Effectiveness of Hallux Valgus Surgery on Improving Health-related Quality of Life: A Follow Up Study

44 (SD: 7.09) to 12.35 (SD: 4.85). SF-12 (physical) changed from 36.26 (SD: 5.32) to 47.06 (SD: 4.82), SF-12 (mental) from 38.23 (SD: 8.04) to 46.49 (SD: 4.16), and EQ5-D from 0.64 (SD: 0.008) to 0.90 (SD: 0.10).


Introduction
Hallux valgus (HV) is a frequent pathological disorder of the forefoot 1 that commonly produces painful disability.HV is used to describe the structural deformity of the rst metatarsophalangeal joint (MTPJ), 2,3 clinically determined when the HV angle is greater than 15º and the 1-2 intermetatarsal angle (IMA) is greater than 9º. 4 Although HV has a clear genetic component, being predominant amongst women, 5 it is a progressive deformity, especially in those aged over 50 years.The prevalence of HV increases from 23% in adults aged 18 to 65 years to 35.7% in people aged over 65 years. 6,7 a result of both pain and disability, HV deformity negatively impacts the health-related quality of life (HRQoL). 8,9The greater the HV angle and rst nger pain, the worse the scores in the physical, psychological, and social aspects of HRQoL. 10,11It has been determined 12 that HV deformity is coupled with signi cant foot-speci c pain and disability, even when global physical functioning and participation in activities are not adversely affected.Over time, as severity of HV deformity increases, progressive muscle weakness occurs around the rst MTPJ and both general and foot-speci c HRQoL scoring progressively decrease. 10,13merous surgical procedures have been developed for the treatment of HV. 14 However, no consensus exists on the optimal surgical technique to correct this deformity. 15,16Thus, the surgical choice should be taken after considering the range of causal factors. 17,18,19Open procedures are the most commonly used surgical techniques, including chevron's distal osteotomy and scarf's diaphyseal in the rst metatarsal. 20,21,22In the last years, the interest in minimally invasive surgery (MIS) has been growing due to its theoretical advantages including lower morbidity and less time required for recovery and rehabilitation. 23wever, the evidence supporting percutaneous techniques is still inconclusive 24,25 per randomized control trials with designed treatment protocols and the use of validated tools for measurements.
Most studies [26][27][28] have reported improvements on the correction of the radiological angles (HVA and 1-2 IMA) as well as in HRQoL assessed by the American Orthopedic Foot and Ankle Society (AOFAS) scale.Although the AOFAS score cannot be considered a true patient-reported outcome measures (PROM) 29 , it is usually the most used tool in clinical studies of the foot and ankle. 30,31Systematic reviews 32 have shown that HV surgery is a more effective procedure than conservative or no treatment in reducing pain 33 .However, high-quality studies comparing similar types of HV treatments and rates of complications or unfavorable outcomes re lacking. 34us, we aimed to evaluate the effect of HV surgery on PROMs (i.e., pain scales, general HRQoL, and region-speci c scales) and radiological angles.Additionally, we aimed to determine whether the effect on these outcomes depends on the type of surgery (including open and percutaneous techniques) and if it is in uenced by potential confounding factors (i.e., age, HVA, 1-2 IMA, body mass index (BMI), and distal metatarsal articular angle (DMAA).

Study Design
This was a longitudinal prospective study conducted between March 2017 and May 2018.The study was designed and conducted in accordance with the Declaration of Helsinki and the Guidelines on Good Clinical Practice, 35 and was approved by the Clinical Research Ethics Committee of the Cuenca Health Area.Written and oral informed consent was obtained from all participants.
We collected the clinical data of all patients who underwent surgery for symptomatic HV deformity in the orthopedic department of the Virgen de la LuH Hospital of Cuenca (Spain).All surgical techniques were performed by orthopedists and included open surgery (chevron osteotomy 36 and scarf osteotomy 37 ) and percutaneous surgery with Reverdin-Isham osteotomy. 38e inclusion criteria for patients were (1) older than 18 years with painful HV; (2) complete conservative treatment (including oral medication, shoe modi cation, and physical therapy for three months); (3) mild-to-moderate HV deformity 39 , a preoperative HVA < 40º and 1-2 IMA < 20º; (4) primary HV surgery, and (5) signed written informed consent.We excluded patients with severe hallux rigidus (radiological suggestion of severe degenerative arthritis) of MTPJ or stiffness, severe instability of the rst tarsometatarsal joint, previous failed HV surgery, infection, peripheral vascular disease, peripheral neuropathy or other systemic diseases affecting the musculoskeletal system, and patients with a follow-up of <12 months.
A total of 81 eligible patients were initially included in our study, but 9 patients were lost to follow-up.The type of osteotomy was decided according to the HVA, the 1-2 IMA, and the surgeon's opinion.Chevron osteotomy was conducted in 16 patients (16 feet), scarf osteotomy was performed in 24 patients (24 feet) ,and percutaneous surgery was completed in 32 patients (32 feet).Akin osteotomy 40 was performed in 22 patients (22 feet) (30%) of the total.

