Validation of an Arabic version of the Diabetes Treatment Satisfaction Questionnaire in Qatar

Validation of an Arabic version of the Diabetes Treatment Satisfaction Questionnaire in Qatar

Accepted Manuscript Title: VALIDATION OF AN ARABIC VERSION OF THE DIABETES TREATMENT SATISFACTION QUESTIONNAIRE IN QATAR Author: Kerry Wilbur Abdulla ...

479KB Sizes 7 Downloads 42 Views

Accepted Manuscript Title: VALIDATION OF AN ARABIC VERSION OF THE DIABETES TREATMENT SATISFACTION QUESTIONNAIRE IN QATAR Author: Kerry Wilbur Abdulla O Al Hammaq PII: DOI: Reference:

S0168-8227(16)00008-5 http://dx.doi.org/doi:10.1016/j.diabres.2015.12.005 DIAB 6521

To appear in:

Diabetes Research and Clinical Practice

Received date: Revised date: Accepted date:

20-9-2015 31-10-2015 28-12-2015

Please cite this article as: K. Wilbur, A.O.A. Hammaq, VALIDATION OF AN ARABIC VERSION OF THE DIABETES TREATMENT SATISFACTION QUESTIONNAIRE IN QATAR, Diabetes Research and Clinical Practice (2016), http://dx.doi.org/10.1016/j.diabres.2015.12.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

VALIDATION OF AN ARABIC VERSION OF THE DIABETES TREATMENT SATISFACTION QUESTIONNAIRE IN QATAR

ip t

Dr. Kerry Wilbura a College of Pharmacy, Qatar University PO Box 2713, Doha, Qatar [email protected]

an

Dr. Kerry Wilbur, Associate Professor College of Pharmacy, Qatar University PO Box 2713 Doha, Qatar T: 974-4403-5581 F: 974-4403-5551 (GMT+3 hrs)

M

Corresponding Author:

us

[email protected]

cr

Dr. Abdulla O Al Hammaqb Manager, IDF/MENA Region b Executive Director, Qatar Diabetes Association Rawdat Al Khalil, Al Mumtazah, Doha, Qatar

pt

ed

This research was made possible by a grant from the Qatar National Research Fund, National Priorities Research Program.

Ac ce

Conflicts of Interest: None

Manuscript word count (2,826) + Tables’ word count (682)= 3,363 Abstract word count: 245

Abstract Aims: Several instruments evaluate patient-reported outcomes in diabetes mellitus (DM), but almost none are validated for use in Arabic language. The aim of this study is

Page 1 of 22

to test the psychometric properties and responsiveness of the Arabic version of the Diabetes Treatment Satisfaction Questionnaire (DTSQs) in Qatar.

cr

ip t

Methods: Ambulatory Arabic speaking DM patients were interviewed at two consecutive time points in Doha, Qatar. The 8-item DTSQs was administered in conjunction with the Medical Outcomes Study 36-Item Short-Form Health Survey (SF36) and the World Health Organization Quality of Life Measure (WHOQOL-BREF) to assess convergent validity. Reliability was evaluated by internal consistency and item analysis. Construct validity was evaluated using ‘‘known groups’’ comparisons (including gender, insulin use, and HbA1c). Sensitivity of DTSQs scores to the subject’s metabolic conditions was determined.

M

an

us

Results: One hundred subjects (mean age 50.7) participated. Half (54%) were female. The majority (93%) had Type 2 DM, but 39 (42%) were using insulin. Results revealed satisfactory internal consistency. Metabolic measures (fasting blood glucose and AIC) had significant inverse correlations with DTSQs scores (interview 1, Pearson’s r= -0.333 and r= -0.401, respectively, p<0.01). Scale criterion and construct validity were found to be satisfactory. Most sub-dimensions of the SF-36 and WHOQOL-BREF were correlated with the DTSQ, indicating a good concurrent validity. As in prior studies, women demonstrated poorer treatment satisfaction.

pt

diabetes treatment; satisfaction; validation; Arabic

Ac ce

Key words:

ed

Conclusions: The Qatar Arabic DTSQs version was found to be a reliable and valid instrument for the assessment of treatment satisfaction in Arabic diabetes mellitus patients in the country.

