Arthroereisis in juvenile flexible flatfoot: Which device should we implant? A systematic review of literature published in the last 5 years

Table 1 Study results

Ref.	Patients (mean age)	AR type	Assessment	Results	Complications	Limits
Ruiz-Picazo et al[8], 2019	16 (32 FFF), 9 yr (range: 7-11 yr).	Subtalar device. Titanium self-locking device.	OxAFQ-C pre- and postoperatively.	Postoperative results were positive, with statistical significance for the “school and play”, “emotional”, and “footwear” domains of the OxAFQ-C scale (P < 0.05) and no differences in the “physical” domain.	4 (25%). 2 extrusion of the implant and 2 overcorrections.	Retrospective nature, no control group, sample size. lack of objectively measurable parameters.
Pavone et al[16], 2019	105 (174 FFF) (12.6 ± 1.3 yr), mean follow-up 67.5 ± 16.4 mo.	Calcaneo-stop.	BMI-for-age AOFAS, FADI, FADI Sport, and SF-36 at 1 and 3 yr. Variation of the angles measured on X-ray images.	AOFAS, FADI, FADI Sport, and SF-36 scores and radiological assessments improved significantly postoperatively (P = 0.001) and remained stable over time (1 yr: P < 0.001; 5 yr: P < 0.001) in all groups of patients. An increased persistence of pain and/or discomfort in obese patients than in normal weight patients (P = 0.02).	20 (11.5%) patients. 1 screw loosening (0.6%); 14 transient pain and discomfort at the level of the surgical incision (8%), 2 contractures of the lateral peroneal muscles (1.15%); 3 superficial infections (1.7%).	Retrospective nature, no control group, sample size. lack of objectively measurable parameters.
Pavone et al[1], 2018	68 (136 FFF) 12.7 yr (9-15 yr), mean follow-up 57.6 mo.	Calcaneo-stop.	AOFAS, Yoo Score, FADI, FADI Sport, OxAFQ-C scores at 1 and 3 yr. variation of the angles measured on X-ray images.	AOFAS (F 2, 201 = 287.51; P < 0.0001), Yoo Score (F 2, 201= 2627.00; P < 0.0001), OxAFQ-C (F2201 = 210.60; P < 0.0001), FADI (F 2, 201 = 372.62; P < 0.0001) and FADI Sport (F 2, 201= 189.32; P < 0.0001). Radiological assessments improved (P < 0.0001).	17 (12.5%); 5 pain at surgical scar (3.7%), 4 local symptoms at the incision (2.9%), 3 screw loosening (1.2%) and 4 superficial Infections (2.9%); 1 screw breakage (0.73).	Retrospective study. No control group. Mid-term follow-up. Lack of objectively measurable parameters.
Memeo et al[26], 2019	202 FFF, 13.6 yr (8-16 yr). Median follow-up was 130 mo (35-150 mo).	Calcaneo-stop.	Clinical evaluation, pain, and variation of the angles measured on X-ray images.	92% Percutaneous lengthening of the Achilles tendon. Improvement of clinical outcome, restoration of the medial arch and the hind foot valgus in orthostasis.	32 (15.8%): 23 (11.4%) incomplete correction; 9 (4.5%) screw breakages.	Retrospective study. No control group. Short-term follow-up. Lack of objectively measurable parameters.
Memeo et al[26], 2019	200 FFF, 12.8 yr (8 to 16 yr). Median follow-up 130 mo (35-150 mo).	Subtalar AR bioabsorbable device.	Clinical evaluation, pain, and variation of the angles measured on X-ray images.	71% Percutaneous lengthening of the Achilles tendon. Improvement of clinical outcome, restoration of the medial arch and the hind foot valgus in orthostasis.	25 (12.5%): 20 (10%) inflammatory process involving soft tissues around tarsal sinus; 5 (2.5%) device. removals and substitutes.	Retrospective study. No control group. Mid-term follow-up. Lack of objectively measurable parameters.
Megremis et al[18], 2019	14 (28 FFF), 10.71 ± 1.58 yr (range 8-14 yr). Mean follow-up duration of 35.14 ± 9.82 mo (19-60 mo).	Subtalar AR.	AOFAS pre- and postoperatively.	The mean postoperative AOFAS score was 88.851 (range 83-97) points (P < 0.0001).	No complication.	Retrospective study. No control group. Short-term follow-up. Lack of objectively measurable parameters.
Martinelli et al[14], 2018	49 (98 FFF) 10.7 yr (7-14 yr), mean follow-up 4.9 yr.	Subtalar AR.	CHQCF; OxAFQ pre- and postoperatively. Number of sessions per week.	The mean OxAFQ scores within the ‘Emotional’ (P < 0.05) and ‘footwear’ item (P < 0.05) (children), and in the ‘school and play’ (P < 0.05) and ‘footwear’ item for the parent scale. 44 (89%) parents were satisfied with the surgical procedure.	Three residual pain; 1 residual deformity, and one sport limitation. 3 subtalar implants removed because of pain at the sinus tarsi.	Retrospective nature; as some of the anthropological measures were not obtained before surgery.
