noticed it in seven patients even after internal fixation. Dendane et al.¹² noted that it was always combined with cephalic necrosis and Togrul et al.²⁶ described premature physeal closure as one of the etiological factors. Eberl et al.¹⁷ suggested extending the fixation across growth plate in Delbet type 2 injuries with short medial fragment and use of interlocking implants in unstable fractures to avoid this complication. Use of spica cast post-operatively can also be useful.^13,16,20 A satisfactory functional outcome and spontaneous improvement of deformity can be anticipated in a child younger than 8 years and with neck-shaft angle more than 110°.¹⁶ Older age and more severe deformities are known to produce shortening and Trendelenberg gait, thereby compromising the functional outcome which can be effectively corrected with a valgus osteotomy.^16,20,22

4.8.3. Nonunion

The incidence of nonunion in children is much less common than in adults due to presence of thick functional periosteum. However, once developed, it is known to be associated with poor outcome.⁴⁹Initial displacement, inadequate reduction and poor fixation are considered to be primary causes.^3,6,16,22The incidence of nonunion in current review was 6.75% and is managed like the late presenters as discussed earlier.

4.8.4. 4.8.4 Premature physeal closure

Premature physeal closure (PPC) was seen in 10.47% of patients. A Significant correlation between type 3 AVN and PPC was observed in many studies.^3,6,16 Morsy et al.⁶ and Togrul et al.²⁶ reported epiphyseal plate penetration with fixation device as a cause for developing PPC. However, when considered alone, PPC was not found to have significant correlation with poor outcome.^13,49 Thus PPC can develop in absence of growth plate penetration and still lead to good functional outcome. Hence one should not hesitate to involve physis in fixation if necessary, and choose stable fixation over physeal preservation.¹⁷

4.8.5. 4.8.5 Limb length discrepancy

The amount of shortening after fracture neck of femur in children depends on the age of child at the time of injury. Causes can be nonunion, coxa vara, AVN and PPC alone or in combination.^6,12,13 Previous studies report an average shortening of 1.5–2 cm (range 1–5 cm) and no significant relation with clinical outcome.^6,13 Shortening of more than 2 cm is usually seen in younger children and in combination with AVN and coxa vara.²⁶

5. Conclusion

The existing literature is inconclusive regarding many facets of this injury in children. However, certain recommendations useful for decision making could be made. The initial fracture displacement is most predictive for development of osteonecrosis of the femoral head, whereas, role of other factors is still debatable. These injuries should be managed aggressively with operative fixation at the earliest and one should not hesitate to open reduce, if anatomical reduction is not obtained by closed manipulation. Choice of implant and their configuration can vary depending on age and weight of patient, type of injury and surgeon’s preference. Use of three titanium screws in an inverted triangle configuration offers most stable fixation and one should avoid crossing the growth plate, unless stability of fixation is in question. Implants should be removed as early as possible after satisfactory union. Functional outcome is primarily affected by osteonecrosis, nonunion and severe coxa vara, whereas mild coxa vara, shortening, and premature physeal closure have no significant influence in short term. Finally, complications can occur regardless of the method of treatment and might not reveal their full impact until many years after injury. Hence, a guarded prognosis should be explained to care-givers at the time of injury. Despite technological advances of recent years, non-unions, avascular necrosis and their sequelae continue to be unresolved issues that need further study.

Submission declaration

This work has not been published previously and is not under consideration for publication elsewhere. Its publication is approved by both authors and by the responsible authorities where the work was carried out. If accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.

CRediT authorship contribution statement

Ritesh Arvind Pandey: Conceptualization, Methodology, Formal analysis, Writing - original draft. Bobby John: Conceptualization, Supervision, Writing - review & editing.

Declaration of competing interest

The authors have no known competing financial interests or personal relationships that could influence the work reported in this paper.