Clinical outcomes
The clinical outcomes were assessed using the AOFAS Hallux metatarsophalangeal-interphalangeal (HMI) scale 41 , visual analogue scale (VAS) 42 , Manchester-Oxford Foot Questionnaire (MOXFQ) 43 , short form health survey (SF-12) 44 , and European Quality of Life-5 Dimensions (EQ-5D). 45These outcomes were evaluated before surgery and at the end of the followup; additionally, the post-operative complications were recorded at end of follow-up.
The AOFAS-HMI scale (range, 0-100 best score) recording system combines subjective and objective data to assess the following clinical factors: pain (40 points), function (45 points), and alignment (15 points). 46The VAS consists of a straight line with the endpoints scoring 0 (no pain at all) to 10 (the worst pain imaginable).The MOXFQ is a patient-based questionnaire for assessing HRQoL in patients receiving foot and/or ankle surgery.It consists of a 16-item summative Likert scale scored from 0 to 4 (most severe) in three dimensions: pain, walking/standing, and social interaction.The total score ranges from 0 to 64.The MOXFQ index is calculated as the percentage of the maximum score dividing the total score by the maximum possible score (i.e., 64) and multiplied by 100. 47e SF-12 questionnaire is a 12-item survey including physical and mental health components (PCS-12 and MCS-12). 48EQ-5D 49 includes 5 questions about the patients' mobility, self-care, usual activities, pain/discomfort, and anxiety/depression.The total score ranges from 1 (without problems) to 5 (unable to perform/extreme problems).

Radiological assessment
The radiological examination of HVA, 1-2 IMA and DMAA, was completed by the same independent observer and according to the same radiographic protocol.Weight-bearing dorsoplantar and lateral images were obtained preoperatively, immediately after the surgery, and at the end of the follow-up. 50The HVA is measured using the anteroposterior view as the angle between the line connecting the center of the rst metatarsal base and the metatarsal head with the line connecting the center of the proximal and distal articular surfaces of the proximal phalanx.The 1-2 IMA is measured as the angle between the former line and the line bisecting the diaphyseal portion of the second metatarsal.Finally, the DMAA is de ned as the angle between the perpendicular line and the longitudinal axis of the rst metatarsal. 51,52rgery procedures The surgical procedures were completed using local or block anesthesia in the supine position and under tourniquet control.During surgery, the foot was kept externally rotated with its lateral edge on the operating table .The distal metatarsal chevron osteotomy 36 was completed through a medial incision starting from the proximal phalanx medially, continuing proximally beside the medial margin of the metatarsal; a longitudinal midline capsulotomy was performed in the identical plane as the incision.The apex of the osteotomy was positioned 1-2 mm superior to the center of the metatarsal head and the angle of the chevron was 60° to 90°.Fixation was achieved by two cannulated screws.Scarf osteotomy 37 was performed with a skin lateral incision of 3-4 cm and a longitudinal capsular incision.The longitudinal osteotomy was oblique from the longitudinal axis of the rst metatarsal and the transverse limbs ought to be parallel to each other, perpendicular to the second metatarsal, and approximately 60º to 80º positioned to the longitudinal osteotomy line.Fixation was achieved with a two headless compression screws.
The adductor hallucis tendon section and a lateral soft tissue release were completed in all cases for both techniques and an intraoperative observation for DMAA was performed.The hallux was then reduced to a neutral position and congruency of the MTPJ was con rmed under uoroscopy.
Otherwise, percutaneous surgery was performed through a 3-5 mm skin incision on the medial and plantar side of the rst metatarsal head.The exostosectomy was removed from this incision with a conical burr.Reverdin-Isham osteotomy 23,38 was performed using a Shannon straight burr.Finally, we completed a lateral soft tissue release of the MTPJ and a tenotomy of the adductor hallucis tendon through a 1-2 mm skin incision.All steps were veri ed under uoroscopy.Osteosynthesis was not used in any case of percutaneous surgery.The Akin osteotomy 40 of the proximal phalanx was chosen for the existence of interphalangeus.The Akin osteotomy was performed without osteosynthesis.

