Translation, Cultural Adaptation and Validation of the Kidney Disease Quality of LifeeShort Form 1.3 in an African Country H. Rhoua, F. Ezaitounia, N. Ouzeddouna, R. Bayahiaa, K. Elhajjib, R. Roudièsb, F.Z. Sekkatb, R. Razinec, R. Abouqalc, A. Alghadid, A. Azennoude, and L. Benamara a Department of Nephrology-Dialysis-Transplantation, Ibn Sina University Hospital, Rabat, Morocco; bDepartment of Psychiatry, Ibn Sina University Hospital, Rabat, Morocco; cDepartment of Epidemiology, Ibn Sina University Hospital, Rabat, Morocco; dDepartment of English, Faculty of letters and Humanities, Rabat, Morocco; and eAl Akhawayn University, Ifrane, Morocco
ABSTRACT Background. The impact of dialysis on patient quality of life has been recognized as an important outcome measure. The Dialysis Outcomes and Practice Patterns Study compared quality of life in 4 continents [1], but very scarce information is available about dialysis patients’ quality of life in Africa. The objective of this study was to translate the Kidney Disease Quality of LifeeShort Form (KDQOL-SF) into Moroccan and measure its psychometric properties. Methods. The questionnaire was first translated into Moroccan by 2 independent translators, and then 2 backward translations into English were performed after pretesting in 10 dialysis patients. The final questionnaire was then administered to 80 dialysis patients. Reliability was estimated by internal consistency and testeretest reliability. Validity was assessed using known group comparisons and correlations between overall health rating and scales scores. Results. Some activities were substituted since they were not common in Morocco. All subscales had a Cronbach a above the recommended value except for 3 scales. All of the items showed good testeretest reliability. Correlation of items within subscales was higher than that of items outside subscales in 87% of cases. Regarding construct validity, all KDQOL-SF scales had significant correlation with overall health rating except for sexual function and dialysis staff encouragement. Furthermore, the questionnaire could be used to discriminate between subgroups of the patients. Conclusions. The psychometric properties of the KDQOL-SF resulting from this firsttime administration of the instrument support the validity and reliability of the KDQOL-SF as a measure of quality of life of patients having hemodialysis in Morocco.
H
EALTH-RELATED QUALITY OF LIFE (QOL) is known to be impaired in patients undergoing dialysis compared to general population [2,3]. In the last decades, advances in dialysis therapy led to increasing life expectancy, but few efforts were made to improve patients’ QOL. Currently, QOL is increasingly recognized as one way of measuring treatment outcome [4,5]. As a consequence, a number of generic and specific questionnaires of QOL have been developed. The Kidney Disease Quality of LifeeShort Form (KDQOLSF) is a disease-specific instrument, developed in the United
States by Ron Hays to assess the functioning and well-being of people with kidney disease [6]. It is one of the most complete instruments currently available in patients with chronic kidney disease [2]. This questionnaire has been translated in multiple languages [7e11] and has been widely used in studies, particularly in the Dialysis Outcomes and
*Address correspondence to Bayane Bouidida, 30, Voie Romaine, Internat Pasteur, 06000, Nice, France. E-mail: b.bayane@hotmail. com
ª 2014 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/14/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2014.02.011
Transplantation Proceedings, 46, 1295e1301 (2014)
1295
1296
Practice Patterns Study [1] and in the United States Renal Data SystemdAnnual Data Report [12]. To date, a Moroccan version of the questionnaire has not yet been developed. Therefore, the purpose of this study was to translate the KDQOL-SF into Moroccan, to adapt it culturally and test its psychometric properties in a Moroccan dialysis population. PATIENTS AND METHODS The Questionnaire The KDQOL-SF consists of 79 items divided into 19 scales (8 generic scales, 11 disease-specific scales) and an overall health rating item [6]: Short-form 36 (SF-36) [13] (8 dimensions/36 items): physical functioning (10 items), role limitations caused by physical problems (4 items), role limitations caused by emotional problems (3 items), pain (2 items), general health (5 items), social functioning (2 items), emotional well-being (5 items), energy/ fatigue (4 items), and 1 item about health status compared to 1 year previously. Kidney diseaseetargeted items (11 dimensions/43 items): symptom/problem list (12 items), effects of kidney disease (8 items), burden of kidney disease (4 items), cognitive function (3 items), quality of social interaction (3 items), sexual function (2 items), sleep (4 items), social support (2 items), work status (2 items), patient satisfaction (1 item), dialysis staff encouragement (2 items). Overall health rating (1 item scored separately). For each scale, a score can be calculated according to a standard procedure available on: http://gim.med.ucla.edu/kdqol. Scores range from 0 to 100, with a higher score indicating better QOL.
