Measurement properties of the Knee Injury and Osteoarthritis Outcome Score Patello-Femoral questionnaire in Saudi Arabians

Study protocol

The present validation process has been approved and with the permission from the original developer (dated 29/08/2018) of the instrument, University of Southern Denmark, Denmark. The process of translation of Arabic abstract can be found on the website (http://www.koos.nu/). KOOS-PF contains 11 questions out of which one is for symptom; nine for pain and one about quality of life. The exact phrase of questions is PF1. How severe is your knee stiffness after exercise? PF2. How often do you experience knee pain after stopping activity? PF3. How often does pain limit your activity? PF4. Rising from sitting (including getting out of the car). PF5. Kneeling PF6. Squatting PF7. Heavy household activities (including carrying and lifting) PF8. Hopping/jumping PF9. Running/jogging PF10. After sport and recreational activities PF11. Have you modified your sport or recreational activities due to your knee pain? Each question contains 5 options never to always (less disability to maximum disability); according to guidelines from the main developers of the KOOS, it is mandatory to mention the direction of scoring the data for better interpretation of results where the highest obtained values indicate no symptoms and lowest were the extreme symptoms. The method of scoring should be to calculate the mean score of the 11 items and divide by the maximum possible score for each item (i.e., 4), using the formula 100–[(mean score PF1-PF11)/4 * 100] (http://www.koos.nu/).

The present study has followed the Beaton et al. (2000) guidelines of the validation process. The religious-cultural activity (prayer/ ‘Salat’) has been taken into consideration while adapting the questionnaire as the Muslim population offer prayers five times every day and involves the knee joint activity in different positions of the prayer. Two translators have involved in Stage I of the process to translate the original (forward) to target language by (T1-informed clinical expert and T2-uninformed non-clinical expert), whereby the English version of KOOS-PF was first translated into the target dialect of the Arabic language. The Questionnaire was synthesized as an Arabic draft (T-12) after analysis of language discrepancies by the two translators (T1&T2) during stage II.  The translated questionnaire was back-translated to English separately by two blinded, Arabic-native English professors to create two English versions (BT1 & BT2) during stage III (Beaton et al., 2000).  All the suggestions were taken into consideration from the initial draft written by the first two translators and back-translation by two qualified English language experts and one Orthopedic surgeon, who could suggest technical discrepancies to produce a pre-final version during stage IV.  During the stage V, the pre-final version was then administered to twenty local Arabic-speaking citizens, who were suffering from anterior knee pain (twelve men and eight women) and checked for the flow of the language and understanding of the items related to their knee problem. The panel then considered the opinions of all these subjects and agreed to add Arabic words ‘Sujud’ and ‘Rakaa’ to the items wherever suitable. The word ‘Sujud’ refers to touching the forehead to the floor during (‘Salat’) prayers, (i.e., “prostrating” and the word ‘Rakaa’ refers to kneel sitting) to the item No. PF1(stiffness) “How severe is your knee stiffness after exercise/prostrating, kneel sitting?”; the word prayer to PF2 (Pain). “How often do you experience knee pain after stopping activity/prayer?”; item No. PF3 “How often does pain limit your activity/prostrating, kneel sitting?”; item No. PF4 “Rising from sitting” (including getting out of the car/ prostrating, kneel sitting); item No. PF5. “Kneeling” (during prayers); item No.PF6 “Squatting” (during prayers); item No.PF10. (pain) “After sport and recreational activities”, prayers/ ‘Sujud’, ‘Rakaa”. The word ‘Salat’ refers to prayer added to item No.PF11. “Have you modified your sport or recreational activities and Salat (prayers) due to your knee pain?” After the subjects were convinced with the items, a final draft evolved as a complete pre-testing questionnaire (stage V). After all the above stages, the measurement properties were measured in target individuals with chronic anterior knee pain.

Study sample

A total of 95 male subjects with anterior knee pain were selected in the present study. The inclusion criteria were patients aged between 32 to 66 years as the patellofemoral pain affects the young and middle-aged (Collins et al., 2019) commonly in whom sports/recreational/praying activity becomes the main reason for suffering the anterior knee pain (PFPS). The final draft of the questionnaire was explained in detail to the patients, including the average time required to fill in the self-reported form. The following exclusion criteria were implemented to prevent biased outcomes: presence of the other variant rheumatic diseases, severe inflammatory arthritis, knee problems other than anterior knee pain, low back pain, corticosteroid therapy as confirmed by Orthopedic Surgeon.