Introduction

The Middle East is currently experiencing a diabetes epidemic.(1) When

considering worldwide data, the highest prevalence of diagnosed diabetes are currently found in Gulf Cooperation Countries (GCC) with estimates reaching 24%, 23%, and 20% of the population in Saudi Arabia, Kuwait, and Qatar, respectively.(2) The high proportion of individuals with impaired glucose tolerance and other associated

Page 2 of 22

modifiable risk factors (central obesity, sedentary behaviour) will only contribute to an increased prevalence of diabetes in the region over the coming years. A major goal for diabetes management is to sustain a high quality-of-life through

ip t

minimization of acute and long-term complications associated with abnormal glucose values.(3) Measuring health status in a population is important for the evaluation of

cr

interventions and the prediction of health and social care needs.(4) While existing

us

performance measures of diabetes care principally focus on quantitative processes and measurements of objectively obtained medical data (e.g. blood pressure, glucose, lipid

an

values), an individual’s perception of their health is becoming one of the most

M

significant qualitative health indicators studied today.(5-7) It is now widely acknowledge that health decisions must take into consideration the patient’s point of view and

ed

perceptions of his or her quality-of-life to complement medical evaluation. A number of tools exist to assess diabetes-specific measures of patient health-related quality of

pt

life.(8) However, most questionnaires available for this purpose are in English. The

Ac ce

development of novel instruments, as well as translation and validation of existing instruments to Arabic is lacking.(9-13) The Diabetes Treatment Satisfaction Questionnaires (DTSQ standard and DTSQ

change) are instruments developed to assess patient satisfaction with diabetes medication and glucose control. Endorsed by the World Health Organization (WHO) and the International Diabetes Federation (IDF), they are widely used in clinical research for assessing outcomes of diabetes care.(14)

Page 3 of 22

These DTSQ instruments have been translated into more than 40 languages, including Arabic for Egypt, Saudi Arabia, the United Arab Emirates, and most recently Qatar.(15) However, translation (and cultural adaptation) alone are insufficient for meaningful use

ip t

of a questionnaire tool. The instrument should be validated in order to determine if the

translation process changed the psychometric properties of the original instrument and

cr

re-estimation of reliability is necessary with each population surveyed.(16) The aim of

us

our study was to evaluate the psychometric properties of the recently adapted Arabic

M

an

version of the Diabetes Treatment Satisfaction Questionnaire (standard).

METHODS

ed

A convenience sample of 100 consenting Arab-speaking adult ambulatory diabetes patients receiving care at the Qatar Primary Health Care Clinics and the Qatar

pt

Diabetes Association and participated in the study. One research assistant interviewed

Ac ce

subjects gathering socio-demographic data (including age, ethnicity, gender, education, marital status and occupation) and health information (diabetes type, duration of disease, therapeutic regimen, most recent fasting blood glucose (FBG) and glycated hemoglobin (HbA1C)) and then administered three questionnaires: 1) The Diabetes Treatment Satisfaction Questionnaire standard (DTSQs) questionnaire in Arabic previously adapted for Qatar whereby each of the 8-items is rated on a sevenpoint Likert scale.(15) Six (items 1, 4-8) of them (item 1: satisfaction with current treatment; item 4: treatment convenience; item 5: flexibility of treatment; item 6:

Page 4 of 22

understanding of diabetes; item 7: continuity of treatment and item 8: recommending treatment to others with diabetes) are summed to get treatment satisfaction score with a possible range of 0 (very dissatisfied) to 36 (very satisfied). Item 2 which

ip t

evaluated perceived frequency of hyperglycemia and item 3 that assesses frequency of hypoglycemia are also rated on a seven point scale (0-6) the same as other items, but

cr

for these two items a score of zero indicates lack of hyper- or hyperglycemia while a

us

higher score indicates a higher frequency.

2) The Arabic World Health Organization Quality of Life Questionnaire (WHOQOL-Bref), a

an

26 question abbreviated version of the WHOQOL-100, based on a four domain

M

structure: physical health, psychological health, social relations, and environmental health. Higher domain scores indicate better quality-of-life.(12)

ed

3) The Arabic Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36), the multi-purpose, short-form health survey yielding an 8-scale profile of functional health

pt

and well-being scores as well as psychometrically-based physical and mental health

Ac ce

summary measures and a preference-based health utility index. Higher domain scores indicate less disability.(17)

Finally, all patients were also asked to rate their satisfaction with their glucose

monitoring services and the overall quality of care according to a Likert scale with five anchor points ranging from “very satisfied” to “very dissatisfied”. These two “benchmark items” were subsequently used as part of the DTSQs validity testing. All subjects were contacted to complete the questionnaires again in 3-4 weeks

Page 5 of 22

(provided no changes in their drug therapy had been made in the interval since the first interview). Ethics approval was obtained from the relevant institutional review boards. Evaluation of the reliability and validity of the DTSQs Qatar Arabic adaptation

ip t

was conducted by confirmatory approach. Reliability was tested by internal

consistency analysis using Cronbach’s alpha with values of 0.70 considered satisfactory Test-retest reliability to determine the instrument’s stability over repeated

cr

(18)

us

administration was assessed comparing baseline and follow-up responses.

