Abstract
Background
Objective of the study was to evaluate and compare psychometric properties such as validity, reliability, floor ceiling effects and interpretability of the modified Harris Hip Score (mHHS) and the Forgotten Joint Score (FJS) in patients undergoing bipolar hemiarthroplasty for femoral neck fracture in Indian patients, because this has not been done before.
Methods
This observational study consisted of 40 consecutively operated patients. Construct validity and reliability were evaluated using correlation coefficient and Intraclass correlation coefficient (ICC), respectively. Interpretability was evaluated by describing mean and standard deviation of mHHS and FJS in five subgroups of patients based on their response to the global rating questions and assessment.
Results
The mean follow-up duration was 15.7 months ± 10.4. There was very high correlation between mHHS and FJS (r = 0.92, p < 0.0001) suggesting convergent construct validity. The results of correlation coefficient were 100% and 66.7% in accordance with the pre-formulated hypotheses for mHHS and FJS, respectively. mHHS and FJS demonstrated adequate construct validity and inadequate construct validity, respectively. The ICC value for mHHS and FJS was 0.80 (p = 0.005) {adequate reliability} and 0.34 (p = 0.06) {inadequate reliability}, respectively. Both mHHS and FJS-12 demonstrated acceptable level of floor (0% for mHHS and 14.3% for FJS) and ceiling effects (12.5% for both mHHS and FJS). There was significant difference in the mHHS and FJS in the five subgroups of patients suggesting adequate interpretability.
Conclusion
We recommend the use of the modified Harris Hip Score over the Forgotten Joint Score for functional outcome evaluation of Indian patients from rural setting undergoing bipolar hemiarthroplasty for femoral neck fracture.
Graphic Abstract
Keywords: Patient-reported outcome measures, Surveys and questionnaires, Hip prosthesis, Arthroplasty, replacement, hip, Femoral neck fractures, Construct validity, Interpretability, Modified Harris Hip Score, Forgotten Joint Score
Introduction
Neck of femur fracture is a common injury in the geriatric population and bipolar hemiarthroplasty as a treatment modality has good long-term outcome [1, 2]. Patient-reported outcome instruments (PROMs) are preferred for assessment of functional outcome after a surgical intervention. As validity and reliability are context-specific terms, it is crucial that the patient-reported outcome instruments have been proven to be valid and reliable in the condition and population of interest [3]. An outcome instrument that is valid and reliable for assessment of functional outcome of elective total hip replacement might not be valid and reliable for assessment of functional outcome after arthroplasty for neck of femur because of differing pathology (degenerative hip condition versus trauma) and different age group (relative young population for elective total hip replacement (THR) versus elderly patients for fragility fracture neck of femur) [4]. The modified Harris Hip Score (mHHS) and the Forgotten Joint Score (FJS) [5, 6] are commonly used patient-reported functional outcome instruments after THR for degenerative hip conditions. The mHHS has also been used to evaluate functional outcome after hemiarthroplasty and THR for fracture neck of femur [7–9].
Most of the patient-reported outcome instruments have been developed and validated in the Western population and there is paucity of validation of those PROMs especially in hip fractures in the Indian population. Moreover, very few patient-reported outcome instruments have been modified to suit the personal, social and cultural needs of the patient from the Indian subcontinent. Sitting cross legged on the floor for eating and religious activities and squatting for the use of eastern toilets are essential activities for patients staying in rural parts of India [10–12].
There is paucity of study reporting psychometric properties such as validity, reliability, interpretability and floor ceiling effects of the mHHS and FJS for functional assessment after bipolar hemiarthroplasty for fracture neck of femur. Also, the FJS has never been validated in patients undergoing hemiarthroplasty of the hip. The objective of the present study was to evaluate and compare the validity, reliability and floor ceiling effects of the mHHS and FJS.
Methods
Institutional Ethics Committee Approval, Inclusion and Exclusion Criteria
This combined retrospective and prospective observational study was conducted in a University teaching hospital after obtaining approval from the Institutional ethics committee approval. Patients aged more than 50 years presenting from January 2017 to July 2019 with fracture neck of femur and treated with bipolar hemiarthroplasty of the hip were considered eligible for inclusion. Patients that underwent bipolar hemiarthroplasty for intertrochanteric hip fractures, bipolar hemiarthroplasty for pathological neck of femur fracture due to metastasis, monopolar endoprosthesis for fracture neck of the femur and patients refusing to give written informed consent for participation in the study were excluded. Written informed consent was obtained from all willing participants in the study.
Sample Size Calculation
Epi Info version 7.2 software developed by Centers for Disease Control and Prevention (CDC) Atlanta, USA ([https://www.cdc.gov/epiinfo/pc.html] was used for the sample size calculation. During the year 2016, 1348 cases were operated in the orthopaedics department out of which 25 cases were bipolar hemiarthroplasty for fracture neck of femur. Accordingly, the prevalence rate for bipolar arthroplasty at our institute was 1.86% and assuming 5% as the acceptable margin of error) it was estimated that a sample size of 34 patients would have 97% power of study.
Data Collection
Data were collected pertaining to age, gender, mechanism of injury, side of injury, injury–surgery interval (in days), days of hospitalization, American Society of Anaesthesiologists (ASA) grading, classification of the neck of femur fracture (garden classification and anatomic classification), type of anaesthesia, surgical approach and the fixation modality of the bipolar endoprosthesis (cemented or uncemented).
On third and fifth postoperative days, the seven patients in whom the data was collected prospectively completed the mHHS and FJS. At final follow-up, patients completed the mHHS, FJS and answered three questions pertaining to global rating of pain, satisfaction and change in their clinical condition. Assessments pertaining to functional ability to sit with legs crossed and ability for squatting were done by clinicians on final follow-up.
Modified Harris Hip Score (mHHS)
The mHHS was developed by making modifications in the pain domain and substituting two items in the functional ability domain (ability to wear socks and shoes replaced by ability to squat; ability to sit in chair replaced by ability to sit with legs crossed) [Table 1]. This modification was done because patients from the rural background in the India usually do not wear shoes and socks. The mHHS is scored from 0 (worst functional outcome) to 100 points (best functional outcome). The values of mHHS were interpreted as follows: score < 70 points (poor result), 70–79 points (fair result), 80–89 points (good result) and 90–100 (excellent result) [13].