Translation Procedure The study protocol was approved by R. Hays and the local ethical committee of Ibn Sina University Hospital Center. The translation steps were carried out according to specifications established by the KDQOL Working Group. Forward Adaptation. 1. Items and response choices of the original version of the KDQOL-SF were independently translated into Moroccan by 2 translators whose first language is Moroccan. Following this, a seminar including translators, nephrologists, psychiatrists, and a sociologist was organized. All the options were reviewed, problems with specific terms and differences in the cultures of the countries were resolved, and a consensus version of the questionnaire was made. After this meeting, the translators independently rated the level of difficulty of the translation between 0 (not at all difficult) and 100 (extremely difficult). 2. Two external translators evaluated the equivalence of each item and response scale according to a scale from 0 (“not at all equivalent”) to 100 (“exactly equivalent”). 3. The Moroccan version of the questionnaire was tested in a sample of 10 patients on hemodialysis to detect comprehension problems. Mean patient age was 39 13 years (20e57 years), 6 were males, and 8 had not completed elementary school or were illiterate. The test took the form of face-to-face interviews immediately followed by retrospective interviews about comprehension problems and choices among a number of alternative renderings.
RHOU, EZAITOUNI, OUZEDDOUN ET AL Backward Adaptation. Two backward translations were done by English native translators and a consensus version was compared to the original English version.
Test Field To measure psychometric properties, the Moroccan version was administered to 80 patients on dialysis at Ibn Sina Hospital. Inclusion criteria were (1) age more than 18 years, (2) dialysis treatment for 3 months without life-threatening disease at the time of the test, (3) patient consent. Exclusion criteria were language difficulties. In this cross-sectional study, 78% of patients were in hemodialysis and 22% in peritoneal dialysis (Table 1). The majority were females and mean age was 43.9 14.2 years. The mean time on dialysis among the patients was 104.8 120 months.
Statistical Analysis and Psychometric Properties Means of each scale and standard deviation were determined. Reliability and validity were determined. Percentages of floor and ceiling were also assessed. Ceiling effects were taken as being the percentage of respondents with scores of 100 and floor effects were the percentage of respondents having a score of 0. Ceiling and floor effects should be less than 30% to ensure that the scale captures the full range of potential responses within the population and that changes over time can be detected. Multitrait/Multi-Item Correlations. Correlation coefficients were calculated by using Pearson correlation to assess the strength of relationships between items within and outside each scale. The item internal consistency assesses item correlation with a corresponding scale, corrected for overlap. A correlation corrected for overlap is the correlation of an item with the remaining items of the scale, which removes the bias of correlating an item with itself. A correlation of 0.4 or more has been used as the standard for supporting the item’s internal consistency. The discriminant validity was assessed by comparing item correlation within and outside each scale. We hypothesized that each Table 1. Patient Characteristics Parameters
Patients on Dialysis (n ¼ 80)
Age (y) Sex ratio Peritoneal dialysis (%) Hemodialysis (%) Educational level (%) Illiterate Elementary High school College Civil status: married (%) Yes No Duration on dialysis (mo) Cause of kidney disease (%) Glomerulonephritis Diabetes Tubulointerstitial nephritis Vascular Unknown etiology Other