Several approaches were employed to evaluate the validity of the DTSQs Qatar

an

Arabic adaptation. The WHOQOL-Bref and SF-36 were administered to patients in

M

order to assess convergent validity, measuring the extent to which satisfaction is related to particular scores on these different instruments: the DTSQs overall scores are

ed

expected to significantly correlate with the SF-36 scale summary measure of physical and emotional health due the chronic nature and wide range of disability and

pt

discomfort associated with diabetes. Similar significant correlations are expected for

Ac ce

associated WHOQOL-Bref domains as well. Construct validity was further assessed to demonstrate whether the instrument

could discriminate between individuals who are known to have low satisfaction with diabetes treatment. Patient gender, age, type of diabetes, duration of diabetes, type of therapy, existence of any diabetic complication or co-morbidities were used for the known-groups validity testing. Older age, females, insulin treatment, elevated HbA1C, and having diabetes complications are patient features previously reported to be associated with poorer DTSQs scores than their counterparts.

Page 6 of 22

Criterion validity, the relationship between the measurement score and the principle, was tested with the two benchmark questions asked to respondents in order to know their self-evaluations (satisfaction) on the quality of overall monitoring of

ip t

diabetes and on the quality of monitoring of blood glucose. Correlation between the

overall DTSQs score and items 2 and 3 (relating to perceived hypo- and hyperglycemia)

cr

with the two parallel instruments were measured. Responsiveness to change of DTSQs

us

scores to fasting blood glucose and HbA1C values at baseline and follow-up were assessed. Regression analysis was used to explore the independent relationships

an

between DTSQs scores and known subgroups as well as metabolic control indices.

M

Confirmatory factor analysis (CFA) was conducted in R Studio software to further assess the ability of the DTSQs to measure the satisfaction construct.

ed

Categorical demographic data were calculated as percentages of frequency and continuous data reported as means with standard deviations. Non-parametric

pt

correlation (Spearman’s Rho) and Mann-Whitney tests were used in the univariate

Ac ce

comparisons. Multilinear regression techniques were used to determine how specific independent demographic and clinical variables were associated with overall treatment satisfaction (dependent variable). Standard version release 21.0 (SPSS Inc., Chicago, Illinois) was used for analyses. RESULTS

Respondent characteristics are summarized in Table 1. One hundred subjects 23 to 76 years old (mean age 50.7) participated and just over half (54%) were female. The majority (93%) were diagnosed with Type 2 diabetes mellitus (DM) but 39 (42%) of

Page 7 of 22

these subjects were using insulin. One-third of T2DM patients were using at least 2 oral drug therapies. The average duration of since diabetes diagnosis was almost one decade (9.8 years) and recent measure of glucose control was elevated according to

ip t

international standards (e.g. AIC < 7%)(3) Few (13%) reported history of macrovascular

complications (the most frequent being problems with circulation (peripheral vascular

cr

disease), but more than one-third indicated presence of microvascular disorders, with

us

30% specifically complaining of neuropathic pain. More than half (64%) possessed comorbid conditions which were mostly cardiovascular in nature (hypertension and

an

dyslipidemia).

M

Each subject completed repeat questionnaires on average one month following initial enrolment (mean time interval 31.3 days (SD 10)). Baseline and follow-up mean

ed

DTSQs item scores are depicted in Figure 1. Floor effects (defined as scores < 1 for item 2 and item 3) and ceiling effects (defined as DTSQs total > 30) were observed (20-45%

pt

and 33%, respectively at baseline and 17-56% and 28% at follow-up).(19) Item responses

Ac ce

were stable across time, with the exception of improvements in perceived hypoglycemia (item 3, 1.31 vs 1.01, p=0.003) and decreased recommendation of their therapy to others (item 7, 3.01 vs 2.42, p=0.001) between interviews. All baseline and most repeat DTSQs scoring items (1, 4-8) were found to be significantly correlated with the overall DTSQs score (Table 2). Greater treatment satisfaction was associated with less perceived frequency hyperglycemia (item 2) both at baseline and follow-up (r= 0.304 and r= -0.323, p<0.001), but not with fewer self-rated incidents of unacceptably low glucose (item 3). Perceived frequency of hypo- or hyperglycemia had almost no

Page 8 of 22

relationship with the individual scale items at either interview. Response results revealed satisfactory internal consistency for DTSQs scores at baseline with some loss at follow-up (Cronbach alpha’s <0.70) (Table 2). Factor loadings determined from the

ip t

CFA exceeded 0.50 and differed reliably from zero (p<0.05) (Table 2). Goodness of fit

indices included Chi square/degrees of freedom (Χ2/df)= 2.12, root mean square error of

cr

approximation (RMSEA) = 0.04, and CFI=0.955 and indicate satisfactory model fit.

us

Measures of construct validity were also found to be satisfactory. All sub-

dimensions WHOQOL-Bref were positively correlated with overall DTSQs scores,

an

indicating good convergent validity (Table 3). Unlike social functioning, physical

M

features in the SF-36 did not seem related to treatment satisfaction scores in our sample, and subject perceptions of hyperglycemia and hypoglycemia frequency were

ed

found to correlate with features of emotional well-being at baseline and follow-up. Hyperglycemia and overall treatment satisfaction were linked to SF-36 component

pt

scores of pain.