Table 1.
The contents of the modified Harris hip score (mHHS)
| Domains and items | Points |
|---|---|
| Pain | |
| None or ignores it | 44 |
| Slight, occasional, no compromise in activities | 40 |
| Mild pain, no effect on average activities, rarely moderate pain with unusual activity, may use diclofenac | 30 |
| Moderate pain, tolerable but makes concessions to pain, some limitation of ordinary activity and work: may require occasional pain medicine stronger than diclofenac | 20 |
| Marked pain, serious limitation of activities | 10 |
| Totally disabled, crippled, bedridden | 0 |
| Function: gait | |
| Limp | |
| None | 11 |
| Slight | 8 |
| Moderate | 5 |
| Severe or not able to walk | 0 |
| Support | |
| None | 11 |
| Cane for long walks only | 7 |
| Cane full time | 5 |
| One crutch | 3 |
| Two canes | 2 |
| Two crutches or walker or not able to walk | 0 |
| Distance walked | |
| Unlimited | 11 |
| Six blocks (600 yards) | 8 |
| Two to three blocks (200–300 yards) | 5 |
| Indoors only | 2 |
| Bed to chair transfer only | 0 |
| Functional activities | |
| Stairs | |
| Normally without using a rail | 4 |
| Normally using a rail | 2 |
| In any manner | 1 |
| Unable | 0 |
| Squatting | |
| With ease | 4 |
| With difficulty | 2 |
| Unable | 0 |
| Sitting cross legged | |
| With ease | 5 |
| With difficulty | 3 |
| Unable | 0 |
| Public transportation | |
| Able to use | 1 |
| Unable to use | 0 |
Forgotten Joint Score (FJS)
The FJS is a 12-item questionnaire and is scored from 0 (maximum awareness of artificial joint; worst outcome) to 100 points (least awareness of artificial joint; best and preferred outcome) [5]. The FJS has demonstrated adequate validity and responsiveness for assessment of total hip arthroplasty for non-traumatic hip pathology [14–16].
Patient-Reported Global Change and Clinician Performed Assessment
Three questions were pertaining to patient-reported global change [Table 2]. Two questions (one-item pain scale and one-item satisfaction scale) were selected based on the recommendations of the working group of International Society of Arthroplasty Registries (ISAR) patient-reported outcome measures [17]. Patient’s perception on change in their clinical condition was evaluated using a single-item clinical condition change questionnaire consisting of five-point Likert scale response based on the previous work by Copay et al. [18].
Table 2.
The three global rating scales from the patient’s perspective
| One-item pain scale |
| During the past 4 weeks, how would you describe the pain you have in your [right/left] operated hip? |
| Severe (1 point) |
| Moderate (2 points) |
| Mild (3 points) |
| Very mild (4 points) |
| None (5 points) |
| One-item satisfaction scale |
| How satisfied are you with your [right / left] hip replacement so far? |
| Very dissatisfied (1 point) |
| Dissatisfied (2 points) |
| Neutral [neither dissatisfied nor satisfied] (3 points) |
| Satisfied (4 points) |
| Very satisfied (5 points) |
| One-item question on change in clinical condition |
| How your clinical condition is presently, compared to your clinical condition after you injured your hip? |
| Very much worse (1 point) |
| Slightly worse (2 points) |
| Neutral [same as before] (3 points) |
| Slightly better (4 points) |
| Very much better (5 points) |
Two assessments [Table 3] pertaining to functional ability were performed by the clinician. One assessment was pertaining to the ability to sit with the legs crossed and the second assessment was pertaining to the ability to squat.
Table 3.
The two functional ability assessments undertaken by the clinician
| Ability to sit cross legged (Oriental sitting position) |
| Unable to sit with legs crossed (1 point) |
| Able to sit with legs crossed but with discomfort (2 points) |
| Able to sit with legs crossed comfortably (3 points) |
| Ability to squat |
| Unable to squat (1 point) |
| Able to squat but with discomfort (2 points) |
| Able to squat comfortably (3 points) |
Construct Validity
Construct validity refers to the relationship that a patient-reported outcome instrument has with other outcome instruments/parameters that either measure similar construct or measure dissimilar construct or measure non-related construct [19, 20]. For adequate construct validity, at least 75% of the results must be in agreement with the hypotheses [21]. Correlation between normally distributed numerical variables (mHHS and FJS) was evaluated using the Pearson correlation coefficient [r]. Correlation between a numerical and categorical variable was determined using the Spearman correlation coefficient [rs]. Outcome instruments measuring similar constructs should have correlation coefficient value ≥ 0.50 [21]. The value of the correlation coefficient 0.70–0.89 and > 0.90 is indicative of high correlation and very high correlation, respectively [22].
The hypotheses for construct validity were as follows: (1) there will be positive significant correlation between mHHS and FJS [convergent construct validity because both outcome instruments measure similar functional outcomes] (2) Patient-reported outcomes (mHHS and FJS) will have positive significant correlation with the response of the patient on the one-item pain scale, one-item satisfaction scale, one-item clinical condition change scale, ability to sit with legs crossed and ability to squat.
Test Retest Reliability
Test retest reliability refers to the stability of values of the outcome instrument in the absence of change in the clinical condition of the patient. Patients were evaluated on third and fifth day postoperatively based on the assumption that no significant difference would occur in the clinical condition of the patients between these 2 days. Test retest reliability was estimated using Intraclass correlation coefficient (ICC) [19] for two-way random effects and absolute agreement. It is recommended that the ICC value should be ≥ 0.70 [19, 21]. The ICC values can be interpreted as follows: ≥ 0.90 (excellent reliability), ≥ 0.70 to < 0.90 (good reliability), ≥ 0.50 to < 0.70 (moderate reliability) and values < 0.50 suggests poor reliability [23].
Floor and Ceiling Effects
Floor effect is defined as patients achieving the least possible score of the outcome instrument, whereas ceiling effect is defined as patients achieving the highest possible score of the outcome instrument. The proportion of patients scoring the highest and the least possible score of mHHS and FJS was estimated at the third postoperative day and at final follow-up. It is recommended that the floor and ceiling effect of should be seen in ≤ 15% of patients [19].
