The effects of hallux valgus and hallux rigidus surgery on gait biomechanics: a systematic review

During the preparation of the article, the PRISMA guidelines were followed. The review was limited to articles analyzing gait biomechanics in patients with hallux valgus (HV) or hallux rigidus (HR) deformity. The electronic search of databases was performed on 19th March 2020. The articles were limited to those published from January 1999 to March 2020. The Science Direct, PubMed Central, PubMed and EBSCO databases were searched to identify appropriate literature using the following key: gait analysis AND (hallux limitus OR hallux rigidus OR hallux valgus) AND surgery.

Only full-text articles were selected from the electronic databases. Titles, abstracts and full texts of the retrieved documents were sequentially reviewed by K.L and M.B to determine their relevance to the topic. Also, the reference lists of all the studies included for review were searched manually for additional studies of relevance. The inclusion criteria were: 1) full text of the publication in English; 2) assessing gait kinetic or kinematic parameters and other parameters relating to gait; 3) analyzed gait after HV or HR surgery. Exclusion criteria included: 1) number of participants below 5; 2) arthroplasty surgical methods; 3) participants with other significant health problems than HR or HV; 4) no corrective surgery of HV or HR among participants, 5) conference abstracts and review papers were excluded, 6) articles in which only comparison of surgical methods was made were excluded.

The duplicate articles from different databases were rejected. The title and abstract of the selected articles were first screened according to the eligibility criteria. Furthermore, full-text evaluation was performed if the title and abstract could not provide adequate information for the article screening process. The rejected articles were re-screened to avoid misinterpretation. The titles, abstracts and full text of the papers identified by the search were subsequently screened by two independent reviewers (the authors: K.L. and P.K.) to choose those that met the selection criteria and to extract data. Decisions about which trials should be selected were made by negotiation. The reference lists of all the studies included for review were searched manually for additional studies of relevance by one author (K.L.). One reviewer (K.L.) compiled all the articles using a reference manager software (Mendeley Ltd., USA). Next, the selected and accepted articles were divided into three subgroupsevaluating:1) kinematic, 2) kinetic parameters, 3) plantar pressure, and 4) scales.

Data extraction was performed by two reviewers independently (K.L. and P.K.) and then agreed. Type of intervention; purpose of the study; study group characteristics, test conditions, equipment; results were extracted.

Methodological quality of the trials selected for this review was then evaluated using a checklist for both randomized and non-randomized studies [13]. The checklist consisted of 27 items divided into five sub-scales: (1) Reporting (10 items) – which assessed whether the information provided in the paper was sufficient to allow the reader to make an unbiased assessment of the findings of the study; (2) External validity (3 items) – which addressed the extent to which the findings from the study could be generalized to the population that the study subjects were derived from; (3) Internal validity – bias (7 items) – which addressed bias in the measurement of the intervention and the outcome; (4) Internal validity – confounding (6 items) – which addressed bias in the selection of study subjects; (5) Power (1 item) – which attempted to verify whether the negative findings from a study could be due to chance. The assessment of the first three criteria has been used in this paper, therefore questions concerning only these sections are presented below. Reporting the checklist comprised the following questions: 1. Is the hypothesis/aim/objective of the study clearly described?; 2. Should the main outcomes be measured and clearly described in the Introduction or Methods section?; 3. Are the characteristics of the patients included in the study clearly described?; 4. Are the interventions of interest clearly described?; 5. Are the descriptions of principal confounders in each group of subjects clear?; 6. Are the main findings of the study clearly described?; 7. Does the study provide estimates of random variability in the data for the main outcomes?; 8. Have all the important adverse events that may result from the intervention been reported?; 9. Have the characteristics of patients lost to follow-up been described?; 10. Have actual probability values for the main outcomes (e.g. 0.035 rather than <0.05), except where the probability value is less than 0.00, been reported? External validity checklist included;11. Were the subjects asked to participate in the study representative of the entire population from which they were recruited?; 12. Were those subjects who were prepared to participate representative of the entire population from which they were recruited?; 13. Were the staff, places, and facilities where the patients were treated, representative of the treatment the majority of patients receive? Internal validity – bias checklist included; 14. Was an attempt made to blind study subjects to the intervention they had received ?; 15. Was an attempt made to blind those measuring the main outcomes of the intervention?; 16. If any of the results of the study were based on “data dredging”, was that made clear?; 17. In trials and cohort studies, do the analyses adjust for different lengths of patients’ follow-up, or in case-control studies, is the time period between the intervention and the outcome the same for cases and controls?; 18. Were the statistical tests used to assess the main outcomes appropriate?; 19. Was compliance with the intervention/s reliable?; 20. Were the main outcome measures used accurate (valid and reliable)?