Ac ce

Total DTSQs scores did not differ among various subgroups between baseline and follow-up interviews (Table 4). We found no statistical difference in satisfaction between Type 1 or Type 2 diabetes patients or between diabetes patients receiving insulin treatment or not. Scores among those of older age were lower, but the difference not statistically significant. However, women in our sample were less satisfied with their treatment than men, as were patients who had reported evidence of any diabetes complications (at follow-up only). Dissatisfaction with overall quality of

Page 9 of 22

diabetes care and glucose monitoring corresponded with lower overall treatment satisfaction scores. As expected, metabolic measures (fasting blood glucose and AIC) had significant

ip t

inverse correlations with DTSQs scores (baseline r= -0.333 and r= -0.401, respectively, p<0.01) and repeat, r= -0.252 and r= -0.308, respectively, p<0.05) (Table 3). When

cr

overall DTSQs scores are considered, female gender and high baseline HbA1C values

us

are identified as the significant demographic and clinical predictors of poor satisfaction in multivariate regression analysis with the model explaining 21.5% of the variance.

an

DISCUSSION

M

Diabetes is a chronic disease associated with many long-term micro- and macrovascular complications, but for most patients, the practical daily management

ed

evolves around glucose control. Studies have shown how these treatment choices can directly impact patient satisfaction as it pertains to their dosing and administration,

pt

side effects, perceptions of hyper- or hypoglycemia and need for self-monitoring of

Ac ce

blood glucose.(20, 21) However, few of these findings are from Middle East populations.(22, 23) Given the prevalence of diabetes in this region, it is important to have a tool to assess local diabetes patient treatment satisfaction whereby findings can contribute to and be compared with worldwide data from many prior studies, as well as ensure versions designed elsewhere are not inappropriately used.(24) We found the Arabic version of the DTSQs adapted for Qatar to be a valid and reliable instrument to assess treatment satisfaction of diabetes patients in the country. Like other diabetes patient assessments using a DTSQs instrument, we also found evidence of a ceiling

Page 10 of 22

effect.(25, 26) Such observed patient selection of maximum possible item scores on the questionnaire prompted the development of the DTSQc (change) instrument.(27) The change version poses the same 8 questions as the DTSQs, but reworded to assess

ip t

satisfaction relative to the patient’s experience a few weeks or months ago. The DTSQc scoring scale is altered to -3 to +3 from the 0 to 6 in the DTSQs questionnaire. We did

cr

not opt to pursue validation of this companion instrument – which has also been

us

adapted for Qatar Arabic - as we were seeking to evaluate the test-retest reliability of the DTSQs in patients whose drug therapy had not changed over the one month

an

interval between interviews.

M

The Qatar diabetes patients we surveyed were largely overall satisfied with their treatment, with specific findings varying little over time (@ 30 day follow-up interview).

ed

Internal consistency of the DTSQs instrument was satisfactory given the alpha levels found. Incidents of perceived hypoglycemia did not seem to be problematic and unlike

pt

hyperglycemia, was not correlated with worse overall DTSQs scores. This may not be

Ac ce

surprising as many of our subjects had T2DM and although many were taking insulin, overall reported HbA1C and FBG values were high. Such disparity has also been observed in the Turkish and Italian DTSQs validation studies.(25, 28) Criterion validity of our instrument was demonstrated according to the association of higher DTSQs scores among diabetes patients who reported satisfaction with the quality of their overall disease monitoring (statistically significant findings) or quality of glucose monitoring which is consist with findings among diabetes patients elsewhere.(23, 25, 28)

Page 11 of 22

Significant correlations between overall DTSQs scores and WHO-QOL Bref subdimensions were found at baseline and at follow-up. Association between treatment satisfaction and pain but not physical function or physical limits domain scores of the

ip t

SF-36 was an unexpected finding but we do not believe this indicates failure in

convergent validity. Unlike other DTSQs validity assessments, our study population

cr

had higher prevalence of neuropathy as microvascular complications. Poor self-rated

us

physical and mental scores have been previously associated with lower treatment satisfaction, but pain and mental health status are also inextricably linked to treatment

an

satisfaction and quality of life.(25, 28-30) Physical pain may restrict activities of daily living