Interpretability
Interpretability is defined as the ability to attribute qualitative meaning to quantitative data provided by raw scores of patient-reported outcome instruments [21]. Interpretability was evaluated by comparing mean and the standard deviation (SD) of mHHS and FJS in subgroups based on patient’s response to global rating scale [19] and subgroups based on clinician assessment of functional abilities. It is recommended that the mean and SD be described in at least four subgroups of patients from the same cohort [19].
Statistical Analysis
Numerical data were checked for normality of distribution using the Kolmogorov–Smirnov test, skewness coefficient and Kurtosis coefficient. As both mHHS and FJS demonstrated symmetrical data distribution, parametric tests were used for comparison of mean values of the outcome instruments. For comparing values of mHHS and FJS in various subgroups of patients, ANOVA test and Bonferroni correction test for multiple repeated comparisons were used. The level of significance was set at 5%. Statistical analysis was performed using IBM SPSS version 19 (IBM, Armonk, NY).
Results
Seventy one consecutive patients underwent bipolar hemiarthroplasty for fracture neck of femur at our Institute from 1st January 2017 to 31st July 2019. Forty six patients were eligible for inclusion in the study, because 19 patients died and 6 patients refused to give consent for participation in the study. Out of these 46 patients, 6 patients were lost to follow-up and 40 patients attended final follow-up. Data were collected prospectively in 7 patients (17.5%) and retrospectively in 33 patients (82.5%). The mean follow-up duration was 15.7 months ± 10.4 (range 3–42 months). Demographic features and operative details of the cohort are presented in Table 4.
Table 4.
The demographic features and operative details of the cohort
| Demographic features | Rest expressed as frequency (percentage) |
|---|---|
| Agea [in years] (n = 40) | 68.8 ± 8.3 (50–85) |
| Gender (n = 40) | |
| Male | 19 (47.5%) |
| Female | 21 (52.5%) |
| Side injured (n = 40) | |
| Right | 11 (27.5%) |
| Left | 29 (72.5%) |
| Mechanism of injury (n = 40) | |
| Domestic fall | 35 (87.5%) |
| Fall from bicycle | 2 (5.0%) |
| Road traffic injury | 3 (7.5%) |
| Injury surgery intervalb [days] (n = 40) | 6 (2–10) |
| Days of hospitalizationa (n = 40) | 15 ± 5.8 (6–29) |
| ASA grading (n = 40) | |
| ASA-2 | 3 (7.5%) |
| ASA-3 | 30 (75.0%) |
| ASA-4 | 7 (17.5%) |
| Garden classification (n = 40) | |
| Type 2 | 3 (7.5%) |
| Type 3 | 25 (62.5%) |
| Type 4 | 12 (30.0%) |
| Anatomic classification (n = 40) | |
| Subcapital | 18 (45.0%) |
| Transcervical | 15 (37.5%) |
| Basicervical | 7 (17.5%) |
| Type of anaesthesia (n = 40) | |
| Spinal anaesthesia | 32 (80.0%) |
| Spinal + epidural anaesthesia | 3 (7.5%) |
| General anaesthesia | 4 (10.0%) |
| Spinal + general anaesthesia | 1 (2.5%) |
| Surgical approach (n = 40) | |
| Posterior approach | 37 (92.5%) |
| Lateral approach | 3 (7.5%) |
| Type of prosthesis (n = 40) | |
| Cemented bipolar endoprosthesis | 32 (80.0%) |
| Uncemented bipolar endoprosthesis | 8 (20.0%) |
aExpressed as mean ± standard deviation (range)
bExpressed as median [interquartile range]
Clinical and Functional Outcome
At the final follow-up, the mean mHHS was 80.9 ± 14.4 (range 48.4–100.1) and the mean FJS was 75.2 ± 19.6 (range 25–100). The clinical and functional outcomes are presented in Table 5.
Table 5.
The clinical and functional outcome of the cohort
| Clinical and functional outcome parameters | Frequency (percentage) |
|---|---|
| Outcome grading based on mHHS (n = 40) | |
| Excellent | 12 (30.0%) |
| Good | 11 (27.5%) |
| Fair | 7 (17.5%) |
| Poor | 10 (25.0%) |
| One-item pain scale (n = 40) | |
| No pain | 24 (60.0%) |
| Very mild pain | 8 (20.0%) |
| Mild pain | 5 (12.5%) |
| Moderate pain | 3 (7.5%) |
| One-item satisfaction scale (n = 40) | |
| Very satisfied | 13 (32.5%) |
| Satisfied | 22 (55.0%) |
| Neutral | 2 (5.0%) |
| Unsatisfied | 3 (7.5%) |
| One-item clinical condition change scale (n = 40) | |
| Very much better | 18 (45.0%) |
| Slightly better | 19 (47.5%) |
| Neutral | 3 (7.5%) |
| Ability to sit with legs crossed (n = 40) | |
| Able to sit comfortably | 15 (37.5%) |
| Able to sit with discomfort | 18 (45.0%) |
| Unable to sit | 7 (17.5%) |
| Ability to squat (n = 40) | |
| Able to squat comfortably | 12 (30.0%) |
| Able to squat with discomfort | 9 (22.5%) |
| Unable to squat | 19 (47.5%) |
Construct Validity
There was a statistically significant, positive and high correlation between mHHS and FJS (r = 0.88, p = 0.009) that was evaluated on the third postoperative day. There was a statistically significant, positive and very high correlation between mHHS and FJS (r = 0.92, p < 0.0001) that was evaluated on final follow-up [Fig. 1]. This is indicative of convergent construct validity between both the patient-reported outcome instruments. Correlation of mHHS and FJS with other measures of global perceived change by the patients and the clinicians are presented in Table 6.
Fig. 1.
Positive and very high correlation between mHHS and FJS
Table 6.
Correlation of mHHS and FJS with global perceived change as rated by the patients and functional ability by the clinician
| Global rating scale for perceived change | mHHS | FJS-12 |
|---|---|---|
| One-item pain scale | rs = 0.53, p < 0.0001 | rs = 0.55, p < 0.0001 |
| One-item satisfaction scale | rs = 0.50, p = 0.001 | rs = 0.42, p = 0.007 |
| One-item clinical condition change | rs = 0.60, p < 0.0001 | rs = 0.56, p < 0.0001 |
| Ability to sit with legs crossed | rs = 0.52, p = 0.001 | rs = 0.41, p = 0.009 |
| Ability for squatting | rs = 0.67, p = 0.001 | rs = 0.54, p = 0.009 |
rs Spearman correlation coefficient. p < 0.05 = significant
Test Retest Reliability
The mean mHHS on the third and fifth postoperative day was 23.6 ± 15.4 (range 11–57.2) and 35.4 ± 15.8 (range 15.4–60.5), respectively. The mean FJS on the third and fifth postoperative day was 3.2 ± 3.1 (range 0–10) and 10.7 ± 4.5 (range 3–15), respectively. The two-way random effects, absolute agreement ICC for mHHS and FJS was 0.80 (p = 0.005) and 0.34 (p = 0.06), respectively.
