Visual analogue scale foot and ankle (VAS-FA score) is a new score, validated in previous studies, but never compared to AOFAS score.
Analysis of the two scores using Indian language questionnaire.
Fifty patients with Malleolar fractures were assessed for functional outcome, time for calculation of scores, difficulty in correlation and comprehension of the questionnaire, in Malayalam language. The score parameters were compared by SSPSS.
There was similarity in pattern of score values in both systems but also a difference between values in each category, with VAS-FA having lower values, reflecting its efficacy. There was significant correlation, similar sensitivity and agreement between the scoring systems. VAS-FA correlated better with patient’s outcome and required less time for assessment.
This study shows that Indian language VAS-FA has a similar pattern of extracting scores as AOFAS and can be an efficient tool in ankle outcome assessment in Indian patients.
Ankle injuries after either conservative or surgical treatment are sometimes followed by complications like pain, stiffness, instability and secondary osteoarthritis . A wide variety of outcome measures have been proposed for use in conditions affecting the foot and ankle. Evaluation after ankle injuries can be subjective, objective or combined and in combination with radiological parameters . The American Orthopaedic Foot and Ankle Society’s (AOFAS score) is the most widely used system in foot and ankle but has some drawbacks in practical application, objective assessment  and inadequate validation have been pointed by various authors [4,5]. This study compares a new scoring system VAS-FA, by Dr. Martines Richter from The Hannover Medical School, to the AOFAS score, and check the efficacy of the score and to see its correlation and to elicit the drawbacks of each scoring system. previous studies have validated VAS-FA with SF-36 but never has the VAS-FA and the AOFAS been correlated [5,6]. We did this study in the Indian language of Malayalam spoken by around 38 million people . Translation was done using the Guillemin’s guidelines .
The inclusion criteria were: 1. The study subjects are patients with malleolar fractures who have been admitted in the Department of Orthopaedics, Pushpagiri Medical College Hospital between January 2007 and January 2010. 2. Only inpatients during this period were randomly selected for this study. 3. Skeletally mature patients were selected. The exclusion criterion was: 1. Patients with other fractures of the lower limb and pelvis. 2. Patients with other debilitating conditions of spine – fracture dislocation with neurological deficits or severe head injury. 3. Patients who were not ambulant prior to injury due to any medical and surgical condition. 4. Positive medical history concerning the entire lower extremity, diabetes mellitus, drug abuse, metabolic disorder, psychiatric diseases and rheumatoid arthritis.
- Patient selection and assessment
Fifty patients were randomly selected from the inclusion group. Old medical records were assessed; patients contacted and follow up functional status ascertained. Information regarding the nature of injury, classification, treatment protocol followed, old X-rays and history were verified from medical records. All ankle fractures were classified using Lauge-Hansen (LH) classification . After eliciting history, these patients assessed for their ankle function with the AOFAS and VAS-FA scores Malayalam version made using the Guillemin’s guidelines [6,8] using the technique followed by Angthong et al. .
The VAS-FA score was marked by the patient on a sheet of paper with 10 cm long straight lines, in which the right end of the line indicated maximum score (1 0 0), and left end of the line indicated minimum score (0). We used a 10-cm ruler for measurement that was positioned on the score form for valuation of each question unlike the template method used in the original article . The score was entered on the Microsoft excel calculation table and the pain, function and total score was elicited .
Time for calculation of scores in both systems was noted. Patients were also asked if they found any difficulties in correlation or comprehension while using the scores. They were also asked if time delay was a hindrance while using the systems. The number of questions unanswered was also noted. Measures like alignment and range of movement were observed and documented.The AOFAS being a subjective as well as objective score the objective parameters were done by the following method. Alignment was done by goniometric method using a caliper for finding the axis of the calcaneum. The mid axis of the leg was also drawn in a similar way and the angle between them was calculated. 0–58 valgus was considered normal, 5–108 as mild, >108 was taken as severe valgus deformity, while 0–58 varus was considered mild and 5–108 varus was severe. Foot was considered to be plantigrade when great toe, little toe and heel touched the ground. Hindfoot motion was calculated with the same method drawing the axis and comparing the range of motion to the contralateral side to decide on the degree of restriction. Sagittal motion was calculated with a goniometer and the range compared to the contralateral side to estimate the restriction. Gait was considered to be normal when patient walked without a lurch and the step length was equal, it was considered as an obvious gait abnormality when the there was a lurch of the shoulder but the step length was equal and was as a marked abnormality when lurch was associated with an unequal step length. In this study the demographic data were looked into, along with percentage of different mechanism of injury. The results were analyzed and correlated to the scores. The time spent for evaluating the different scores was recorded and compared. The correlation of pain and function scores were also looked into.