Kubo et al[20], 2020	Group A (5-8 yr): 6 (11 FFF) MA 7.4 ± 1.2 yr; Group B (9-12 yr): 33 (63 FFF) MA: 11.2 ± 1.0 yr; Group C (13-15 yr): 11 (21 FFF). MA 13.6 ± 0.7 yr.	Calcaneo-stop.	CP; lat. TCA; a.p. TCA, kite angle; NCI; Meary angle. Bony maturation.	Improvement of radiological assessment in each group (P < 0.05). Best deformity correction when surgery was conducted between 9 and 12 yr, with significant improvement in all measured parameters without secondary deterioration during FU.	No complications.	Retrospective study; no clinical scores.
Indino et al[28], 2020	56 (112 FFF), 9-14 yr (MA 15.5 ± 1.2 yr), mean follow-up 40.1 ± 23.6 mo.	Subtalar AR.	AOFAS, SEFAS, SF-12.	AOFAS 97.3 ± 4.5, SEFAS 47.2 ± 1.5, SF-12.MCS 51.1 ± 8.8, SF-12.PCS 55.6 ± 9.1, Clinical scores were not correlated with the foot radiographic parameters at follow-up period.	No complications.	No preoperative clinical or functional scores.
Hsieh et al[15], 2019	102 (204 FFF) MA 9.1 ± 0.2 yr).	Subtalar AR.	CP; lat. TCA; a.p. TCA, kite angle; NCI; Meary angle.	22 cases of bilateral extrusion in the overweight group (39%) vs 13 cases bilateral extrusion in the low body weight group (23%) (P = 0.0004). The inter-observer correlation was 0.95. Improvement of radiological assessment at 3 mo postoperative follow-up (P < 0.05).	19%.	Retrospective nature, no control group, sample size. lack of objectively measurable parameters. No BMI loss evaluation surgical outcomes.
Hagen et al[32], 2019	7 (13 FFF), MA 12.43 ± 1.27 yr.	Calcaneo-stop.	Pedobarographic measurements before surgery, 3, 14, 28 d after.	The ground force increased significantly in lateral foot areas (P < 0.001) and decreased in medial areas (P < 0.001).	1 minor soreness.	Sample size, short follow-up.
Hagen et al[31], 2020	14 (27 FFF), MA 12.4 ± 1.4 yr.	Calcaneo-stop.	Heel angle, rearfoot angle, leg axis angle, step length, and walking speed, pre- and postoperatively.	Heel valgus (F 1, 24 = 110.465, P < 0.001); Dynamic heel angles vs static heel angles (F 1, 24 = 38.498, P < 0.001). Correlation between heel angle and rearfoot angle (static: r = 0.647, P < 0.001; dynamic: r = 0.640, P = 0.001). Dynamic rearfoot angle vs static rearfoot angle (F 1, 24 = 166.55, P < 0.001).	No complications.	Small size. No clinical assessment.
Giannini et al[21], 2017	44 (88 FFF). MA 11.7 yr (8-14 yr).	Bioabsorbable Calcaneo-stop.	Patient satisfaction; Meary angle; talocalcaneal angle.	33 excellent, 9 good outcome, and 2 poor clinical. Meary’s, talocalcaneal angle had improved (P < 0.001).	2 breakages.	Retrospective nature, no control group, sample size. lack of objectively measurable parameters.
Faldini et al[24], 2018	173 (283 FFF), MA 11.2 yr. MF 49.5 mo.	Bioabsorbable Calcaneo-stop.	FFI, SEFAS.	FFI score 4; SEFAS score 47.2, well into the normal range. No statistically significant differences between males and females (P value > 0.05).	3 implant breakages, 1 persistent pain secondary to a local inflammatory response.	No radiological assessment. No control group.
Caravaggi et al[33], 2018	13 (26 FFF), MA 11.3 ± 1.6 yr, MF 1 yr (12.5 ± 3.7 mo).	Endo-orthotic implant and Calcaneo-stop.	A 10-point VAS, Kinematic and kinetic analysis during normal walking, radiological parameters.	All radiological parameters and VAS were significantly improved at 1-yr follow-up (P < 0.001). No significant differences were detected in spatiotemporal and ground-reaction-force parameters between pre-op evaluation and control in either implant groups.	No complications.	Multisegment foot protocol. Small size. No clinical assessment.
Bernasconi et al[27], 2020	31 (62 FFF), MA 10.5 ± 1.6 yr, MF 62 ± 15 mo.	Subtalar AR.	ROM; AOFAS; VAS-FA, radiological parameters.	Improvement clinical score and radiological parameters except talonavicular coverage angle (P = 0.49) and calcaneo-fifth metatarsal angle (P = 0.53). Improvement of dorso-plantar view. No loss of correction was found after removal of the implant.	17 (24%) sinus tarsi syndromes.	Retrospective nature, no control group, sample size. lack of objectively measurable parameters.
Elmarghany et al[38], 2020	42 (84 FFF); MA 9.92 ± 2.2 yr; range (7-15 yr). MF 29.1 mo.	Calcaneo-stop.	AOFAS; radiological parameters.	Improvement clinical score and radiological parameters (P value < 0.000).	3 minor complications, 1 (0.02%) under correction. 1 sunken screw. 1 synovits around screw.	Retrospective nature, No control group, sample size. lack of objectively measurable parameters.