Floor and Ceiling Effect
On third postoperative day, none of the patients achieved the lowest possible score for mHHS whereas one patient (14.3%) achieved the lowest possible score of 0 point for FJS. At final follow-up, 5 out of 40 patients (12.5%) scored highest possible score on mHHS and FJS.
Interpretability
There was improvement in scores of mHHS and FJS in subgroups of patients based on reduction in pain (Table 7) [Fig. 2]. Contrary to expectation, both mHHS and FJS showed deterioration in score from unsatisfied to neutral subgroup of patients and thereafter, both patient-reported outcome instruments demonstrated improvement in their mean values as satisfaction level of patient improved (Table 6) [Fig. 3]. There was increase in the mean values of mHHS and FJS in subgroups of patients consistent with improvement in their clinical condition (Table 6) [Fig. 4]. There was very small improvement in mean values of mHHS and FJS between patients who were unable to sit with legs crossed and those who were able to sit with discomfort. However, there was big improvement in mean values of mHHS and FJS between patients who were able to sit with discomfort and those who were able to sit with legs crossed comfortably (Table 6) [Fig. 5]. There was increase in the mean values of mHHS and FJS consistent with improvement in the ability of patients to perform squatting (Table 6) [Fig. 6].
Table 7.
Interpretability of mHHS and FJS in five subgroups of patients
| Ordinal external anchors and subgroups of patients based on their response /ability | mHHS presented as mean ± SD | FJS presented as mean ± SD |
|---|---|---|
| One-item pain scale | ||
| Moderate pain (n = 3) | 59.0 ± 9.2 | 46.7 ± 8.0 |
| Mild pain (n = 5) | 68.2 ± 13.7 | 60.0 ± 15.4 |
| Very mild pain (n = 8) | 80.3 ± 10.8 | 72.5 ± 20.1 |
| No pain (n = 24) | 86.4 ± 12.0 | 82.8 ± 16.2 |
| ANOVA test result | p = 0.001 | p = 0.002 |
| Result of repeated comparison using Bonferroni correction | Significant difference between no pain and mild pain subgroups (p = 0.02) and between no pain and moderate pain subgroups (p = 0.004) | Significant difference between no pain and moderate pain subgroups (p = 0.006) |
| One-item satisfaction scale | ||
| Unsatisfied (n = 3) | 70.5 ± 10.7 | 70.2 ± 15.4 |
| Neutral (n = 2) | 57.2 ± 12.5 | 41.3 ± 5.3 |
| Satisfied (n = 22) | 79.4 ± 13.3 | 73.0 ± 18.7 |
| Very satisfied (n = 13) | 89.4 ± 11.3 | 85.2 ± 16.6 |
| ANOVA test result | p = 0.004 | p = 0.013 |
| Result of repeated comparison using Bonferroni correction | Significant difference between neutral and very satisfied (p = 0.01) | Significant difference between neutral and very satisfied (p = 0.013) subgroups of patients |
| One-item clinical condition change scale | ||
| Neutral (n = 3) | 51.7 ± 4.0 | 35.8 ± 10.1 |
| Slightly better (n = 19) | 77.4 ± 11.1 | 71.6 ± 15.8 |
| Very much better (n = 18) | 89.3 ± 10.1 | 85.4 ± 14.3 |
| ANOVA test result | p < 0.0001 | p < 0.0001 |
| Result of repeated comparison using Bonferroni correction | Significant difference between neutral and slightly better (p = 0.001), between neutral and very much better (p < 0.0001) and between slightly better and very much better (p = 0.004) subgroups of patients | Significant difference between neutral and slightly better (p = 0.001), between neutral and very much better (p < 0.0001) and between slightly better and very much better (p = 0.023) subgroups |
| Ability for cross leg sitting | ||
| Unable to sit (n = 7) | 72.8 ± 17.1 | 66.8 ± 22.5 |
| Able to sit with discomfort (n = 18) | 75.7 ± 11.7 | 69.4 ± 18.6 |
| Able to sit comfortably (n = 15) | 90.8 ± 10.5 | 86.0 ± 14.9 |
| ANOVA test result | p = 0.001 | p = 0.02 |
| Result of repeated comparison using Bonferroni correction | Significant difference between unable to sit cross legged and able to sit cross legged comfortably (p = 0.01) and between able to sit cross legged with discomfort and able to sit cross legged comfortably (p = 0.004) subgroups of patients | Significant difference between able to sit cross legged with discomfort and able to sit cross legged comfortably (p = 0.04) subgroup of patients |
| Ability for squatting | ||
| Unable to squat (n = 19) | 72.6 ± 12.9 | 66.1 ± 20.6 |
| Able to squat with discomfort (n = 9) | 79.7 ± 10.9 | 72.5 ± 14.2 |
| Able to squat comfortably (n = 12) | 94.8 ± 6.4 | 91.5 ± 8.8 |
| ANOVA test result | p < 0.0001 | p = 0.001 |
| Result of repeated comparison using Bonferroni correction | Significant difference between unable to squat and able to squat comfortably (p < 0.0001) between able to squat with discomfort and able to squat comfortably (p = 0.01) subgroup of patients | Significant difference between unable to squat and able to squat comfortably (p = 0.001) and between able to squat with discomfort and able to squat comfortably (p = 0.04) subgroup of patients |
Fig. 2.
Trend in mean values of mHHS and FJS in subgroups of patients based on response to the one-item pain scale
Fig. 3.
Unexpected deterioration in mean mHHS and mean FJS between unsatisfied and neutral subgroup of patients and thereafter gradual increase in score consistent with improvement in satisfaction level of the patients
Fig. 4.
Increase in mean mHHS and mean FJS consistent with perception of improvement in clinical condition of the patients
Fig. 5.