M

and interfere with social engagement, which may compound existing intrusive effects for some individuals living with diabetes.(31)

ed

Age and time since diabetes diagnosis were not related to treatment satisfaction in our cohort. Poor glucose control (as indicated by high fasting blood glucose and

pt

HbA1C were significant predictors of treatment satisfaction in univariate analysis, this

Ac ce

effect was lost when gender was entered into the multi-linear regression model. Many other studies have like ours found poorer treatment satisfaction among women, but the effect of gender is inconsistent as other researchers have not found any relationship.(25, 28, 29, 32, 33) In our cohort, women reported more limits of their diabetes to their emotional and physical functioning (SF-36) than men, which may aversely influence their opinions related to their therapy and glucose control. Documented comorbidities did not demonstrably influence overall DTSQs scores. Although they were present in over half of our cohort, the nature of the specific reported diseases in

Page 12 of 22

question (hypertension, dyslipidemia) is such that symptoms are largely not discernable to the patient. The use of a research assistant to administer the questionnaire may be

ip t

considered a limitation to the study if it were felt that subjects would offer favourable responses when compared to answers given during self-administration (acquiescence

cr

bias). However, as we have found in our earlier work administering diabetes

us

questionnaires to patients, many need additional orientation to item scoring scales, which would also assist avoidance of “extreme responses”. Others have reported such

an

challenges assessing a DTSQs instrument within an Arab context.(34) Additionally, one-

M

fifth of our population had not received education past the primary school level necessitating research assistant support for questionnaire completion.(35) Although one

ed

of our comparator instruments itself has not been validated in Qatar Arabic (WHO-QoL Bref) and we used the Saudi Arabic version as other researchers in Kuwait have

pt

employed in order to complement the comparisons with the Qatar Arabic SF-36.(36-38)

Ac ce

We did not gather information about patient weight to evaluate effects on treatment satisfaction. While obesity is prevalent in Qatar, other study in populations with above normal body mass index has failed to find an association.(29, 32, 33) Like other validation studies, our sample is largely of subjects with Type 2 diabetes and reflects the make-up of diabetes diagnosis in the country. Although nearly one-third of these subjects were receiving insulin therapy, it is possible that treatment satisfaction findings among Type 1 diabetes patients and subsequent relationship with our comparator questionnaires might be different.

Page 13 of 22

CONCLUSION In a sample of Qatar diabetes patients in ambulatory care, treatment satisfaction as scored by the DTSQs was tested against dimensions of health measured

ip t

by the WHOQOL-Bref and SF-36 across two separate time periods and demonstrated favourable psychometric properties. The Qatar Arabic adaptation of DTSQs may be

cr

considered a culturally relevant and validated instrument for use by clinicians and

us

researchers to evaluate diabetes treatment satisfaction in the Arabic-speaking

M

an

population in Qatar.

2.

3.

Klautzer L, Becker J, Mattke S. The curse of wealth - Middle Eastern countries need to address the rapidly rising burden of diabetes. Int J Health Policy Manag 2014;2:109-14.

Ac ce

1.

pt

REFERENCES

ed

ACKNOWLEDGMENTS: The authors wish to acknowledge research assistant, Ms. Reem Dajani, staff at Qatar Diabetes Association and the Primary Health Care Corporation, including Director Dr. Maryam Abdulmalik.

International Diabetes Federation. IDF Diabetes Atlas update poster, 6th edition. Brussels, Belgium: International Diabetes Federation 2014 [cited 2015 March 2 ]. Available from URL: http://www.idf.org/diabetesatlas.

Association AD. Standards of Medical Care in Diabetes - 2015. Diabetes Care 2015;38(Supplement 1).

4. Sabbah I, Drouby N, Sabbah S. Quality of life in rural and urban populations in Lebanon using SF-36 Health Survey. Health Qual Life Outcomes 2003;1:30-44. 5.

Harper J, Gunning K. Standards, safety and quality – where next? JICS 2008;9:114-5.

Page 14 of 22

6. Wens J, Driven K, Mathieu C, Paulus D, Van Royen P. Quality indicators for type-2diabetes care in practice guidelines: an example from six European countries. Prim Care Diabetes 2007;1:17-23. Martirosyan L, Braspenning J, Denig P, de Grauw WJC, Bouma M, Storms F, et al. Prescribing quality indicators of type 2 diabetes mellitus in ambulatory care. Qual Saf Health Care 2008;17:318-23.

ip t

7.

cr

8. Garratt AM, Schmidt L, Fitzpatrick R. Patient-reported health outcome measures for diabetes: a structured review. Diabet Med 2002;19:1-11.

us

9. Anderson RT, Aaronson NK, Bullinger M, McBee WI. A review of the progress towards developing health-related quality of life instruments for international clinical studies and outcomes research. Pharmacoeconomics 1996;10:336-45.