43.9 14.2 0.7 22 78 28 22 23 27 61 39 104.8 120 26 14 18 5 34 3
KDQOL-SF 1.3 IN AN AFRICAN COUNTRY item’s correlation with its intended scale would be at least two standard errors greater than its correlation with other scales. Internal Consistency Reliability and TesteRetest Reliability. Cronbach a coefficient was used to assess internal consistency reliability. As recommended by Nunnally [14], a Cronbach value of 0.70 was taken as an indicator of adequate internal consistency reliability. Testeretest reliability was assessed by convenience sample (n ¼ 20) of the original respondents. The same questionnaire was administered to these patients after 10 to 14 days. Testeretest reliability was measured using intraclass correlation coefficient. Construct Validity. Construct validity was assessed by comparing scales of KDQOL-SF with overall health rating by means of the Pearson and Spearman correlation coefficient. External Discriminant Validity. External discriminant validity assesses the ability of the KDQOL-SF subscales to differentiate between patient groups with different clinical characteristics. We used the t test to examine the differences of QOL scores in subgroups of patients based on sex, educational level, marital status. We also looked at the correlations of scales scores with age, and dialysis duration by using Pearson and Spearman correlations. The SPSS 13 statistical package for Windows (SPSS, Chicago, Ill, United States) was used for statistical analysis.
RESULTS Translation Procedure
Forward Translation. Difficulty mean rates ranged from 0 to 60 with an average difficulty rating of 13.1 3.5 (results not shown). Items and response choice equivalence between the Arabic and the US version ranged between 75 and 100, and the mean equivalence rating was 93.3. Cultural adaptation concerned some activities rather uncommon in Morocco like “pushing a vacuum cleaner,” “playing golf,” and “bowling” (item SF3b), which needed to be replaced by activities requiring the use of similar body parts. Changes also involved distances that were rendered in meters rather than blocks and miles (items SF3g, SF3h, SF3i). Lexical problems were met with the following words: Physical health as the word health implies physical health in Moroccan. It was finally rendered by your health (ie, of your body). Activities rendered by matters, things to do. It was preferred to a word-for-word translation, as the seemingly equivalent word is too formal. Dissatisfied changed to not at all satisfied and then somewhat dissatisfied was rendered by not very satisfied. Definitely true and definitely false, respectively, rendered by 100% true and 100% false. Syntactic changes were often necessary when there were differences in structure between the English and the Arabic sentence. Few words were changed during the pilot test and some explanations were added in parenthesis (ie, “You respond after a delay to things that were said or done around you” was added in parentheses in item SF13b). The patients were a bit puzzled by the item SF11b where they were asked if their health was “excellent.” The interviewers explained that they were asked about their subjective point of view and not about their medical state.
1297
Backward Translation. The backward translation caused few problems. A few words were corrected before the tests. Field Test