Statistical Analysis
The sample size was estimated considering a pre-post-surgery difference of 2.5% (a error 0.05; statistical power of 0.80) in AOFAS scale mean score, obtaining a minimal estimated sample size of 43 patients.To have su cient statistical power for examining other outcomes and estimating a 15% dropout rate, the minimum sample size was estimated on 72 patients.
Both statistical (Kolmogorov-Smirnov test) and graphical methods (normal probability plots) were used to assess the normal distribution for each continuous variable.In the descriptive statistics section, categorical variables are presented as sample (n) and percentage (%), and continuous variables are presented as mean and standard deviation (SD).Repeated measures ANCOVA models were used to test the mean differences of the pre-post-surgery clinical and radiological outcomes, adjusted by baseline values of clinical outcomes (i.e., VAS, MOXFQ, SF-12, and EQ-5D) in Model 1, and adjusted by baseline values, age, BMI, HVA, 1-2 intermetatarsal angle, and distal metatarsal articular in Model 2. We estimated the crude change and the effect size (ES) for each outcome using Cohen's d index. 53The ES was categorized as 0.2 (weak effect), 0.5 (moderate effect), 0.8 (strong effect), and >1.0 (very strong effect).Additionally, Student's t-test and ANOVA were used to test mean differences in clinical outcomes at the end of follow-up in the categories of age, HVA ( 30 and ≥30), 1-2 IMA ( 13 and ≥13), BMI (normal weight, overweight, and obesity), and DMAA.Statistical analyses were performed using SPSS-IBM (V.24.0 SPSS Inc., Armonk, NY, USA), and statistical signi cance was set at p ≤ 0.05.

Characteristics of the case series
The nal number of patients included in the study was 72 (72 feet), mean age 59.67 years (SD: 11.39) and including 70 women (97.2%).Each patient was followed for at least 12 months with a total mean follow-up period of 17.68 months (SD: 3.11 months).Demographic baseline characteristics are summarized in Table 1.
Clinical and radiological outcomes.
Non-signi cant differences were found in clinical outcomes at the end of the follow-up by type of surgery except for the PCS-12 between scarf osteotomy and percutaneous surgery (p = 0.006).Regarding the angles, greater improvements in the 1-2 IMA were found for chevron osteotomy versus scarf osteotomy, chevron osteotomy versus percutaneous surgery, and scarf osteotomy and percutaneous (p = 0.001).In addition, better improvements in DMAA were found for chevron versus scarf osteotomy, chevron osteotomy versus percutaneous surgery, and scarf osteotomy versus percutaneous (p = 0.001; Supplementary Material Table S1).Finally, after controlling for baseline clinical outcomes (model 1), as well as for baseline clinical outcomes, age, BMI, HVA, 1-2 IMA, and DMAA, the differences in clinical and radiological outcomes were not substantially modi ed (Tables 3 and 4).
Overall, no differences were found when we analyzed the mean of the clinical outcomes at the end of follow-up by categories of HVA, 1-2 IMA, age and BMI.(Supplementary material table S2)

Complications
At the end of follow-up, a total of seven complications occurred in six patients (six feet) (8.4%).There were two patients suffering transfer metatarsalgia, three patients with recurrent asymptomatic HV, one patient had rst MTPJ stiffness, and another patient had complex regional pain syndrome type 1.One patient suffered metatarsalgia and recurrent asymptomatic HV simultaneously.No patient suffered head necrosis or troughing, 54 failure of xation, delayed wound healing, or Varus deformity.