We have resigned from using Internal validity – confounding and Power assessment sections due to a large number of questions not applicable to the articles included in our review. The answers were scored 0 or 1, except item number 5 which scored 0 to 2. Questions 1–10 were scored 1 if the answer was “yes” or 0 if the answer was “no”. Questions 11–20 were scored 1 if the answer was “yes” or 0 if the answer was “no” or “unable to determine”. The maximum number of points in each section was as follows: Reporting – 10, External validity – 3 and Internal validity – bias – 7. Therefore, the total maximum score was 21 points.

All the papers considered for analysis were rated (11–15/21 points)using Downs and Black Quality Assessment Tool [13] (Table 1). The lower rating was mainly due to the External validity section, where most of the papers (9 papers) had 0 or 1 point (6 papers). Only one article [14] was rated at the maximum number of points.

In the Reporting section, all the papers were rated 6–8/11 points. The scores were affected by negative answers to questions 5 (Are the descriptions of principal confounders in each group of subjects clear?) and 8 (Have all the important adverse events that may result from the intervention been reported?) in all of the cases. Only one article [18]was rated the lowest 6/11 points. In that case, a negative answer to question 4 (Are the interventions of interest clearly described?) had an impact. In the Internal validity – bias section all the papers were rated 4-5/7 points. The scores were mainly affected by negative answers to questions 14 (Was an attempt made to blind study subjects to the intervention they had received?) and 15 (Was an attempt made to blind those measuring the main outcomes of the intervention?) in all of the cases. Summarizing, Overall Quality Index Score ranged from 48% to 71%. Quality assessment revealed a problem with external validity in most of the papers, e.g. the participant groups were not representative for the population and some aspects of internal validity.

The subsections that follow contain analysis of kinematic, kinetic and plantar pressure parameters as well as different patient assessment tools such as AOFAS score, which are referred to separately as scales for HV and HR. There are no articles concerning kinetic parameters for HV because none of them met the inclusion criteria.

Kinematics parameters during walking after HV surgery were described in five papers [15,16,17,19] (Tab. 2).

In one of the analyzed studies [15], the type of operation was not specified. In two works: [17,19] – Lapidus procedure was performed, and in two papers [16,18] first ray osteotomy was carried out. In two papers [15,18] motion capture systems were used. On average, 14 patients were analyzed in the above studies, with the maximum number of 19. Most of the spatio-temporal parameters remained the same as before the surgery, except for the push-off duration and toe-off pitch angle in one study [17]. The temporo-spatial parameters were significantly worse than in the control groups [17]. After Scarf surgery the only one reduced parameter was the gait speed [16].

Kinematic parameters during walking after HR surgery were described in seven papers ( Tab. 3). In three of those articles plantar pressure parameters were characterized additionally, in three authors assessed kinetic parameters, and in all of them scales were used to evaluate patient’s progress.

Cheilectomy was performed in three articles [22,23,24] and MTP I arthrodesis also in three [20,21,26]. In five papers motion analysis systems were used [20,22,23,24,26]. In papers where cheilectomy was done, the average number of subjects was 15.66, whereas in papers where arthrodesis was used it was 13. Cheilectomy and MTP I arthrodesis do not influence knee or hip kinematics [20]. The influence of MTP I arthrodesis on ankle joint kinematics is unclear [21,26]. Cheilectomy in most cases increases MTP I motion in the sagittal plane [22,23].

Kinetic parameters during walking after HR surgery were described in 3 papers (Tab. 4). In one of those articles plantar pressure parameters were characterized additionally, in three authors assessed kinematic parameters (Tab. 2) and in all of them scales were used. Both arthrodesis and MTP I cheilectomy favorably influence ankle push-off power [21,24].

Plantar pressure parameters after HV surgeries were assessed in six articles (Tab. 5). Percutaneous distal soft tissue release (DSTR)-Akin procedures were carried out in two of them, in two others first ray osteotomies were performed, and in another two modified Lapidus procedure was used.

In four articles [14,17,19,27] there were insole pressure systems used and in one it was platform measurement systems [28,29]. The average number of patients in plantar pressure studies was 30.66, with the minimum of 10 and the maximum of 79. In osteotomy studies, plantar pressure under hallux and force time integral significantly increased [28,29]. Changes of plantar pressure parameters in the hallux region are inconclusive and depend on the study.

Plantar pressure parameters after HR surgeries were assessed only in three articles ( Tab. 6). Defrino et al. and Nawoczenski et al. [20,22] compared pre – and postoperative results, whereas Stevens et al. [26] compared post-operative results with the control group.

For HV in seven papers different scales and angles were discussed ( Tab. 7). HVA and IMA were evaluated in six articles [14,15,17,19,27,29], AOFAS scale was used also in six articles [14,17,19,27,28,29]. HVA and IMA significantly decreased after surgeries in all the articles that they were measured in. AOFAS results were also significantly better than the pre-operative ones. For HR AOFAS scale was used in four papers ( Tab. 7.) [20,23,24] and the results increased post-operatively. HVA and IMA also significantly decreased in two articles [20,26]

For HV AOFAS scale was used in six out of seven papers and in all of them the results significantly increased [14,17,19,27,28,29]. HVA and IMA were also assessed in six articles and were significantly smaller postoperatively [14,15,17,19,27,29].