Completeness of Data. Completeness at the scale level was of 100% except for portions on dialysis staff encouragement (99%) and sexual function (40%). Floor and Ceiling. The scales work status, role limitations caused by physical problems, and role limitations caused by emotional problems suffered from high percentage of floor scores, whereas social support dialysis and staff encouragement had a high percentage of ceiling scores (Table 2). Multitrait/Multi-item Correlations, Internal Consistency Reliability, and TesteRetest Reliability. The item internal consistency, corrected for overlap, was above 0.4 for 94% of items. Correlation of items within subscales was higher than that of items outside subscales in 87% of cases. Internal consistency reliability was adequate, with Cronbach a values above 0.70 for all scales except 3 dimensions: quality of social interaction (0.67), work status (0.38), and cognitive function (0.57; Table 2). Cronbach a did not improve after removing 1 item from quality of social interaction and cognitive function scales. No item could be removed from work status scale as it only consisted of 2 items. The intraclass coefficient correlation of the kidney diseaseetargeted scales ranged from 0.67 (pain) to 0.9 (energy/fatigue, symptom/problems and overall health rating; Table 2). Construct Validity. Health rating was significantly correlated with all KDQOL-SF scales except sexual function and dialysis staff encouragement and patient satisfaction scales (Table 3). External Discriminating Validity. Table 4 shows the correlation coefficients between the scales of the KDQOL-SF and age and dialysis duration. There are significant correlations between age and 5 subscales of the KDQOLSF: physical functioning (r ¼ 0.27; P ¼ .006), energy/ fatigue (r ¼ 0.33; P ¼ .001), work status (r ¼ 0.23; P ¼ .023), burden of kidney disease (r ¼ 0.22; P ¼ .037), overall health rating (r ¼ 0.23; P ¼ .024). Duration of dialysis correlated significantly with 6 scales of the KDQOL-SF: pain (r ¼ 0.27; P ¼ .008), energy/fatigue (r ¼ 0.26; P ¼ .01), work status (r ¼ 0.29; P ¼ .004), sexual function (r ¼ 0.28; P ¼ .09), sleep (r ¼ 0.26; P ¼ .001), and overall health rating (r ¼ 0.23; P ¼ .022). We also compared mean KDQOL-SF scores in groups expected to differ in their QOL (Fig 1). Regarding gender, men scored significantly higher than women on symptom/ problem, cognitive function, physical functioning, social functioning, and pain (Fig 1A). Literate patients compared with illiterate patients showed significant higher scores on 5 scales: burden of kidney disease, work status, role limitations caused by physical problems, role limitations caused by emotional problems, and energy/fatigue (Fig 1B). Illiterate patients showed significant higher scores on dialysis staff encouragement and patient satisfaction. Regarding marital
1298
RHOU, EZAITOUNI, OUZEDDOUN ET AL
Table 2. Scores Obtained for the Moroccan Version, Percentage of Floor and Ceiling, Internal Consistency, and TesteRetest Reliability Dimensions
No. of Items
Physical functioning Role limitations caused by physical problems Pain General health perceptions Emotional well-being Role limitations caused by emotional problems Social functioning Energy/fatigue Symptom/problem Effects of kidney disease Work status Cognitive function Quality of social interaction Sexual function Sleep Social support Burden of kidney disease Dialysis staff encouragement Patient satisfaction Overall health rating
10 4 2 5 5 3 2 4 12 8 2 3 3 2 4 2 4 2 1 1
Mean SD
Floor (%)
Ceiling (%)
Cronbach a
Intraclass Correlation
2.1 46.3 5.3 1.1 0 35.8 1.1 0 0 11 37.9 0 0 3.2 2.1 0 7.4 2.1 3.2 3.2
6.3 14.7 29.5 0 3.2 25.3 17.9 0 6.3 5.3 18.9 12.6 24.2 16.8 3.2 55.8 3.2 47.4 13.7 9.5
0.86 0.81 0.89 0.71 0.83 0.74 0.7 0.8 0.82 0.8 0.38 0.57 0.67 0.9 0.74 0.7 0.77 0.82 NA NA
0.89 0.81 0.67 0.84 0.87 0.83 0.8 0.9 0.9 0.88 0.82 0.85 0.8 0.81 0.84 0.78 0.88 0.81 0.87 0.9
68 33 67 42 61 43 66 51 66 63 40 73 74 73 64 83 43 85 57 59
23 38 30 23 20 40 25 21 20 22 37 20 22 31 21 24 29 21 24 22
Abbreviations: NA, not applicable for a single-item measure; SD, standard deviation.