AOFAS: American Orthopedic Foot and Ankle Society; AR: Arthroereisis; FADI: Foot and Ankle Disability Index; FFF: Flexible flatfoot; FFI: Foot Function Index Questionnaire; MA: Mean age; MF: Mean follow-up; OxAFQ-C: Oxford Ankle Foot Questionnaire for Children; ROM: Range of motion; SEFAS: Self-reported Foot and Ankle Score; VAS: Visual Analogue score.

Table 2 Risk of bias of the included studies

Ref.	Dutch checklist form for prognosis
	No participant selection took place	Groups are comparable regarding age	Validated measuring system used	Independent (blind) determination of outcomes	Clear description of groups available
Ruiz-Picazo et al[8], 2019	+	-	+	?	+
Pavone et al[16], 2019	+	+	+	-	+
Pavone et al[1], 2018	+	+	+	-	+
Memeo et al[26], 2019	+	+	+	?	+
Megremis et al[18], 2019	+	-	+	-	+
Martinelli et al[14], 2018	+	?	+	?	+
Kubo et al[20], 2020	+	+	+	?	+
Indino et al[28], 2020	+	-	+	?	+
Hsieh et al[15], 2019	+	+	+	-	+
Hagen et al[32], 2019	+	+	+	?	+
Hagen et al[31], 2020	+	+	+	?	+
Giannini et al[21], 2017	+	+	+	?	+
Faldini et al[24], 2018	+	+	+	-	+
Caravaggi et al[33], 2018	+	+	+	-	+
Bernasconi et al[27], 2020	+	+	+	?	+
Elmarghany et al[38], 2020	+	+	+	-	+

+: Low risk; -: High risk; ?: Unclear.