M

an

10. Anderson RT, Aaronson NK, LePlage AP, Wilkin D. In: Spiker B, editor. Quality of Life and Pharmacoeconomics in Clinical Trials. 2nd ed. Philadelphia: LippincottRaven Publishers; 1996. p. 613-32. 11. Aburuz S, Bulatova N, Twalbeh M, Gasawi M. The validity and reliability of the Arabic version of the EQ-5D: a study from Jordan. Ann Saudi Med 2009;29:304-8.

ed

12. Chaeri JU, Awadalla AW. The reliability and validity of the short version of the WHO Quality of Life Instrument in an Arab general population. Ann Saudi Med 2009;29:98-104.

pt

13. Al Robaee AA. Assessment of general health and quality of life in patients with acne using a validated generic questionnaire. Acta Dermatoven 2009;8:157-64.

Ac ce

14. Bradley C, Gansu DS. Guidelines for encouraging psychological well-being: report of a working group of the World Health Organization Regional Office for Europe and International Diabetes Federation Europe Region St Vincent Declaration Action Programme for Diabetes. Diabet Med 1994;11:510-57. 15. Wilbur K. Diabetes Treatment Satisfaction Questionnaire - an Arabic adaptation for Qatar. Diabetes Res Clin Pract 2013;99:e24-6. 16. Polit DF, Beck CT. Nursing research: Principles and methods. Philadelphia: Lippincott, Williams & Wilkins; 2004. 17. Al Abdulmohsin SA, Coons SJ, Draugalis JR, Hays RD. Translation of the RAND 36Item Health Survey 1.0 (aka SF-36) into Arabic. 1997. 18. Bland JM, Altman DG. Statistics Notes. Cronbach's alpha. BMJ 1997;514:572.

Page 15 of 22

19. Bradley C, Plowright R, Stewart J, Valentine J, Witthaus E. The Diabetes Treatment Satisfaction Questionnaire change version (DTSQc) evaluated in insulin glargine trials shows greater responsiveness to improvements than the original DTSQ. Health Qual Life Outcomes 2007;5:57.

ip t

20. Alcubierre N, Rubinat E, Traveset A, Martinez-Alonso M, Hernandez M, Jurjo C, et al. A prospective cross-sectional study on quality of life and treatment satisfaction in type 2 diabetic patients with retinopathy without other major late diabetic complications. Health Qual Life Outcomes 2014;12:131-41.

us

cr

21. Ashwell SG, Bradley C, Stephens JW, Witthaus E, Home PD. Treatment satisfactoin and quality of life with insulin glargine plus insulin lispro compared with NPH insulin plus unmodified human insulin in individuals with Type 1 Diabetes. Diabetes Care 2008;31:1112-7.

an

22. Al Shahrani A, Baraja M. Patient satisfaction and it's relatin to diabetic control in a primary care setting. J Family Med Prim Care 2014;3:5-11.

M

23. Biderman A, Noff E, Harris SB, Friedman N, Levy A. Treatment satisfaction of diabetic patients: what are the contributing factors? Fam Pract 2009;26:102-8.

ed

24. Bener A, A.O. A-H, Yousafzai MT, Abdul-Ghani M. Relationship between patient satisfactions with diabetes care and treatment. Niger J Clin Pract 2014;17:218-25.

pt

25. Ozmen B, Eser E, Ozkaya Kafesciler S, Pala T, Guclu F, Hekimsoy Z. Psychometric properties and responsiveness of the Turkish version of the Diabetes Treatment Satisfaction Questionnaire (s) on a sample of diabetics of three consecutive monitoring periods. Acta Diabetol 2010;47 Suppl 1:123-31.

Ac ce

26. Power F, Snoek FJ, Heine RJ. Ceiling effect reduces the validity of the diabetes treatment satisfaction questionnaire. Diabetes Care1998;21:2039. 27. Bradley C. Diabetes Treatment Satisfaction Questionnaire: Change version for use alongside status version provides appropriate solution where ceiling effects occur. Diabetes Care 1999;22:530-2. 28. Nicolucci A, Giorgiono R, Cucinotta D, Zoppini G, Muggeo M, Squatrtio S, et al. Validation of the Italian version of the WHO-Well-Being Questionnaire (WHOWBQ) and the WHO-Diabetes Treatment Satisfaction Questionnaire (WHO-DTSQ). Diabetes Nutr Metab 2004;17:235-43. 29. Wredling R, Stalhammer J, Adamson U, Berne C, Larsson Y, Ostman J. Well-being and treatment satisfaction in adults with diabetes: a Swedish population-based study. Qual Life Res 1995;4:515-22.