status, no significant differences were observed for scores of married and unmarried patients (results not shown). DISCUSSION
To our knowledge, this has been the first study to evaluate the reliability and validity of the KDQOL-SF among dialysis patients in an African country. This questionnaire would then give information about QOL in this continent and allow international comparisons of QOL in dialysis. Table 3. Correlation Coefficients (With P Values) Between the Kidney Disease-Targeted Scales and the Overall Health Rating Scale
Dimensions
Overall Health Rating
P Value
Physical functioning Role limitations caused by physical problems Pain General health perceptions Emotional well-being Role limitations caused by emotional problems Social functioning Energy/fatigue Symptom/problem Effects of kidney disease Work status Cognitive function Quality of social interaction Sexual function Sleep Social support Burden of kidney disease Dialysis staff encouragement Patient satisfaction
0.53 0.45 0.51 0.52 0.29 0.36 0.28 0.49 0.55 0.32 0.38 0.23 0.23 0.16 0.32 0.4 0.36 0.13 0.12
ABSTRACT Background. The impact of dialysis on patient quality of life has been recognized as an important outcome measure. The Dialysis Outcomes and Practice Patterns Study compared quality of life in 4 continents [1], but very scarce information is available about dialysis patients’ quality of life in Africa. The objective of this study was to translate the Kidney Disease Quality of LifeeShort Form (KDQOL-SF) into Moroccan and measure its psychometric properties. Methods. The questionnaire was first translated into Moroccan by 2 independent translators, and then 2 backward translations into English were performed after pretesting in 10 dialysis patients. The final questionnaire was then administered to 80 dialysis patients. Reliability was estimated by internal consistency and testeretest reliability. Validity was assessed using known group comparisons and correlations between overall health rating and scales scores. Results. Some activities were substituted since they were not common in Morocco. All subscales had a Cronbach a above the recommended value except for 3 scales. All of the items showed good testeretest reliability. Correlation of items within subscales was higher than that of items outside subscales in 87% of cases. Regarding construct validity, all KDQOL-SF scales had significant correlation with overall health rating except for sexual function and dialysis staff encouragement. Furthermore, the questionnaire could be used to discriminate between subgroups of the patients. Conclusions. The psychometric properties of the KDQOL-SF resulting from this firsttime administration of the instrument support the validity and reliability of the KDQOL-SF as a measure of quality of life of patients having hemodialysis in Morocco.
H
EALTH-RELATED QUALITY OF LIFE (QOL) is known to be impaired in patients undergoing dialysis compared to general population [2,3]. In the last decades, advances in dialysis therapy led to increasing life expectancy, but few efforts were made to improve patients’ QOL. Currently, QOL is increasingly recognized as one way of measuring treatment outcome [4,5]. As a consequence, a number of generic and specific questionnaires of QOL have been developed. The Kidney Disease Quality of LifeeShort Form (KDQOLSF) is a disease-specific instrument, developed in the United
States by Ron Hays to assess the functioning and well-being of people with kidney disease [6]. It is one of the most complete instruments currently available in patients with chronic kidney disease [2]. This questionnaire has been translated in multiple languages [7e11] and has been widely used in studies, particularly in the Dialysis Outcomes and
*Address correspondence to Bayane Bouidida, 30, Voie Romaine, Internat Pasteur, 06000, Nice, France. E-mail: b.bayane@hotmail. com
ª 2014 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710
0041-1345/14/$esee front matter http://dx.doi.org/10.1016/j.transproceed.2014.02.011
Transplantation Proceedings, 46, 1295e1301 (2014)
1295
1296
Practice Patterns Study [1] and in the United States Renal Data SystemdAnnual Data Report [12]. To date, a Moroccan version of the questionnaire has not yet been developed. Therefore, the purpose of this study was to translate the KDQOL-SF into Moroccan, to adapt it culturally and test its psychometric properties in a Moroccan dialysis population. PATIENTS AND METHODS The Questionnaire The KDQOL-SF consists of 79 items divided into 19 scales (8 generic scales, 11 disease-specific scales) and an overall health rating item [6]: Short-form 36 (SF-36) [13] (8 dimensions/36 items): physical functioning (10 items), role limitations caused by physical problems (4 items), role limitations caused by emotional problems (3 items), pain (2 items), general health (5 items), social functioning (2 items), emotional well-being (5 items), energy/ fatigue (4 items), and 1 item about health status compared to 1 year previously. Kidney diseaseetargeted items (11 dimensions/43 items): symptom/problem list (12 items), effects of kidney disease (8 items), burden of kidney disease (4 items), cognitive function (3 items), quality of social interaction (3 items), sexual function (2 items), sleep (4 items), social support (2 items), work status (2 items), patient satisfaction (1 item), dialysis staff encouragement (2 items). Overall health rating (1 item scored separately). For each scale, a score can be calculated according to a standard procedure available on: http://gim.med.ucla.edu/kdqol. Scores range from 0 to 100, with a higher score indicating better QOL.