Discussion
This follow up study was aimed to examine the effect of HV surgery on PROMs (i.e., pain scales, general HRQoL, and regionspeci c scales) and radiological angles, and to determine whether the observed effects could be modi ed by the type of surgery (including open and percutaneous techniques) and by confounding factors.Our data indicated that HV surgery improves patients' PROMs and radiological angles regardless the type of surgery.These effects persist after controlling for potential confounders (i.e., age, HVA, 1-2 IMA, BMI, and DMAA).
We found pre-post improvements in the VAS, AOFAS (total score and speci c domains), MOXFQ (total score and speci c domains), PCS and MCS of SF-12, and EQ-5D scores, similar to ndings reported in previous studies. 55,56,57Although most studies only reported data from the AOFAS and VAS scales, the use of PROMs 58 is encouraged to assess HV surgery outcomes. 59Their inclusion could provide valuable information for the evaluation of patients' outcomes, 33 and could play an important role in determining the impact of HV surgery on HRQoL.
The correction of radiological angles is related to improvement in PROMs, 29 although we observed a correction in the angles and poor scores in the PROMs, especially in the pain scoring.From this perspective, our data showed improvements after surgery in the three radiological angles (i.e., HVA, 1-2 IMA, and DMAA) similar to those previously reported in other studies. 60,61en we analyzed whether the observed effects on clinical outcomes could be modi ed by the type of surgery and confounding factors, we did not nd substantial differences in the results after adjusting for different sets of confounders.The differences found in the 1-2 IMA and DMAA angles between types of surgery may be due to the surgeon's decision based on the initial angles.Additionally, the role of BMI in HV surgery outcomes has been previously studied, 62,63 and BMI did not act as a predictor for functional outcomes; thus, patients with obesity should not be excluded from HV surgery. 64e overall complication rate in our study was 8.4%.Although these gures are slightly higher than those reported by other studies, 65,66,67 our complication rate is acceptable, considering the surgical variability and the lack of appearance of serious complications.Additionally, recurrence is a feared complication after HV surgery, with rates ranging from 3% to 73%, 61,63,68,69 being 4.2% in our study.HV recurrence has been correlated with preoperative HVA, 1-2 IMA, and DMAA. 70,71,72Preoperative HVA has been considered the most important predictor of outcomes after HV surgery, with 1-2 IMA and DMAA playing secondary roles in postoperative outcomes.We found an acceptable recurrence rate for the three surgical techniques and found no differences between them. 73ere are some limitations to this study that should be acknowledged.First, the follow-up period was relatively short for evaluating long-term HV recurrence.Second, the variability in the surgical techniques and the involvement of different surgeons might have affected the results, although it provides some advantages such as it might better re ect actual practice than when reporting data from a single surgeon or technique.Third, although AOFAS is the most used instrument, this tool has some limitations in assessing the impact of HV surgery on HRQoL in terms of validity, reliability, and responsiveness, which may impact the statistical power of our effect estimates.Finally, numerous factors may have in uenced clinician-and patientrelated outcomes, such as type of surgery, surgeon skill, other associated forefoot diseases (lesser toe deformities, metatarsalgia), and comorbidities.

Table 2 .
Mean differences of the preoperative and end of follow-up clinical and radiological outcomes.valuesfor all comparisons between preoperative and end of follow-up were statistical signi cance p ≤ 0.001.*Thesize of the effect was categorized as 0.2 (considered a weak effect), 0.5 (considered a moderate effect), 0.8 (considered a strong effect), and >1.0 (considered a very strong effect) byCohen.
Results are shown as mean and (±) Standard deviation.For categorical variables the values are expressed in percentages.Abbreviation: BMI, body mass index.aHigher scores indicate a better health-related quality of life.bLower scores indicate a better health-related quality of life.cLower scores indicate a better radiological outcome.P

Table 3 .
Pre-post mean differences on the clinical outcomes by types of surgery, controlling for baseline outcome (model 1) and baseline outcome, age, body mass index, hallux valgus angle, 1-2 intermetatarsal angle and distal metatarsal articular angle(model 2).Results are shown as mean and (±) Standard deviation.Bold values indicate statistical signi cance p ≤ 0.05 Abbreviations: VAS,visual analogue scale; AOFAS,American Orthopaedic Foot and Ankle Society; MOXFQ, Manchester-Oxford Foot Questionnaire; SF-12,Short Form-12 Health Survey; PCS, Physical Component Score; MCS, Mental Component, Score; EQ-5D, EuroQol-5D; ES, effect size.Higher scores indicate a better health-related quality of life b Lower scores indicate a better health-related quality of life.P values for all comparisons between preoperative and end of follow-up were statistical signi cance p ≤ 0.001.* The size of the effect was categorized as 0.2 (considered a weak effect), 0.5 (considered a moderate effect), 0.8 (considered a strong effect), and >1.0 (considered a very strong effect) by Cohen.Model 1: adjusted by baseline outcome; Model 2: adjusted by baseline outcome, age, body mass index, hallux valgus angle, 1-2 intermetatarsal angle and distal metatarsal articular angle. a

Table 4 .
Pre-post mean differences on the radiological outcomes by types of surgery, controlling for baseline outcome (model 1) and baseline outcome, age, body mass index, hallux valgus angle, 1-2 intermetatarsal angle and distal metatarsal articular angle (model 2).