Page 16 of 22

ip t

30. McCarberg BH, Nicholson BD, Todd KH, Palmer T, Penles L. The impact of pain on quality of life and the unmet needs of pain management: results from pain sufferers and physicians participating in an Internet survey. Am J Ther 2008;15:31220. 31. Javed S, Maqsood A. Diabetes associated distress: Implications for coping and treatment. Applied Psychology 2015;12:1-19.

us

cr

32. Kontodimopoulos N, Arvanitaki E, Aletras VH, Niakas D. Psychometric properties of the Greek Diabetes Treatment Satisfaction Questionnaire. Health Qual Life Outcomes 2012;10-17.

an

33. Petterson T, Lee A, Hollis S, Young B, Newton B, Donran T. Well-being and treatment satisfaction in older people with diabetes. Diabetes Care 1998;21:930-5.

M

34. Dawsey R, Sweeney E, Plowright R, Wilson A, Bradley C. Linguistic validation of the DTSQ: Challenges with arabic and french for Algeria. ISPOR 19th Annual International Meeting Research Abstracts: Value in Health; 2014. p. A196-7.

ed

35. Meisenberg G, Williams A. Are acquiescent and extreme response styles related to low intelligence and education? Pers Individ Dif 2008;44:1539-50.

pt

36. Ohaeri JU, Awadalla AW. The reliability and validity of the short version of the WHO Quality of Life Instrument in an Arab general population. Ann Saudi Med 2009;29:98-104.

Ac ce

37. Alshubaili AJ, Ohaeri JU, Awadaala AW, Mabrouk AA. Family caregiver quality of life in multiple scelorosis among Kuwaitis: a controlled study. Saudi Med J 2009;30:1328-35. 38. Ohaeri JU, Awadalla AW. Characteristics of subjects with comordity of symptoms of generalized anxiety and major depressive disorders and the corresponding threshold and subthreshold condittions in an Arab general population. Med Sci Monit 2012;18:CR160-CR173.

Page 17 of 22

Table 1. Patient Demographics, N=100 50.7 (10.2) 54

cr

ip t

16 29 11 29 26.1 (17.9)

an

us

85 5 5 5

M

Age, years (mean ± SD) Female Nationality* Qatari Egyptian Sudanese Other Arab Residing in Qatar, years (mean ± SD) Marital Status Married Single Divorced or Separated Widowed Highest Education none primary school highschool diploma university Type 2 diabetes (T2DM)

ed

Duration of diagnosis, years (mean ± SD)

Ac ce

pt

Treatment Metformin Combination oral therapy T2DM receiving insulin Insulin monotherapy Last documented AIC, % (mean ± SD) Last documented FPG, mmol/L, (mean ± SD) Co-morbid conditions Macrovascular complications Microvascular complications Hypertension Dyslipidemia

9 12 25 11 43 93

9.8 (7.7) 78 27 27 18 7.5 (1.3) 8.6 (3.8) 13 37 31 28

* 15 did not give nationality; AIC=hemoglobin A1C; FPG= fasting blood glucose Macrovascular= self-reported MI/angina, stroke/TIA, peripheral vascular disease Microvascular= self-reported neuropathy, nephropathy, retinopathy

Page 18 of 22

ip t cr

6 7 8 Total DTSQs score

0.001 -0.471** -0.381** -0.399 -0.216** -0.221** -0.32 -0.34 0.065 -0.083 -0.335** -0.419 -0.304** -0.323**

0.631** -0.490 -0.08 -0.078 -0.08 -0.078 -0.136 -0.21 0.90 0.78 -0.126 -0.011 -0.57 0.091

an

Item 3

M

5

4.67 ± 1.5 4.81 ± 1.3 4.58 ± 1.6 4.99 ± 4.0 4.42 ± 1.6 4.48 ± 1.5 4.63 ± 1.5 4.48 ± 1.5 3.01 ± 2.3 2.42 ± 2.4 4.54 ± 1.7 4.58 ± 1.7 25.8 ± 7.3 25.7 ± 7.7

Item 2

d

4

Baseline Follow-Up Baseline Follow-Up Baseline Follow-Up Baseline Follow-Up Baseline Follow-Up Baseline Follow-Up Baseline Follow-Up

Mean ± SD

ep te

1

Response

Ac c

Item number

us

Table 2. Item Descriptive Results, Item-Scale Correlations and Internal Consistency

Total DTSQs score 0.731** 0.740** 0.769** 0.814 0.769** 0.814 0.754** 0.731** 0.586** 0.580** 0.833** 0.720** 1.00 1.00

Item deleted 0.745 0.694 0.746 0.688 0.742 0.687 0.763 0.702 0.752 0.708 0.729 0.690 0.770 0.722