Translation Procedure The study protocol was approved by R. Hays and the local ethical committee of Ibn Sina University Hospital Center. The translation steps were carried out according to specifications established by the KDQOL Working Group. Forward Adaptation. 1. Items and response choices of the original version of the KDQOL-SF were independently translated into Moroccan by 2 translators whose first language is Moroccan. Following this, a seminar including translators, nephrologists, psychiatrists, and a sociologist was organized. All the options were reviewed, problems with specific terms and differences in the cultures of the countries were resolved, and a consensus version of the questionnaire was made. After this meeting, the translators independently rated the level of difficulty of the translation between 0 (not at all difficult) and 100 (extremely difficult). 2. Two external translators evaluated the equivalence of each item and response scale according to a scale from 0 (“not at all equivalent”) to 100 (“exactly equivalent”). 3. The Moroccan version of the questionnaire was tested in a sample of 10 patients on hemodialysis to detect comprehension problems. Mean patient age was 39 13 years (20e57 years), 6 were males, and 8 had not completed elementary school or were illiterate. The test took the form of face-to-face interviews immediately followed by retrospective interviews about comprehension problems and choices among a number of alternative renderings.
RHOU, EZAITOUNI, OUZEDDOUN ET AL Backward Adaptation. Two backward translations were done by English native translators and a consensus version was compared to the original English version.
Test Field To measure psychometric properties, the Moroccan version was administered to 80 patients on dialysis at Ibn Sina Hospital. Inclusion criteria were (1) age more than 18 years, (2) dialysis treatment for 3 months without life-threatening disease at the time of the test, (3) patient consent. Exclusion criteria were language difficulties. In this cross-sectional study, 78% of patients were in hemodialysis and 22% in peritoneal dialysis (Table 1). The majority were females and mean age was 43.9 14.2 years. The mean time on dialysis among the patients was 104.8 120 months.
Statistical Analysis and Psychometric Properties Means of each scale and standard deviation were determined. Reliability and validity were determined. Percentages of floor and ceiling were also assessed. Ceiling effects were taken as being the percentage of respondents with scores of 100 and floor effects were the percentage of respondents having a score of 0. Ceiling and floor effects should be less than 30% to ensure that the scale captures the full range of potential responses within the population and that changes over time can be detected. Multitrait/Multi-Item Correlations. Correlation coefficients were calculated by using Pearson correlation to assess the strength of relationships between items within and outside each scale. The item internal consistency assesses item correlation with a corresponding scale, corrected for overlap. A correlation corrected for overlap is the correlation of an item with the remaining items of the scale, which removes the bias of correlating an item with itself. A correlation of 0.4 or more has been used as the standard for supporting the item’s internal consistency. The discriminant validity was assessed by comparing item correlation within and outside each scale. We hypothesized that each Table 1. Patient Characteristics Parameters
Patients on Dialysis (n ¼ 80)
Age (y) Sex ratio Peritoneal dialysis (%) Hemodialysis (%) Educational level (%) Illiterate Elementary High school College Civil status: married (%) Yes No Duration on dialysis (mo) Cause of kidney disease (%) Glomerulonephritis Diabetes Tubulointerstitial nephritis Vascular Unknown etiology Other
43.9 14.2 0.7 22 78 28 22 23 27 61 39 104.8 120 26 14 18 5 34 3
KDQOL-SF 1.3 IN AN AFRICAN COUNTRY item’s correlation with its intended scale would be at least two standard errors greater than its correlation with other scales. Internal Consistency Reliability and TesteRetest Reliability. Cronbach a coefficient was used to assess internal consistency reliability. As recommended by Nunnally [14], a Cronbach value of 0.70 was taken as an indicator of adequate internal consistency reliability. Testeretest reliability was assessed by convenience sample (n ¼ 20) of the original respondents. The same questionnaire was administered to these patients after 10 to 14 days. Testeretest reliability was measured using intraclass correlation coefficient. Construct Validity. Construct validity was assessed by comparing scales of KDQOL-SF with overall health rating by means of the Pearson and Spearman correlation coefficient. External Discriminant Validity. External discriminant validity assesses the ability of the KDQOL-SF subscales to differentiate between patient groups with different clinical characteristics. We used the t test to examine the differences of QOL scores in subgroups of patients based on sex, educational level, marital status. We also looked at the correlations of scales scores with age, and dialysis duration by using Pearson and Spearman correlations. The SPSS 13 statistical package for Windows (SPSS, Chicago, Ill, United States) was used for statistical analysis.