Increase in mHHS and FJS consistent with improvement in the ability for cross leg sitting
Fig. 6.
Increase in mean mHHS and mean FJS and improvement in ability of patients to perform squatting
Discussion
Our study has demonstrated that the modified version of the mHHS to suit the rural Indian population has adequate construct validity, reliability, interpretability and acceptable level of floor ceiling effect in patients undergoing bipolar hip hemiarthroplasty for fracture neck of the femur.
Construct Validity
Significant, positive, very high correlation was observed between mHHS and FJS thereby suggesting that both patient-reported outcome instruments have convergent construct validity and they measure similar sub-domains and constructs. The value of the correlation of mHHS with other constructs was more than or equal to the threshold value of 0.50 and hence, 100% of the results pertaining to the correlation of mHHS were in agreement to the formulated six hypotheses. Whereas, the FJS satisfied only four out of the six hypotheses (66.7%), because the value of correlation coefficient was lower than the threshold value of 0.50 for association with one-item satisfaction scale and functional ability for sitting cross legged. The FJS might not have adequate construct validity, because less than 75% of the hypotheses were confirmed. Both mHHS and FJS demonstrated highest correlation to squatting ability followed by one-item clinical condition change scale.
Test Retest Reliability
mHHS demonstrated good reliability, whereas FJS demonstrated poor and unacceptable test retest reliability and its p value was insignificant.
Ceiling and Floor Effects
Both mHHS and FJS demonstrated acceptable level of floor and ceiling effects in the present study. Presence of significant floor and ceiling effects suggests compromised content validity of an outcome instrument [24].
Interpretability
The mean and SD values of mHHS and FJS were described in five subgroup of patients thereby suggesting adequate interpretability of both mHHS and FJS. In the present study, both mHHS and FJS demonstrated trend of increase in their mean values consistent with improvement in clinical condition, pain improvement, ability to sit with legs crossed and ability to perform squatting. Only in the one-item satisfaction scale, there was decrease in the values of mHHS and FJS, whereas patients had reported improvement in satisfaction levels (from unsatisfied to neutral). One possible explanation for this paradoxical observation is that a previous study demonstrated that patient satisfaction might not be necessarily related to the hip arthroplasty procedure, because about 13% of very satisfied and satisfied patients paradoxically demonstrated lower or equal FJS scores postoperative as compared to preoperatively [25]. Another study reported weak correlation between patient satisfaction and the score obtained on the patient-reported outcome instrument thereby suggesting that patient-reported outcome instrument solely are not an optimal indicator of patient satisfaction [26]. Review of the figures and tables demonstrated that the mean values of mHHS were always higher than that of FJS though both the patient reported outcome instruments are scored on an identical scale from 0 to 100 points.
The mean value of mHHS in the published literature on functional outcome of bipolar hip hemiarthroplasty for fracture neck of the femur in elderly patients tends to range from 58.9 to 83.9 and the mean mHHS of 81.5 observed in our study was comparable to that already published [7, 8, 27] despite the alterations in the mHHS devised by us. As FJS has not been used to assess functional outcome of bipolar hemiarthroplasty, the published literature is on the use of FJS in THR. The mean FJS observed after THR for elective hip conditions varies from 54.3 to 75.1 [15, 28, 29] and the FJS observed in our study was 75.9 was comparable to the previously observed values.
The FJS refers to the level of perception of the prosthetic joint by the patient and consists of 12 questions during performance of various daily functional activities. Two questions of the FJS which were not answered by patients of the present study were: awareness while travelling in a car and awareness while playing favourite sport, because both activities were not being performed and were marked not applicable by majority of the patients. Also patients who were in the hospital on the third and fifth postoperative day could not be expected to perform household activity and gardening and hence this question was also marked as not applicable by the patients. One possible explanation for low test retest reliability and validity of the FJS is 3 out of 12 items of FJS being not relevant to the rural Indian population. Previous study [5] has shown that patients with university educational background tended to have significantly higher FJS compared to patients that had no college level education. Though we did not formally evaluate the educational background of the patients, it is possible that most patients from rural background probably had education only till school level and this could be another possible explanation for lower reliability and validity of the FJS in our study.
Comparison with Other Outcome Instruments for Indian Population
Rai et al. [30] described their modification of the HHS and the objective of the outcome tool was to evaluate the clinical outcome of bipolar hip hemiarthroplasty for femoral neck fracture. Their mHHS included both patient completed (85 points) and clinician completed items (15 points). The standard HHS and the mHHS have 44 points allotted to pain assessment, whereas the modification by Rai et al. only allotted 25 points for pain with fewer options thereby suggesting that pain assessment might probably be lesser with the modification suggested by Rai et al. Rai et al. added patient satisfaction (25 points) to the functional assessment heading and the assessment of satisfaction was never included in the standard HHS or the mHHS. Rai et al. gave maximum 10 points for assessment of squatting and 10 points for assessment of ability to sit with legs crossed, whereas the mHHS devised by us allotted maximum 5 points for being able to sit with legs crossed and 4 points for being able to squat with ease. Rai et al. gave only two options each for ability to sit cross legged and ability for squatting (unable to do/able to do) whereas we gave three options each for the ability to sit with legs crossed and squatting (unable to do/able to do with difficulty/able to do with ease). Giving more options to the patient helps to capture the experience of the patients more realistically. Rai et al. did not mention the mean and standard deviation of the modified HHS scored by the patients in their cohort. There was no description of the validity or interpretability of the outcome instrument thereby limiting the acceptability of the assessment tool. The modification of HHS was also not compared with another functional outcome instrument.
Vishwanathan et al. [31] modified the HHS to develop the mHHS by incorporating functional activities such as ability to squat and sit with legs crossed. The mHHS measured 10 items and was completed both by the patient (pain, function) [91 points] and the clinician (deformity, range of motion) [9 points]. The objective of the outcome tool was to evaluate the clinical outcome of proximal femoral nail fixation of pertrochanteric fractures. The validity and responsiveness of the mHHS were described but the mHHS was not compared with another functional outcome instrument in the study by Vishwanathan et al. [31].
