The FPI-6 was developed in response to a requirement for a quick, easy, and reliable method for measuring foot position in a variety of clinical settings, while minimizing the subjectivity of clinical evaluation methods [21]. It is a useful assessment tool and the only approach that captures information about standing foot postures in multiple foot segments using palpation and observation of deformity without a requirement for complex measurement techniques for clinicians [22]. Especially, each individual FPI-6 score is considered to be affinitive to radiologic findings due to its reliance on visualization or palpation of deformity. Hence, in this study, we presented that the assessment of foot posture using FPI-6 was reproducible in asymptomatic people and patients with foot and ankle symptom, and that FPI-6 scores were correlated with radiographic measurements, suggesting that FPI-6 measurement is an adequate alternative for clinical application of foot postures in the absence of radiographic examinations.
Based on the data presented in this study, the range of mean total FPI-6 score in each rater was 0.73 to 1.15 in asymptomatic people, which means nearly neutral and slightly pronated in foot posture. Consistent with our study, previous studies have confirmed this tendency towards normal feet as being pronated rather than completely neutral [23, 24]. Also employing a large sample indicates that in the normal sample, the mean FPI-6 score is 2.4 (SD = 2.3), confirming that a slightly pronated foot posture is the normal position at rest [22] The range of mean FPI-6 score in each rater was − 0.37 to 0.4 in patients with foot and ankle symptom. Since the patient group is composed of individuals with heterogenous disease related deformities, the FPI-6 range should not be interpreted to have significant clinical implications. Nevertheless, limiting value of each FPI-6 score in the patient group was broader than those of the healthy group as expected. Notably, in contrast to the patient group, some outliers were reported in the asymptomatic group, which demonstrated that the absence of symptoms does not always suggest normal posture of foot and ankle.
When comparing inter-rater reliability of FPI-6 in our study, the findings in the asymptomatic people group showed good repeatability (ICC; 0.608) as in line with the previous studies [25–27]. Whereas the result analyzed in the patient group demonstrated excellent inter-rater reliability (ICC; 0.878)
In hoc, there have been many studies of factors affecting the inter-rater discrepancy in asymptomatic people group and patient group, clinically. First, FPI-6 manual does not show example figures of intermediate scores. In addition, for item 4, the manual figure shows a foot part that is not exactly the talonavicular joint region, which may lower inter-rater reliability, especially in people without any deformity [3]. Also, one could hypothesize that the soft-tissue influence, which is indeed what the assessor is basing the FPI-6 score on, plays an integral and perhaps more important role than the underlying bony structure in normal population [28]. Another study explained the difficulty, as reported by examiners, of visualizing bony structures. The presence of calluses, bunions and edema is more common in elderly people [29]. Meanwhile, the group with neurogenic pes cavus (mean FPI logit score = -2.78, SD = 2.32) and idiopathic pes cavus (mean = -2.63, SD = 1.25) had FPI-6 scores significantly different from the normal population (mean logit score = + 2.4), indicating that the FPI-6 data was sensitive to disease-related postural changes [22]. Similarly, another article reported the high sensitivity of the FPI-6 to postural change associated with pathological pes planovalgus (median FPI raw score = + 12) [30].
There appears to be a scope for using FPI-6 scores and associated normative values to help identify groups with structural pathology and to assist in the clinical decision-making process. Hence, distinct characteristics of palpation as well as clear visual observation about morphological deformity can be demonstrated the higher inter-rater reliability in patient group.
Interestingly, inter-rater reliability was excellent in rater 1 and 2, who were most experienced in taking FPI measurements. Their knowledge of the musculoskeletal system, palpation skills and foot surface anatomy knowledge may have allowed them to discern the FPI-6 criteria with reproducible precision. Previous studies reported that even novice examiners who have a background in musculoskeletal assessment were able to produce reliable inter-examiner results using the FPI-6 with minimal training [27]. However, the current study demonstrated that inter-rater reliability was relatively lower in inexperienced examiner, showing that experience and training would be valuable for application in clinical situation.
Previously, there have been few studies about the correlation between FPI-6 and radiographic parameters. They reported that total FPI-6 demonstrated weaker correlations with the radiographic parameters (CPA, 0.36; calcaneal first metatarsal angle, 0.42) in older people [13]. In pediatric flatfoot, FPI-6 was correlated with the MA (r = 0.422, p = 0.008) and CPA (–0.411, p = 0.01)[31]. Inconsistent with previous studies, CPA was not correlated with total FPI-6 in our study. Instead, TNCA, TMA, lateral TCA, MA, and HAA were correlated with total FPI-6 with statistical significance. Especially, TNCA and HAA were more strongly correlated among them, having almost moderate correlation 0.665 and − 0.773 respectively.
More specifically, we compared individual FPI-6 scores with several radiographic parameters. This result was compatible with a previous study which reported that 4 individual components of the FPI score were poorly correlated with relevant radiographic parameters in young healthy subjects, with the exception of a moderate correlation between talar head palpation and the TNCA (0.42, p < 0.001) [28].
At first, we speculated that radiographic parameters which correspond to individual FPI-6 scores would exist, such as a relationship between congruence of the medial longitudinal arch and CPA. But, the bulging in talonavicular joint was most correlated with HAA (-0.748, p < 0.001) and the congruence of the medial longitudinal arch was correlated with TNCA (0.686, p < 0.001). Hence, although some radiographic parameters that were speculated to correspond with specific individual FPI-6 did show some correlation, the speculated match was not the most correlated as we expected.
When we compare the asymptomatic group and the patient group, the patient group exhibited higher correlation between FPI-6 and radiographic parameters. In general, in asymptomatic participants, distinct characteristics in radiographic parameters might not be present in contrary to the patients group. Furthermore, in the patients group, TNCA was highly correlated with talar head palpation, and HAA was highly correlated with inversion/eversion of calcaneus. This correlation demonstrates that some individual FPI-6 could be applied in midfoot and hindfoot assessment in clinical settings regarding foot and ankle deformities. When we specified each disease in 60 patients with foot and ankle symptom using descriptive statistics, not only the individual FPI-6 score but the total FPI-6 score were negative (supinated) in cavus foot, varus osteoarthritis and rheumatic arthritis, and the scores were positive (pronated) in flat foot, valgus osteoarthritis and hallux valgus patients
The primary limitation of this study is that gender and age were not controlled. Group A was composed only of men of young age, and Group B was relatively older, so there could be a limitation in the analysis of repeatability and correlation of this study. However, we think age was not an important issue because we focused on the repeatability of FPI-6 scoring system rather than actual scores of each case. Second, the distribution of disease was uneven, having high number of varus OA patient. However, the purpose of our study was not only to identify correlation between FPI-6 and radiologic measurements in asymptomatic volunteers and patients with radiographic deformity separately, but also to confirm correlation of total 100 participants overall, regardless of whether or not a disease is present. Third, whereas repeatability of FPI-6 was confirmed by each 4 raters, radiographic parameter was only analyzed by 1 orthopedic surgeon, whereby reliability could be poor. Therefore, further research should be undertaken to evaluate the effect of these potential confounders and to overcome each limitation described above.