RESULTS Translation Procedure
Forward Translation. Difficulty mean rates ranged from 0 to 60 with an average difficulty rating of 13.1 3.5 (results not shown). Items and response choice equivalence between the Arabic and the US version ranged between 75 and 100, and the mean equivalence rating was 93.3. Cultural adaptation concerned some activities rather uncommon in Morocco like “pushing a vacuum cleaner,” “playing golf,” and “bowling” (item SF3b), which needed to be replaced by activities requiring the use of similar body parts. Changes also involved distances that were rendered in meters rather than blocks and miles (items SF3g, SF3h, SF3i). Lexical problems were met with the following words: Physical health as the word health implies physical health in Moroccan. It was finally rendered by your health (ie, of your body). Activities rendered by matters, things to do. It was preferred to a word-for-word translation, as the seemingly equivalent word is too formal. Dissatisfied changed to not at all satisfied and then somewhat dissatisfied was rendered by not very satisfied. Definitely true and definitely false, respectively, rendered by 100% true and 100% false. Syntactic changes were often necessary when there were differences in structure between the English and the Arabic sentence. Few words were changed during the pilot test and some explanations were added in parenthesis (ie, “You respond after a delay to things that were said or done around you” was added in parentheses in item SF13b). The patients were a bit puzzled by the item SF11b where they were asked if their health was “excellent.” The interviewers explained that they were asked about their subjective point of view and not about their medical state.
1297
Backward Translation. The backward translation caused few problems. A few words were corrected before the tests. Field Test
Completeness of Data. Completeness at the scale level was of 100% except for portions on dialysis staff encouragement (99%) and sexual function (40%). Floor and Ceiling. The scales work status, role limitations caused by physical problems, and role limitations caused by emotional problems suffered from high percentage of floor scores, whereas social support dialysis and staff encouragement had a high percentage of ceiling scores (Table 2). Multitrait/Multi-item Correlations, Internal Consistency Reliability, and TesteRetest Reliability. The item internal consistency, corrected for overlap, was above 0.4 for 94% of items. Correlation of items within subscales was higher than that of items outside subscales in 87% of cases. Internal consistency reliability was adequate, with Cronbach a values above 0.70 for all scales except 3 dimensions: quality of social interaction (0.67), work status (0.38), and cognitive function (0.57; Table 2). Cronbach a did not improve after removing 1 item from quality of social interaction and cognitive function scales. No item could be removed from work status scale as it only consisted of 2 items. The intraclass coefficient correlation of the kidney diseaseetargeted scales ranged from 0.67 (pain) to 0.9 (energy/fatigue, symptom/problems and overall health rating; Table 2). Construct Validity. Health rating was significantly correlated with all KDQOL-SF scales except sexual function and dialysis staff encouragement and patient satisfaction scales (Table 3). External Discriminating Validity. Table 4 shows the correlation coefficients between the scales of the KDQOL-SF and age and dialysis duration. There are significant correlations