Kumar et al. [32] demonstrated perfect correlation between the HHS and the mHHS. The work by Kumar et al. was the first study to validate the use of mHHS in the Indian population for assessment of functional outcome of total hip replacement. But the study excluded patients that underwent hemiarthroplasty of the hip and did not include assessment of squatting ability and ability to sit cross legged. Our study fills the gap by testing the construct validity of the mHHS in the Indian population and this being the first study remains the main highlight and strength of the present work. The mHHS used by Kumar et al. [32] did not have functional assessment of ability to sit with legs crossed and squatting ability of the patients, whereas the mHHS revised by us comprised of assessment of both ability for cross leg sitting and squatting. The population in the series of Kumar et al. was much younger (mean age 43 years; range 17–80 years) than the cohort of our study. Hence, due to the differences in the patient, intervention and outcome evaluation, we feel the reliability would be different in our cohort of patients that underwent bipolar hemiarthroplasty. Moreover, Kumar et al. mentioned the assessment of reliability using the Intraclass correlation coefficient of the mHHS but the study did not mention the time interval between the re-assessment of the mHHS for test retest reliability.
Strengths of the Study
Published literature has not described the use of mHHS and FJS to assess functional outcome after bipolar hemiarthroplasty for fracture neck of the femur in the Indian population and the present study is the first study to fill in that gap in knowledge. This is also the first study to describe the validity, reliability, floor ceiling effects and interpretability of mHHS and FJS in bipolar hemiarthroplasty of the hip in any population.
Limitations of the Study
The limitations of the study include a combined prospective and retrospective nature of the study and limited sample size. As per COSMIN checklist, our sample can be considered as moderate sample size (30–49 patients) [33]. The power of the study of our sample size was more than 97%. A retrospective study is prone to few biases. Selection of consecutively operated bipolar hemiarthroplasty cases helped to reduce selection bias in our study [34]. Performance bias was reduced, because multiple orthopaedic surgeons performed the surgical procedure at a single tertiary referral and university teaching hospital [34]. Reporting bias was mitigated by the use of commonly used patient reported instruments in hip arthroplasty [34]. Though mHHS and FJS were not specifically validated for bipolar hip arthroplasty, both the outcome instruments are commonly used to assess functional outcome after total hip arthroplasty. Recording bias was reduced by having well defined, objective and validated patient completed questionnaires and assessment of clinical outcome was performed by junior doctors who were not the operating surgeons of the patients [34]. Transfer bias or lost to follow-up bias was reduced, because we had data on 40 out of 46 patients (87%) and missing data (13%) were lesser than the maximum acceptable value of 20% [34]. Patients that were lost to follow-up had migrated to other places and were not willing to allow telephonic conversation for recording of their functional outcomes. Test retest reliability was tested in a small prospective sample of seven patients and it is possible that the results of reliability could be different if evaluated in a larger sample.
Utility of mHHS in Other Countries
Cross leg sitting (oriental sitting position) is considered crucial for certain customary, social and religious practices for people from the Middle Eastern countries [35] and Far East Asian countries [36]. Hence, the mHHS can be used in other Asian and Middle Eastern countries where sitting cross leg and squatting are important functions for activities of daily living. A context-specific outcome instrument is likely to be a better patient responsive measure [37] and the mHHS is one such outcome instrument.
Conclusion
The modified Harris hip score demonstrated adequate validity, reliability and interpretability in patients that underwent bipolar hemiarthroplasty of the hip for fracture neck of the femur. The mHHS is well suited to evaluate functional outcome of patients from the rural setting of the Indian subcontinent wherein local customs require ability to squat and sit cross legged FJS demonstrated inadequate construct validity and reliability and hence might not be suitable for assessment of functional outcome after bipolar hemiarthroplasty, because it might not be capturing information relevant to the population of interest. The mHHS was better than the FJS in capturing functional abilities such as sitting with legs crossed and squatting.
Acknowledgements
The authors wish to thank Professor Milind B Bhatt, Department of Statistics, Sardar Patel University, Anand, Gujarat for his help and guidance regarding relevant statistical analysis. The authors also wish to thank Dr Maulik and Dr Mohit for their assistance with data collection.
Author Contributions
CRediT author statement: KV: conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing—original draft, visualization, supervision, project administration. SKP: validation, supervision, project administration, writing—review and editing. RM: validation, resources, writing—review and editing. CC: validation, investigation, writing—review and editing.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Compliance with Ethical Standards
Conflict of interest
Karthik Vishwanathan, Shahrukh Khan Pathan, Ravi Makadia and Chintan Chaudhary declare that they have no conflict of interest.
Ethical standard statement
The study was conducted after obtaining approval from the Institutional Ethics Committee of HM Patel Centre for Medical Care and Education, Karamsad. The approval number for the project was IEC/HMPCMCE/93/Faculty/15/.
Informed consent
The study involved human subjects and the participants were given information sheet and the participants gave voluntary written informed consent for participating in the study.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Karthik Vishwanathan, Email: karthik.vishwanathan@paruluniversity.ac.in, Email: karthik_vishwanathan@yahoo.com.
Shahrukh Khan Ataullah Pathan, Email: shahrukhpathan98@gmail.com.
Ravi Chandulal Makadia, Email: risingravi123@gmail.com.
Chintan Bahecharbhai Chaudhary, Email: delly4959.cc@gmail.com.
