SMS delivered messages to family members to improve social support

Calculated from Weight and Height. As a measure of adiposity, it correlates positively with cardiovascular disease outcomes. While imperfect, it is an easily measured and validated marker.

Calculated from Weight and Height. As a measure of adiposity, it correlates positively with cardiovascular disease outcomes. While imperfect, it is an easily measured and validated marker.

A measure of central adiposity, it correlates with cardiovascular outcomes.

A measure of central adiposity, it correlates with cardiovascular outcomes.

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the systolic blood pressure for that visit. Systolic blood pressure is associated with cardiovascular complications

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the systolic blood pressure for that visit. Systolic blood pressure is associated with cardiovascular complications

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the diastolic blood pressure for that visit.

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the diastolic blood pressure for that visit.

Summary of Diabetes Self-care Activities (Toobert, Hampson, Glasgow, & RE, 2000). The Summary consists of 6 subscales representing different domains of diabetes related healthy behaviors and self-care. It has been validated in over 10 studies, with the results published in two manuscripts: one with three studies (Toobert & Glasgow, 1994), and one with seven studies(Toobert et al., 2000). Each measure ranges from Each range from 0-7, indicating number of days per week patient reports engaging in these behaviors.

Summary of Diabetes Self-care Activities (Toobert, Hampson, Glasgow, & RE, 2000). The Summary consists of 6 subscales representing different domains of diabetes related healthy behaviors and self-care. It has been validated in over 10 studies, with the results published in two manuscripts: one with three studies (Toobert & Glasgow, 1994), and one with seven studies(Toobert et al., 2000). Each measure ranges from Each range from 0-7, indicating number of days per week patient reports engaging in these behaviors.

Medication adherence(Wilson et al.,2014). The 3item medication adherence scale is a self-report measure that assessing no. of days medication missed in prior month, days taken medication as advised & self-rated medication adherence. Developed in HIV research, it has been validated for non-HIV medications (Wilson, et al,2016). Each component contributes 33% of the scale. Total score ranges from 0-100. It is the combination of 3 sub-scores, (days missed dose(negative scored), self-rating, days taking medications as indicated). Higher scores indicate better medication adherence.

Medication adherence(Wilson et al.,2014). The 3item medication adherence scale is a self-report measure that assessing no. of days medication missed in prior month, days taken medication as advised & self-rated medication adherence. Developed in HIV research, it has been validated for non-HIV medications (Wilson, et al,2016). Each component contributes 33% of the scale. Total score ranges from 0-100. It is the combination of 3 sub-scores, (days missed dose(negative scored), self-rating, days taking medications as indicated). Higher scores indicate better medication adherence.

Healthcare utilization by EMR review of clinic appointments, ED visits and hospitalizations within enrollment through 6 months at each follow up visit.

Healthcare utilization by EMR review of clinic appointments, ED visits and hospitalizations within enrollment through 6 months at each follow up visit.

Change in Self-efficacy. Diabetes Empowerment Scale Short Form (Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), is a measure of a patient's overall diabetes related self-efficacy, shortened by the original from 37 to an 8 item scale. It has reliability of alpha=0.84; and has been shown to be sensitive to change in multiple populations and is correlated with improved glycemic control measured by A1C (Anderson et al., 2003). It ranges from 8-40 points; higher score indicates higher self-efficacy.

Change in Self-efficacy. Diabetes Empowerment Scale Short Form (Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), is a measure of a patient's overall diabetes related self-efficacy, shortened by the original from 37 to an 8 item scale. It has reliability of alpha=0.84; and has been shown to be sensitive to change in multiple populations and is correlated with improved glycemic control measured by A1C (Anderson et al., 2003). It ranges from 8-40 points; higher score indicates higher self-efficacy.

Diabetes related distress (Diabetes Distress Scale) (Polonsky et al., 2005), is a 17 item scale measuring distress related to emotional burden, physician-related distress, regimen-related distress, and diabetes-related interpersonal issues. It has a Cronbach's alpha of 0.88-0.93 in various studies. Higher scores are negatively correlated with several healthy behaviors. Each question is a Likert scale ranking of how serious a particular issue is from 1-6, 6 indicating a more significant problem. The scores are then averaged across all item, with final scores ranging from 1-6, with 6 indicating higher levels of distress

Diabetes related distress (Diabetes Distress Scale) (Polonsky et al., 2005), is a 17 item scale measuring distress related to emotional burden, physician-related distress, regimen-related distress, and diabetes-related interpersonal issues. It has a Cronbach's alpha of 0.88-0.93 in various studies. Higher scores are negatively correlated with several healthy behaviors. Each question is a Likert scale ranking of how serious a particular issue is from 1-6, 6 indicating a more significant problem. The scores are then averaged across all item, with final scores ranging from 1-6, with 6 indicating higher levels of distress

Depression as measured by the PHQ-9 (Kroenke, Spitzer, Williams, & Lowe, 2010), is a widely used scale of depression used in clinical practice and research. The Cronbach's alpha ranges from 0.86-0.89, and it has been validated in multiple languages. Depression is worse with higher scores. Score ranges from 0 to 27. Higher levels of depression are associated with fewer healthy behaviors and worse glycemic control.

Depression as measured by the PHQ-9 (Kroenke, Spitzer, Williams, & Lowe, 2010), is a widely used scale of depression used in clinical practice and research. The Cronbach's alpha ranges from 0.86-0.89, and it has been validated in multiple languages. Depression is worse with higher scores. Score ranges from 0 to 27. Higher levels of depression are associated with fewer healthy behaviors and worse glycemic control.