- Statistical analysis
Data were analyzed using statistical package for social sciences (SPSS). Data were expressed in its frequency and percentage as well as mean and standard deviation. For the associations and comparisons between different parameters, Chi square (x2) test was used as nonparametric test. Student’s t test was used to compare the mean values between two groups. Multivariate Pearson correlation analysis was employed to find out relationship between various parameters. Receiver operated characteristic (ROC) analysis was carried out to conclude upon the cut-off value and the ROC curve was drawn. To compare different groups each other, non-parametric Mann Whitney’s U test was employed. Cohen’s kappa was estimated for tables that have the same categories to estimate the agreement between variables. For all statistical evaluations, a two-tailed probability of value, <0.05 was considered significant. 3. Observations and results The distribution of data in different age groups, the side of fracture, the type of fracture (open/closed), classification according to LH system, mechanism of injury and the occupa-tional distribution are noted in Tables 1–3. Varied distribution was noted with maximum number of cases in 40 – 49 age group at 24% and least incidence in less than 20 years (Fig. 1). According to LH classification 46% fractures were supination external rotation type, followed by pronation external rotation at 34%, and supination adduction and pronation abduction at 10% each. There were no pronation dorsiflexion injuries (Fig. 2). According to stage of injury maximum number of cases were noted in stage 4 at 26% followed by stage 5–22%, stage 1–20%, stage 2–18%, least was in stage 3–14% In different mechanism under LH classification (Fig. 3). 4. Comparison of scoring systems 4.1. Pain scores Pain score was divided into 3 groups and number of patients in each group noted with their scores according to AOFAS and VAS-FA systems. Maximum distribution of patients were noted in the (higher group >80%) followed by the (second 60–80%) group and then by the (third 60%) in both systems (Fig. 4).
Function score was divided into 4 groups and number of patients in each group noted with their scores according to
Comparison of mean total score values in mild (Stages 1 and 2) with severe stages (3–5).
Total score Stage Mean SD t value P value
AOFAS Mild stage 92.37 10.83 1.809 >0.05
Severe stage 84.90 15.83
VAS-FA Mild stage 93.13 9.55 4.087 <0.001 Severe stage 75.89 16.74
Comparison of mean function score values in closed and open fractures. Function score Type Mean SD t value P value AOFAS Closed 88.97 17.87 2.481 <0.05 Open 69.56 33.46 VAS-FA Closed 79.12 21.31 1.551 >0.05
Open 66.41 26.21
Comparison of mean total score values in closed and open fractures.
Total score Compound Mean SD t value P value
AOFAS Closed 89.60 12.22 2.231 <0.05 Open 78.11 20.41, VAS-FA Closed 83.57 15.76 1.334 >0.05 Open 75.47 19.57
AOFAS and VAS-FA systems. Most number of patients were noted to be in the (higher group 80%) followed by (60–79), then by (40–59) and least no of patients were noted in (20–39), in both scoring systems (Fig. 5).
Distribution of patients according to total score groups
The total scores were also divided into 3 groups and patients in each group were noted. There was maximum number of cases in (>80%) category followed by (60–79%) category and the least number of cases were noted in (40–59%) (Fig. 6).
Mann Whitney U test
Score parameters of pain, function and the total scores in both systems were compared. It was observed that, the mean VAS-FA score was persistently lower in category of pain, function with a significant P value <0.05. There was also a more significant difference in total score with VAS-FA system. Having a highly significant P value <0.01 (Fig. 7). 4.5. Student t-test 4.5.1. Comparison of mean and score values in mild (stages 1 and 2) with severe stage (3, 4 and 5) Comparison was done for total scores. And there was a significantly lower total score in severe stage compared to milder stage in both systems P < 0.01 VAS-FA and P < 0.05 AOFAS (Table 1). 4.5.2. Comparison of mean function score values in closed and open fractures Student t-test was used to compare mean scores in open and closed fractures. There was a significantly lower score in compound compared to closed in both systems though VAS-FA being less significant (Tables 2 and 3). 4.6. Multivariate correlation analysis (MCA) MCA for two score systems was done. The correlation between pain – 0.692, function – 0.788 and total – 0.801 scores in VAS-FA and AOFAS systems were very highly significant. *P < 0.05; **P < 0.001 (Table 4). 4.7. Measure of sensitivity and specificity using ROC curve Measure of sensitivity and specificity using ROC curve was done. Cut-off value for AOFAS under ROC curve was 72.0 with a sensitivity of 0.88 and specificity of 0.44. For VAS-FA the cut-off value was 64.45, sensitivity 0.88 and specificity was 0.22 (Figs. 8 and 9, Table 5). 4.8. Cohen’ kappa Measure of agreement between two scoring systems using Cohen’ kappa was done and there was very highly significant agreement between the pain, function and total scores in AOFAS and VAS-FA systems (Table 6). 4.9. TIme for assessment Average time AOFAS system requires is less, 6 min, for assessment when compared to VAS-FA system’s 9 min.