References
- 1.Kannan A, Kancherla R, McMahon S, Hawdon G, Soral A, Malhotra R. Arthroplasty options in femoral-neck fractures: answers from the national registries. International Orthopaedics. 2012;36(1):1–8. doi: 10.1007/s00264-011-1354-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.von Roth P, Abdel MP, Harmsen WS, Berry DJ. Cemented bipolar hemiarthroplasty provides definitive treatment for femoral neck fractures at 20 years and beyond. Clinical Orthopaedics and Related Research. 2015;473(11):3595–3599. doi: 10.1007/s11999-015-4462-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Thorborg K, Roos EM, Bartels EM, Petersen J, Hölmich P. Validity, reliability and responsiveness of patient-reported outcome questionnaires when assessing hip and groin disability: a systematic review. British Journal of Sports Medicine. 2010;44(16):1186–1196. doi: 10.1136/bjsm.2009.060889. [DOI] [PubMed] [Google Scholar]
- 4.Pynsent PB. Choosing an outcome measure. Journal of Bone and Joint Surgery. British Volume. 2001;83(6):792–794. doi: 10.1302/0301-620X.83B6.0830792. [DOI] [PubMed] [Google Scholar]
- 5.Behrend H, Giesinger K, Giesinger JM, Kuster MS. The “forgotten joint” as the ultimate goal in joint arthroplasty: Validation of a new patient-reported outcome measure. Journal of Arthroplasty. 2012;27(3):430–436.e1. doi: 10.1016/j.arth.2011.06.035. [DOI] [PubMed] [Google Scholar]
- 6.Loth FL, Giesinger JM, Giesinger K, MacDonald DJ, Simpson AHRW, Howie CR, Hamilton DF. Impact of comorbidities on outcome after total hip arthroplasty. Journal of Arthroplasty. 2017;32(9):2755–2761. doi: 10.1016/j.arth.2017.04.013. [DOI] [PubMed] [Google Scholar]
- 7.van den Bekerom MP, Hilverdink EF, Sierevelt IN, Reuling EM, Schnater JM, Bonke H, Goslings JC, van Dijk CN, Raaymakers EL. A comparison of hemiarthroplasty with total hip replacement for displaced intracapsular fracture of the femoral neck: A randomised controlled multicentre trial in patients aged 70 years and over. Journal of Bone and Joint Surgery. British Volume. 2010;92(10):1422–1428. doi: 10.1302/0301-620X.92B10.24899. [DOI] [PubMed] [Google Scholar]
- 8.Stoffel KK, Nivbrant B, Headford J, Nicholls RL, Yates PJ. Does a bipolar hemiprosthesis offer advantages for elderly patients with neck of femur fractures? A clinical trial with 261 patients. ANZ Journal of Surgery. 2013;83(4):249–254. doi: 10.1111/ans.12048. [DOI] [PubMed] [Google Scholar]
- 9.Tol MC, van den Bekerom MP, Sierevelt IN, Hilverdink EF, Raaymakers EL, Goslings JC. Hemiarthroplasty or total hip arthroplasty for the treatment of a displaced intracapsular fracture in active elderly patients: 12-year follow-up of randomised trial. Bone Joint Journal. 2017;99B(2):250–254. doi: 10.1302/0301-620X.99B2.BJJ-2016-0479.R1. [DOI] [PubMed] [Google Scholar]
- 10.Kumar S, Sharma R, Gulati D, Dhammi IK, Aggarwal AN. Normal range of motion of hip and ankle in Indian population. Acta Orthopaedica et Traumatologica Turcica. 2011;45(6):421–424. doi: 10.3944/AOTT.2011.2612. [DOI] [PubMed] [Google Scholar]
- 11.Prakash V, Ganesan M, Vasanthan R, Hariohm K. Do commonly used functional outcome measures capture activities that are relevant for people with stroke in India? Topics in Stroke Rehabilitation. 2017;24(3):200–205. doi: 10.1080/10749357.2016.1234190. [DOI] [PubMed] [Google Scholar]
- 12.Vijayvargiya M, Shetty V, Makwana K, Suri HS. Mid-term results of an uncemented tapered femoral stem and various factors affecting survivorship. Journal of Clinical Orthopaedics and Trauma. 2019;10(2):368–373. doi: 10.1016/j.jcot.2018.02.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ng DZ, Lee KB. Unipolar versus bipolar hemiarthroplasty for displaced femoral neck fractures in the elderly: Is there a difference? Annals of the Academy of Medicine, Singapore. 2015;44(6):197–201. [PubMed] [Google Scholar]
- 14.Hamilton DF, Giesinger JM, MacDonald DJ, Simpson AH, Howie CR, Giesinger K. Responsiveness and ceiling effects of the Forgotten Joint Score-12 following total hip arthroplasty. Bone Joint Research. 2016;5(3):87–91. doi: 10.1302/2046-3758.53.2000480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hamilton DF, Loth FL, Giesinger JM, Giesinger K, MacDonald DJ, Patton JT, Simpson AH, Howie CR. Validation of the English language Forgotten Joint Score-12 as an outcome measure for total hip and knee arthroplasty in a British population. Bone Joint Journal. 2017;99B(2):218–224. doi: 10.1302/0301-620X.99B2.BJJ-2016-0606.R1. [DOI] [PubMed] [Google Scholar]
- 16.Adriani M, Malahias MA, Gu A, Kahlenberg CA, Ast MP, Sculco PK. Determining the validity, reliability, and utility of the Forgotten Joint Score: A systematic review. Journal of Arthroplasty. 2019 doi: 10.1016/j.arth.2019.10.058. [DOI] [PubMed] [Google Scholar]
- 17.Rolfson O, Bohm E, Franklin P, Lyman S, Denissen G, Dawson J, Dunn J, Eresian Chenok K, Dunbar M, Overgaard S, Garellick G, Lübbeke A. Patient-reported outcome measures working group of the International Society of Arthroplasty Registries. Patient-reported outcome measures in arthroplasty registries Report of the Patient-Reported Outcome Measures Working Group of the International Society of Arthroplasty Registries Part II. Recommendations for selection, administration, and analysis. Acta Orthopaedica. 2016;87(Suppl 1):9–23. doi: 10.1080/17453674.2016.1181816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Copay AG, Martin MM, Subach BR, Carreon LY, Glassman SD, Schuler TC, Berven S. Assessment of spine surgery outcomes: inconsistency of change amongst outcome measurements. Spine Journal. 2010;10(4):291–296. doi: 10.1016/j.spinee.2009.12.027. [DOI] [PubMed] [Google Scholar]
- 19.Terwee CB, Bot SD, de Boer MR, van der Windt DA, Knol DL, Dekker J, Bouter LM, de Vet HC. Quality criteria were proposed for measurement properties of health status questionnaires. Journal of Clinical Epidemiology. 2007;60(1):34–42. doi: 10.1016/j.jclinepi.2006.03.012. [DOI] [PubMed] [Google Scholar]
