A prospective consecutive, non-controlled, clinical follow-up study was performed. A cohort of 32 patients with symptomatic flexible flatfoot, after failure of conservative treatment, were treated with subtalar arthroereisis from February 2017 to December 2018 in the hospital of first author. The exclusion criteria comprised treatment of pes eqiunovarus, the presence of genetic or neurological disorders, previous foot injury or surgical intervention on the foot and inability to follow directions during gait examination.
Two patients were lost to follow up and discharged from the study, leaving 30 patients (41 feet) included in the final assessment. Mean age was 10 years (6 to 16 years). Assessment was performed before surgery and after a mean period of eight months (6 to 12 months).
The final follow-up was based on a simple patient foot health survey prepared for the study. The survey included the following: pain of operated foot during last week (no pain, mild, moderate, severe), pain at activities (no pain, at every day activities, long walking, sport activities, walking on uneven surface, walking on stairs) and well-being in context of the operated foot (very good, good, moderate, poor).
The initial assessment included a clinical examination comprising indications for surgery: flexibility of deformation, degree of heel valgus, presence of laxity, symptoms and Achilles tendon contracture.
Heel valgus was assessed by the patient standing backwards on a treadmill. A photo was taken with a camera placed on a tripod 40 cm above the ground. Following this, the angle of valgus of the heel before and after surgery was determined using a protractor in the treadmill gait analysis software (Freestep, Sensor Medica, Rome, Italy).
Tissue flaccidity was assessed using the Beighton scale. A Beighton score of 0–4 was assumed as indicating a norm, while values of 5–9 indicated tissue flaccidity [9].
Dorsoplanar and lateral radiographs were obtained during full weightbearing in a standing position. The radiographs were measured digitally with CGM DiagRAD software (CompuGroup Medical, Lublin, Poland). Radiographic measurements were obtained at lateral and dorsoplanar talo-1st metatarsal angle (TMT 1), calcaneal inclination angle, talar declination angle and longitudinal arch angle, as described by Davids et al. [10, 11].
Static and dynamic pedobarography was performed on a treadmill with diagnostic platform (Run time treadmill, Sensor Medica, Rome, Italy) and Freestep software (Sensor Medica, Rome, Italy).
During dynamic pedobarography, the patient was instructed to walk on the treadmill. Whole stance time, double support/swing phase, stance phase time, initial contact phase (ICP), forefoot contact phase (FFCP), flat foot phase (FFP), and foot contact area were measured. In addition, the percentage loading of the forefoot, hindfoot, medial and lateral parts of the foot was evaluated while in the flatfoot phase.
During static pedobarography, each subject was instructed to stand on a pressure sensing device with their body weight distributed equally over both feet. The feet were subjected to digital mapping during static examination; this included the lateral and medial forefoot, lateral and medial midfoot and lateral and medial hindfoot.
Footprints were recorded using a footprint digital scanner (Sensor Medica, Rome, Italy) using Freestep software (Sensor Medica, Rome, Italy). The flatness of the footprint was measured by measuring, Staheli’s arch index (SAI), Chippaux-Smirak index (Fig. 1) and Clark’s angle (CSI) (Fig. 2) [1, 12].
The B/C ratio was used to determine SAI: a value of 0.44–0.89 indicated a normal rate, < 0.44 as hollow foot and > 0.89 as flat foot [1, 12]. The A/B*100% ratio was used to calculate CSI; a value of 26–45% indicated a normal value, 46–49% flat foot I°, 50–75% flat foot II°, > 75% extreme flat foot and ≤ 25% hollow foot [1, 12].
Surgical technique. A skin incision measuring approximately 1–2 cm was made over the sinus tarsi. A blunt dissection was then made through the superficial tissues and further to the sinus tarsi to spread apart the interosseus ligament. A blunt guide wire was then placed in the sinus and canalis tarsi, so that the end of the wire was felt below medial malleolus. The correct size of the implant was determined using trial sizers; the position of the guide wire and sizers were checked with fluoroscopic imaging. Following this, the movements of the foot were examined, with the goal being to achieve up to 5° of pronation. Care was taken especially to avoid overcorrection. The exact size of the implant was established based on x-ray imaging and clinical examination, the implant was placed and the wound was closed. A titanium conical screw implant was used (Fig. 3, Fig. 4).
If the gastrocnemius muscle was found to be contracted, the muscle was receded using the Strayer technique (three patients). In such cases, a below-knee plaster cast was applied for four weeks after surgery.
The study was approved by Bioethical Committee of the institution of the first author (project number 2017/IV/62-MN). It was performed in line with the principles of the Declaration of Helsinki. Informed consent was obtained from the parents and/or legal guardians of the participants included in the study. If the patients were 13 years old or above, informed consent was obtained from both the participants and their parent and/or legal guardians, as required by Polish law.
Statistical analysis. The investigated values were described by weighted arithmetic means, standard deviation, 95% confidence interval and minimum-to-maximum values. All data was first tested for normality, skewness and kurtosis, and the equality of variances was confirmed using Levene’s test.
For normally-distributed variables, the differences in the values before and after surgery were analysed using multifactor analysis of variance (ANOVA) with repeated measurements. For non-normally distributed variables, generalized estimating equations (GEE) with repeated measurements and robust standard errors (i.e. sandwich estimators) were fitted. All the statistical models were controlled for sex. As the measurement unit was a single foot, and that some of the study participants underwent surgery on one foot and others on both feet, an intra-subject correction was applied. Any comparisons with P < 0.05 were deemed statistically significant. All the statistical computations were carried out using Stata/Special Edition, release 14.2 (StataCorp LP, College Station, Texas, USA). Patients lost to follow-up were not included in the statistical analysis.