between age and 5 subscales of the KDQOLSF: physical functioning (r ¼ 0.27; P ¼ .006), energy/ fatigue (r ¼ 0.33; P ¼ .001), work status (r ¼ 0.23; P ¼ .023), burden of kidney disease (r ¼ 0.22; P ¼ .037), overall health rating (r ¼ 0.23; P ¼ .024). Duration of dialysis correlated significantly with 6 scales of the KDQOL-SF: pain (r ¼ 0.27; P ¼ .008), energy/fatigue (r ¼ 0.26; P ¼ .01), work status (r ¼ 0.29; P ¼ .004), sexual function (r ¼ 0.28; P ¼ .09), sleep (r ¼ 0.26; P ¼ .001), and overall health rating (r ¼ 0.23; P ¼ .022). We also compared mean KDQOL-SF scores in groups expected to differ in their QOL (Fig 1). Regarding gender, men scored significantly higher than women on symptom/ problem, cognitive function, physical functioning, social functioning, and pain (Fig 1A). Literate patients compared with illiterate patients showed significant higher scores on 5 scales: burden of kidney disease, work status, role limitations caused by physical problems, role limitations caused by emotional problems, and energy/fatigue (Fig 1B). Illiterate patients showed significant higher scores on dialysis staff encouragement and patient satisfaction. Regarding marital
1298
RHOU, EZAITOUNI, OUZEDDOUN ET AL
Table 2. Scores Obtained for the Moroccan Version, Percentage of Floor and Ceiling, Internal Consistency, and TesteRetest Reliability Dimensions
No. of Items
Physical functioning Role limitations caused by physical problems Pain General health perceptions Emotional well-being Role limitations caused by emotional problems Social functioning Energy/fatigue Symptom/problem Effects of kidney disease Work status Cognitive function Quality of social interaction Sexual function Sleep Social support Burden of kidney disease Dialysis staff encouragement Patient satisfaction Overall health rating
10 4 2 5 5 3 2 4 12 8 2 3 3 2 4 2 4 2 1 1
Mean SD
Floor (%)
Ceiling (%)
Cronbach a
Intraclass Correlation
2.1 46.3 5.3 1.1 0 35.8 1.1 0 0 11 37.9 0 0 3.2 2.1 0 7.4 2.1 3.2 3.2
6.3 14.7 29.5 0 3.2 25.3 17.9 0 6.3 5.3 18.9 12.6 24.2 16.8 3.2 55.8 3.2 47.4 13.7 9.5
0.86 0.81 0.89 0.71 0.83 0.74 0.7 0.8 0.82 0.8 0.38 0.57 0.67 0.9 0.74 0.7 0.77 0.82 NA NA
0.89 0.81 0.67 0.84 0.87 0.83 0.8 0.9 0.9 0.88 0.82 0.85 0.8 0.81 0.84 0.78 0.88 0.81 0.87 0.9
68 33 67 42 61 43 66 51 66 63 40 73 74 73 64 83 43 85 57 59
23 38 30 23 20 40 25 21 20 22 37 20 22 31 21 24 29 21 24 22
Abbreviations: NA, not applicable for a single-item measure; SD, standard deviation.
status, no significant differences were observed for scores of married and unmarried patients (results not shown). DISCUSSION
To our knowledge, this has been the first study to evaluate the reliability and validity of the KDQOL-SF among dialysis patients in an African country. This questionnaire would then give information about QOL in this continent and allow international comparisons of QOL in dialysis. Table 3. Correlation Coefficients (With P Values) Between the Kidney Disease-Targeted Scales and the Overall Health Rating Scale
Dimensions
Overall Health Rating
P Value
Physical functioning Role limitations caused by physical problems Pain General health perceptions Emotional well-being Role limitations caused by emotional problems Social functioning Energy/fatigue Symptom/problem Effects of kidney disease Work status Cognitive function Quality of social interaction Sexual function Sleep Social support Burden of kidney disease Dialysis staff encouragement Patient satisfaction
0.53 0.45 0.51 0.52 0.29 0.36 0.28 0.49 0.55 0.32 0.38 0.23 0.23 0.16 0.32 0.4 0.36 0.13 0.12