- 20.Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, Bouter LM, de Vet HC. The COSMIN study reached international consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient-reported outcomes. Journal of Clinical Epidemiology. 2010;63(7):737–745. doi: 10.1016/j.jclinepi.2010.02.006. [DOI] [PubMed] [Google Scholar]
- 21.Prinsen CAC, Mokkink LB, Bouter LM, Alonso J, Patrick DL, de Vet HCW, Terwee CB. COSMIN guideline for systematic reviews of patient-reported outcome measures. Quality of Life Research. 2018;27(5):1147–1157. doi: 10.1007/s11136-018-1798-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lenguerrand E, Wylde V, Brunton L, Gooberman-Hill R, Blom A, Dieppe P. Selecting, assessing and interpreting measures of function for patients with severe hip pathology: The need for caution. Orthopaedics and Traumatology: Surgery and Research. 2016;102(6):741–746. doi: 10.1016/j.otsr.2016.04.006. [DOI] [PubMed] [Google Scholar]
- 23.van Lummel RC, Walgaard S, Hobert MA, Maetzler W, van Dieën JH, Galindo-Garre F, Terwee CB. Intra-rater, Inter-rater and test-retest reliability of an instrument Timed Up and Go (iTUG) test in patients with Parkinson’s disease. PLoS ONE. 2016;11(3):e0151881. doi: 10.1371/journal.pone.0151881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.de Fátima-Costa-Oliveira N, Oliveira-Pena-Costa L, Nelson R, Maher CG, Beattie PF, de Bie R, Oliveira W, Camara-Azevedo D, da Cunha-Menezes-Costa L. Measurement properties of the Brazilian Portuguese version of the MedRisk instrument for measuring patient satisfaction with physical therapy care. Journal of Orthopaedic and Sports Physical Therapy. 2014;44(11):879–889. doi: 10.2519/jospt.2014.5150. [DOI] [PubMed] [Google Scholar]
- 25.Loth FL, Giesinger JM, Giesinger K, Howie CR, Hamilton DF. Single-item satisfaction score mask large variations in pain, function and joint awareness in patients following total joint arthroplasty. European Journal of Orthopaedic Surgery and Traumatology. 2020;30(2):267–274. doi: 10.1007/s00590-019-02560-x. [DOI] [PubMed] [Google Scholar]
- 26.Halawi MJ, Jongbloed W, Baron S, Savoy L, Cote MP, Lieberman JR. Patient-reported outcome measures are not a valid proxy for patient satisfaction in total joint arthroplasty. Journal of Arthroplasty. 2020;35(2):335–339. doi: 10.1016/j.arth.2019.09.033. [DOI] [PubMed] [Google Scholar]
- 27.Sonaje JC, Meena PK, Bansiwal RC, Bobade SS. Comparison of functional outcome of bipolar hip arthroplasty and total hip replacement in displaced femoral neck fractures in elderly in a developing country: a 2-year prospective study. European Journal of Orthopaedic Surgery and Traumatology. 2018;28(3):493–498. doi: 10.1007/s00590-017-2057-y. [DOI] [PubMed] [Google Scholar]
- 28.Larsson A, Rolfson O, Kärrholm J. Evaluation of forgotten joint score in total hip arthroplasty with Oxford Hip Score as reference standard. Acta Orthopaedica. 2019;90(3):253–257. doi: 10.1080/17453674.2019.1599252. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Rosinsky PJ, Chen JW, Lall AC, Shapira J, Maldonado DR, Domb BG. Can we help patients forget their joint? Determining a threshold for successful outcome for the Forgotten Joint Score. Journal of Arthroplasty. 2020;35(1):153–159. doi: 10.1016/j.arth.2019.08.014. [DOI] [PubMed] [Google Scholar]
- 30.Rai AK, Agarwal R, Singh S, Ratan R. The BHU bicentric bipolar prosthesis in fracture neck femur in active elderly. Journal of Trauma Management and Outcomes. 2008;2(1):7. doi: 10.1186/1752-2897-2-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Vishwanathan K, Akbari K, Patel AJ. Is the modified Harris hip score valid and responsive instrument for outcome assessment in the Indian population with pertrochanteric fractures? Journal of Orthopaedics. 2018;15(1):40–46. doi: 10.1016/j.jor.2017.12.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kumar P, Sen R, Aggarwal S, Agarwal S, Rajnish RK. Reliability of Modified Harris Hip Score as a tool for outcome evaluation of total hip replacements in Indian population. Journal of Clinical Orthopaedics and Trauma. 2019;10(1):128–130. doi: 10.1016/j.jcot.2017.11.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Terwee CB, Mokkink LB, Knol DL, Ostelo RW, Bouter LM, de Vet HC. Rating the methodological quality in systematic reviews of studies on measurement properties: A scoring system for the COSMIN checklist. Quality of Life Research. 2012;21(4):651–657. doi: 10.1007/s11136-011-9960-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.ISAKOS Scientific Committee. Audigé L, Ayeni OR, Bhandari M, Boyle BW, Briggs KK, Chan K, Chaney-Barclay K, Do HT, Ferretti M, Fu FH, Goldhahn J, Goldhahn S, Hidaka C, Hoang-Kim A, Karlsson J, Krych AJ, Laprade RF, Levy BA, Lubowitz JH, Lyman S, Ma Y, Marx RG, Mohtadi N, Marcheggiani-Muccioli GM, Nakamura N, Nguyen J, Poehling GG, Roberts LE, Rosenberg N, Shea KP, Sohani ZN, Soudry M, Voineskos S, Zaffagnini S. International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine. A practical guide to research: design, execution, and publication. Arthroscopy. 2011;27(4Suppl):S1–112. doi: 10.1016/j.arthro.2011.02.001. [DOI] [PubMed] [Google Scholar]
- 35.Assi CC, Barakat HB, Caton JH, Najjar EN, Samaha CT, Yammine KF. Mortality rate and mid-term outcomes of total hip arthroplasty using dual mobility cups for the treatment of femoral neck fractures in a middle eastern population. Journal of Arthroplasty. 2019;34(2):333–337. doi: 10.1016/j.arth.2018.10.030. [DOI] [PubMed] [Google Scholar]
- 36.Zhou H, Liu A, Wang D, Zeng X, Wei S, Wang C. Kinematics of lower limbs of healthy Chinese people sitting cross-legged. Prosthetics and Orthotics International. 2013;37(5):369–374. doi: 10.1177/0309364612470964. [DOI] [PubMed] [Google Scholar]
- 37.Suk M, Norvell DC, Hanson B, Dettori JR, Helfet D. Evidence-based orthopaedic surgery: What is evidence without the outcomes? Journal of American Academy of Orthopaedic Surgeons. 2008;16(3):123–129. doi: 10.5435/00124635-200803000-00003. [DOI] [PubMed] [Google Scholar]