Data collection

Patients were selected through a convenient sampling technique from the outpatient departments of three different hospitals in and around Al Majmaah city, KSA who were diagnosed by an Orthopedic surgeon, from 1st November 2019 to 31st January 2020. Patients who were eligible with the inclusion criteria were asked to participate in the study. All included patients were provided with informed written consent; they were then asked to complete the translated Arabic questionnaire on two different occasions, with a 48-hour gap in-between; to ensure that their answers were reliable. The questionnaire was explained to all patients with mention of its 11 items of KOOS-PF related to their anterior knee pain. Patients were given the questionnaire soon after the diagnosis of their knee condition. Among the 95 subjects who were asked to come back after 48 h to refill the questionnaire, 84 patients (88.4% compliance rate) responded. I chose the 48-hour time gap to avoid measuring the effects of prescribed drugs on pain; as well as to prevent recall bias in questionnaire administration. Subjective pain was measured using the numeric visual analog scale (VAS). The psychometric measurement properties of the translated version such as internal consistency, test-retest reliability, floor-ceiling effect, inter-item correlation, the percentage of agreement and construct validity were done using IBM-SPSS statistical (Version 24.0) software.

Measurement of psychometric properties

Test-retest reliability

The Intraclass correlation coefficient (ICC) (single measure) was measured to see the difference between the first and second (Test and retest) self-rated form for its reliability calculated with 95% confidence interval (Rattray & Jones, 2007; Juul et al., 2016). The two-way random effect model and an absolute agreement were utilized to measure the ICC. ICC value should be >0.80 for good test-retest reliability. Forty-eight hours (2 days) of time interval was opted for evaluating test-retest reliability to avoid the outcome bias due to treatment drug effect.

Inter-item correlation matrix

This matrix within the subscale should be between 0.3 –0.8; anything less than 0.3 and more than 0.8 should be considered as redundant hence removed (Juul et al., 2015).

Percentage of agreement

The percentage of the agreement was calculated in two ways, coefficient of variance (CV) in percentage (Simonsen, 1995) and minimal detectable change (MDC) in % (Lee et al., 2013). The CV was calculated with SD divided by mean which provides a ratio for variability. This ratio can be multiplied by 100 for percentile. The MDC could be calculated by two steps, the first measurement of error (SEM) by SD_Pooled $\sqrt{\left(1-r\right)}$; secondly measurement of MDC by 1.96* $\sqrt{2}$*SEM. Thirdly MDC value will be divided into mean and this value can be multiplied by 100 for percentile. Values less than 30% is acceptable and if values are less than 10% that is excellent.

Construct validity

The construct validity was assessed at obtaining baseline data through Spearman correlations to check the relationship between the KOOS-PF scale and with the numeric visual analog scale (VAS). Standard values of more than 0.70 were considered strong, between 0.70–0.50 moderate and less than 0.50 weak for Correlation coefficients (Terwee et al., 2007).

Floor and ceiling effect

The floor and ceiling effects were evaluated for the items of the KOOS-PF questionnaire. The obtained scores are statistically acceptable when less than 15% of the participants reporting the lowest or highest possible score (Roos et al., 1998). Ceiling (highest 100%) and floor (Lowest 0%) effect of each individual subscale has been observed individually in percentage (Roos et al., 1998).

Inter-item correlation matrix

Table 5 shows the Coefficient of variation (CV) as a percentage, of all items in the Arabic KOOS-PF. These CV values indicate that no item was precise but overall items in the Arabic KOOS-PF fell within the acceptable range (<30%).

When the correlation analysis was done between all the items of the KOOS-PF subscale, all the values were found below 0.8. Minimum inter-item correlation found between PF3 and PF11 (0.318) whereas, the maximum inter-item correlation found between PF2 and PF8 (0.784). None exceeded the standard value of 0.8 which indicates a good inter-item relationship (Table 6).

Construct Validity

VAS has a moderate correlation with the KOOS-PF items (−0.568).

Floor and Ceiling effect

Table 7 shows the floor (lowest) and ceiling (highest) values of individual items (questions) as there was no floor or ceiling effect for individual items in Arabic KOOS-PF (Table 7). The lowest floor effect opted was 4.7% for the item No. PF8 and the highest was 13.9% for the PF4. The lowest Ceiling effect opted was 2.2% for the item No. PF1 and the highest was 13.4% for the PF11.

Strength and limitations

The strength of this study is the only available Saudi dialect of Arabic questionnaire to evaluate the patellofemoral pain syndrome in the Saudi Arabian Muslim populace, which is well adapted based on the religious practice.

Duration between test and retest for reliability is too short but the author decided to use short interval to avoid any analgesic effect of the drug. Author used only the male subjects in the validation process as the male therapists are not allowed to treat/evaluate female patients in the hospitals of Saudi Arabia hence, the results may not apply to female patients. Future studies should attempt to measure this property in Saudi female patients. No validation studies of KOOS-PF have been done either in any other Arabic dialect or language other than the original English KOOS-PF version to support the results. The responsiveness property was not measured and intends to measure in the future by the same author.