There have been several scoring systems used for evaluation in foot and ankle [10–15]. Some evaluation schemes incorporate subjective and objective clinical variables into a numerical scale, such as the forefoot score for bunionette, forefoot score for bunion, ankle scores, heel score, tarsometatarsal or resection arthroplasty score, or ankle inversion injury scale, or into a graded nonnumerical scale, such as excellent, good, fair, and poor [10–22]. Some scores also have radiological and clinical parameters. VAS-FA is the adequate technique for data acquisition regarding objectivity and reliability in numerous studies [23,24]. Considering the limitations of existing scoring systems it was hypothesized by various authors that a new scoring system was required like Halasi et al. who used Tegner score for evaluating ankle-related activity changes [3–5,25].
The AOFAS score
It is the most popular score [2,3,5,26,27] but has some disadvantages as it is not validated, it cannot be obtained if answers are missing, and contains problematic pseudo-objective assessment [3–5] Richter et al. also defines difficulties because of non- or pseudo-objective assessment pattern of the AOFAS-score .
The VAS-FA score
Martinus Richter et al. constructed and validated a new score taking into consideration the flaws of existing scores, the VAS-FA score with the following features was constructed: questionnaire based on 20 subjective questions, visual-analogue-scale (VAS-FA) based rating, computerized evaluation . The three different question categories pain, function, and other complaints compa-rable to the AOFAS score were included. These categories were weighed differently, because function was considered to be more important for the clinical outcome than pain alone or other complaints. More questions were included for function (n = 11) than for pain (n = 4) and other complaints (n = 5). Therefore, function is weighed higher for the final score than pain or other complaints. All questions, all categories and the entire score were designed to have the same range of possible points (0–100) for easy and distinct evaluation. Moreover, a high ‘‘stability’’ of the score against missing answers was expected. This makes it possible to obtain the entire score result and the result of the different score categories if questions are unanswered . It was validated with SF-36 and Hannover questionnaire with good correlation and further the time needed for evaluating the scores was significantly lower for VAS-FA than for SF-36 and Q .
There have been previous studies comparing ankle scoring systems in search of a better and comprehensive system. Nilsson et al. evaluated the functional outcome in ankle fractures after surgery using Olerud and Molander ankle score (OMAS), SF 36, linear analogue scale and self rated ankle function and found that only SF-36 scores indicated that females had functional status below the age- and gender matched normative data of the Swedish population . Nilsson et al. in another study on effects of a training program after surgically treated ankle fracture in a prospective randomized controlled trial using OMAS and SF-36 and found that there was a correlation between these two scores in three out of nine outcome measures . Van Der Wees et al. studied the validity and responsiveness of the ankle function score (AFS) after acute ankle injury, AFS was compared with OMAS, patient-specific complaints (PSC), and global perceived effect. This study found limited evidence for the AFS as a prognostic and evaluative instrument . The ankle osteoarthritis score (AOS) is another instrument that was developed for the assessment of pain in patients with ankle osteoarthritis [31,32]. The score demon-strated a high ‘‘vulnerability’’ regarding other musculoskeletal problems than at the ankle resulting in a questionable specificity for foot and ankle disorders . The foot function index (FFI), was validated at the 21st Annual Summer Meeting, Boston, MA, USA, 15–17 July, 2005 and was correlated with the SF-36 for validation Budiman-Mak et al, Kuyvenhoven et al. [33–35]. Kurup et al. evaluated the function with AOFAS and visual analogue scale for pain following total ankle replacements and concluded that there was a correlation between the scores .
The most recent study with respect to VAS-FA score was by Angthong et al, they translated the original English version of VAS-FA into the Thai version and evaluated the validity and reliability of Thai VAS-FA in patients with foot and ankle-related problems and correlated with SF 36 score . They found similar outcome as the study by Richter et al. .
Our study is unique in comparing the AOFAS and the VAS-FA scoring systems for the first time. A similar study is not available for comparison. Furthermore, in spite of the limitations, we have tried to correlate the various comparable parameters like pain and function. By testing the systems in a new language we have tried to assess the reproducibility and sensitivity. The drawback of this study is that AOFAS is a combined subjective and objective score and the VAS-FA is purely subjective score. Furthermore, only the pain and function parameters of the AOFAS could be compared
since the remaining variable of alignment was not present in VAS-FA.
Alignment and objective parameters were checked but could not be compared since such parameter was not present in VAS-FA. Pain and function were comparable entities along with total score. Comparing pain and function categories gave specific correlation in theses assessment categories along with total score.
Interpretations of results
- In comparison of AOFAS with VAS-FA
The distribution of patients in the three Pain score groups was similar and this suggested that there was a similarity in distribution of patients in both scores. Similarly, distribution of patients in the four Function score groups indicated similarity between the two scores. The total scores were divided into three groups and patient distribution in each was similar.
- Mann Whitney U test