Kinematic analysis after HV surgeries was performed at different time intervals depending on the research. Klugarova et al. [18] tested patients 4 months after the surgery, whereas Canseco et al. [15] had their mean time of follow up 16 months, Moerenhout et al. [19] 12 months, Chopra et al. [17] 6 months. The conducted analysis showed that spatio-temporal parameters were worse [15,17,18] or the same as before the surgery [15,17,19]. Moreover, Klugarova et al. [18] proved that spatio-temporal parameters of the non-operated limb had also been affected. Lack of significant improvement in spatio-temporal parameters may suggest that the healing process may not have finished by the time of testing or the gait pattern had not been properly restored in the process of physiotherapy. The number of physiotherapy sessions e.g. 9 [17,19] may not be enough. It seems that surgery does not influence knee and hip biomechanics even when spatio-temporal parameters are disturbed.

Due to a small number of papers comparing preand post-operative patients who underwent arthrodesis, it is hard to draw consistent conclusions. It seems that some spatio-temporal parameters may be altered [21]. Increased time of single limb support and smaller step width may suggest that arthrodesis improves gait quality in patients with severe MTP I arthrosis and does not affect hip or knee kinematics [20,21,26]. The cheilectomy performed on patients with mild and moderate HR appears to be a valid procedure in case of gait kinematics. It significantly increases MTP I active range of motion in the sagittal plane [22,23,25]. The range of motion at MTP I in the coronal and transverse planes improves [22] or stays the same [23]. There are, however, differences in reports for spatio-temporal parameters: Canseco et al. [23] report significant improvements, Smith et al. [24] report no differences between pre – and post-operative analysis and Kuni et al. [25] report that the level walking speed did not change, but only eight patients were tested in that research.

Ankle power during the push-off phase in a limb with fused MTP I is significantly lower than in a healthy limb [20], but compared to the pre-operative values it significantly increased [21]. Similar conclusions comparing pre – and post-cheilectomy patients were drawn by Smith et al. [24]. In case of arthrodesis and cheilectomy, it is possible that a decreased pain level allowed patients to walk freely without pain in the push-off phase connected with MTP I dorsiflexion. Pain usually occurs while osteophytes on the dorsal surface of the first metatarsal and proximal phalanx compress each other or irritate cutaneous nerves.

Patients with HV usually shift load to the lateral side of the foot and unweight hallux [30]. Therefore, increased mean contact area, maximal force and force time integral under hallux [28,29] suggest that first ray osteotomies positively affect gait biomechanics. Restoration of proper gait pattern is among the objectives of the rehabilitation process after HV and HR surgeries. One of its elements is to impose weight on the MTH I during the terminal stance phase, but it may sometimes be hard to achieve due to the characteristics of osteotomies. Osteotomies causes elevation of the MTH I and shortening of the first ray, which automatically causes unweighting of the MTH I during gait and increases amount of weight transferred by the MTH II. Decreased mean pressure under hallux after DSTR surgery may be explained by the fact that in HV lateral deviation is increased by the hallux abductor muscle which starts acting like a plantar flexor. Therefore, correction of the deformity should decrease pressure under the hallux. There are two different outcomes, increased and decreased parameters under the hallux after surgery, but then which outcome is really beneficial to the patient? To assess if change of the plantar pressure parameter is beneficial to the patient, we suggest that the results should always be compared with the control group. The closer the results to the control group, the better the treatment outcome is.

Plantar pressure analysis performed among patients with HR who were subjected to MTP I arthrodesis shows that this type of surgery reestablishes weightbearing function of the first ray and hallux [20]. However, it does not reestablish normal gait pattern [26]. According to Nawoczenski et al. [22] cheilectomy does not induce any significant plantar pressure changes.

All HV surgical procedures included in this study significantly decreased HVA [14,15,17,19,27,29] and IMA [14,15,17,19,27,29]. What is more, AFOAS [14,17,19,27,29] score was significantly better after surgeries. HVA or IMA are less often measured in patients after HR surgical treatment. The main reason is that in HR deformity HVA and IMA are usually not significantly greater than in healthy people. In this group of patients AOFAS scores are significantly better postoperatively than pre-operatively [20,23,24,25,26].

One limitation of this study is the fact that most articles had a low number of participants. The other weaknesses of this study are incoherent age of participants in analyzed articles and different post-operative evaluation periods. Another drawback is comparison of studies where different procedures were used to treat patients with the same deformity, although we believe it is justified by a small number of articles on the topic and our awareness of the differences between procedures.