Item factor loading 0.736 0.752 0.790 0.503 0.629 0.098 1.00 1.00

** p<0.001  Correlation (Spearman’s Rho) of Items 2 and 3 (frequency of hypoglycemia and hyperglycemia, respectively) and total DTSQs score with items 1, 4-8.  Cronbach alpha values of scale internal consistency if items 1, 4-8 were singularly deleted  Confirmatory factor analysis correlation coefficients between treatment satisfaction and scale items

Page 19 of 22

ip t cr

0.187* 0.064 0.494**

-0.122 -0.022 0.30**

-0.102 0.105 0.287**

-0.288** 0.126 0.528**

-0.115 0.090 0.33**

-0.227* 0.226* 0.289*

SF

EF

EW

PF

EL

PL

Pain

0.193* 0.064 0.087

0.001 0.044 0.325**

0.165 0.058 -0.40

0.252** 0.195* 0.004

-0.151 -0.113 -0.071

-0.167 -0.90 0.022

-0.151* 0.195 0.004

-0.332** 0.075 0.375**

0.087 0.163 0.055

-0.129 0.279** 0.289**

0.079 0.031 -0.127

0.258** 0.230** 0.064

-0.076 -0.146 -0.33

-0.127 -0.067 0.076

0.258** 0.230** 0.064

-0.297** 0.036 0.409**

Fasting Blood Glucose

HbA1C

-0.347** -0.252*

-0.414** 0.308*

Ac c

Baseline Follow-up

-0.179* 0.068 0.451**

M

Follow-up item 2 item 3 Overall

Enviro

d

-0.206* 0.50 0.410**

SR

ep te

DTSQs Baseline item 2 item 3 Overall

SF-36 GH

an

WHOQOL-Bref Phys Psych

us

Table 3. Relationship Between DTSQs Scores and Specific WHOQOL-Bref , SF-36 Domain Scores, and Metabolic Parameter at Baseline and Follow-Up Interviews

PHYS – Physical, PSYCH – Psychological, SR – Social Relationships, Enviro – Environmental, GH – General Health, SF – Social Functioning, EF – Energy/Fatigue, EW – Emotional Wellbeing, PF – Physical Function, EL -Emotional Limits, PL – Physical Limits, * p<0.05; ** p<0.001

Page 20 of 22

Gender (n) Female (54) Male (46)

Baseline Mean ± SD

Follow-up Mean ± SD

24.2 ± 6.8* 27.8 ± 7.4

ip t

Table 4. Relationships between Total DTSQs Score and Specific Diabetes Patient Subgroups

24.6 ± 8.1* 27.0 ± 7.1

Ac ce

pt

ed

M

an

us

cr

Age, years (n) 24.4 ± 7.0 >55 (36) 25.1 ± 8.0 27.0 ± 8.2 26.9 ± 6.8 40-54 (46) 25.7 ± 6.9 25.2 ± 6.2 <40 (17) Diabetes Type (n) Type 1 (7) 24.7 ± 6.7 24.1 ± 7.7 25.9 ± 7.3 25.8 ± 7.7 Type 2 (93) Diabetes Duration, years (n) 26.6 ± 9.3 >15 (24) 26.1 ± 7.7 24.8 ± 6.2 25.0 ± 7.0 5-15 (36) 26.1 ± 8.0 26.5 ± 7.4 < 5 (39) Therapy (n) Oral only (54) 27.1 ± 7.1 26.9 ± 8.4 24.4 ± 7.2 24.2 ± 6.5 Insulin (46) Diabetes Complications (n) Present (38) 24.5 ± 7.2 23.4 ± 6.6* 26.7 ± 7.2 27.1 ± 8.0 Absent (62) Co-morbidities (n) Present (64) 24.9 ± 7.5 25.1 ± 8.7 27.4 ± 6.5 26.7 ± 5.6 Absent (36) Satisfaction with quality of diabetes monitoring 26.8 ± 6.6* 27.3 ± 6.8* Very/Satisfied (81) 19.4 ± 7.8 17.6 ± 7.3 Very/Dissatisfied (9) Satisfaction with quality of glucose monitoring Very/Satisfied (86) 26.4 ± 6.9 26.3 ± 6.5 25.2 ± 8.1 23.7 ± 7.8 Very/Dissatisfied (8) No statistically significant differences found across subgroups between baseline and follow-up *subgroup differences p<0.05

21

Page 21 of 22

Figure 1. DTSQs Item Scores at Baseline and Follow-Up

Baseline

Repeat

ip t

6

cr

5

4

us

3

an

2

*

M

1

0 1

2

*

3

4

5

6

7

8

Ac ce p

te

* p<0.05

d

DTSQs Item

Page 22 of 22