Fatalism is measured by the Diabetes Fatalism Scale (Egede & Ellis, 2010), which measures three subscale emotional distress, religiosity and coping and perceived self-efficacy, which are summed together. The total score ranges from 12-72, higher score indicates higher fatalism, and has a Cronbach's alpha of 0.80. The score is associated with glycemic control, depression, self-care behaviors and social factors.

Fatalism is measured by the Diabetes Fatalism Scale (Egede & Ellis, 2010), which measures three subscale emotional distress, religiosity and coping and perceived self-efficacy, which are summed together. The total score ranges from 12-72, higher score indicates higher fatalism, and has a Cronbach's alpha of 0.80. The score is associated with glycemic control, depression, self-care behaviors and social factors.

Quality of life (World Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015) is a widely used measure of quality life, validated in many languages and consists of only 5 items. Each self-reported item is scored between 0 and 5, summed together and then multiplied by 4. The total range is from 0 to 100, with 0 being the lowest quality of life.

Quality of life (World Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015) is a widely used measure of quality life, validated in many languages and consists of only 5 items. Each self-reported item is scored between 0 and 5, summed together and then multiplied by 4. The total range is from 0 to 100, with 0 being the lowest quality of life.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

Diabetes-specific social support is measured by the Diabetes Care Profile Support Questions,(Fitzgerald et al., 1996) with sub-scores for perceived disease specific support needs and perceived disease specific support received. Diabetes Care Profile support questions: Support wanted, and support received. Each range from 5 to 30, high scores indicating high desire for support and higher support received.

Diabetes-specific social support is measured by the Diabetes Care Profile Support Questions,(Fitzgerald et al., 1996) with sub-scores for perceived disease specific support needs and perceived disease specific support received. Diabetes Care Profile support questions: Support wanted, and support received. Each range from 5 to 30, high scores indicating high desire for support and higher support received.

Measure Description: General social support is captured by the Norbeck Social Support Questionnaire Emotional and&Tangible subscales.(Norbeck, Lindsey, & Carrieri, 1981) General social support: Norbeck Social Support Questionnaire, emotional sub score (ranges 0-16, with higher scores indicating higher perceived emotional support) & tangible sub score (0-8, higher score indicating higher perceived tangible support). While the subscores are highly correlated, the authors do not recommend Cronbach's alpha as a test of internal validity.(http://eileengigliotti.com/uploads/1/1/0/2/110241155/nssq-psychometric.pdf)

Measure Description: General social support is captured by the Norbeck Social Support Questionnaire Emotional and&Tangible subscales.(Norbeck, Lindsey, & Carrieri, 1981) General social support: Norbeck Social Support Questionnaire, emotional sub score (ranges 0-16, with higher scores indicating higher perceived emotional support) & tangible sub score (0-8, higher score indicating higher perceived tangible support). While the subscores are highly correlated, the authors do not recommend Cronbach's alpha as a test of internal validity.(http://eileengigliotti.com/uploads/1/1/0/2/110241155/nssq-psychometric.pdf)

Supporter diabetes-related distress is measured by the Partner Distress Scale (Polonsky, Fisher, Hessler, &Johnson, 2016). The self-report scale consists of 21 items in 4 domains: "my partner's diabetes management", "how best to help", "diabetes & me", & "hypoglycemia". Domains are summed together. Each item is score from 0 to 4, lower scores indicate less distress. The total score is an average of the 21 items. The Cronbach's alpha was 0.76 for total scores. Greater partner distress is correlated with higher A1C among patients, worse self-care & lower satisfaction with relationship

Total of all reporting groups

Ethnicity: Collected by patient self-report, with option for decline to state

Race: collected by patient self-report, with option for more than one race and decline to state

BMI: calculated from Weight and Height. As a measure of adiposity, it correlates positively with cardiovascular disease outcomes. While imperfect, it is an easily measured and validated marker.

2 participants were unable to stand to provide height

A measure of central adiposity, it correlates with cardiovascular outcomes.

20 participants were unable to provide abdominal circumference measurement

Systolic blood pressure mmHg: Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the systolic blood pressure for that visit. Systolic blood pressure is associated with cardiovascular complications

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the diastolic blood pressure for that visit.

Summary of Diabetes Self-care Activities (Toobert, Hampson, Glasgow, & RE, 2000). The Summary consists of 6 subscales representing different domains of diabetes related healthy behaviors and self-care. It has been validated in over 10 studies, with the results published in two manuscripts: one with three studies (Toobert & Glasgow, 1994), and one with seven studies(Toobert et al., 2000). Each measure ranges from Each range from 0-7, indicating number of days per week patient reports engaging in these behaviors.

Medication adherence(Wilson et al.,2014). The 3item medication adherence scale is a self-report measure that assessing no. of days medication missed in prior month, days taken medication as advised & self-rated medication adherence. Developed in HIV research, it has been validated for non-HIV medications (Wilson, et al,2016). Each component contributes 33% of the scale. Total score ranges from 0-100. It is the combination of 3 sub-scores, (days missed dose(negative scored), self-rating, days taking medications as indicated). Higher scores indicate better medication adherence.

1 participant decline to answer questions

Healthcare utilization by the patients' report of clinic appointments, ED visits and hospitalizations within enrollment through 6 months at each follow up visit. EMR review used instead, as self-report was inconsistent.

Self-efficacy (Diabetes Empowerment Scale Short Form) (Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), is a measure of a patient's overall diabetes related self-efficacy, shortened by the original from 37 to an 8 item scale. It has reliability of alpha=0.84; and has been shown to be sensitive to change in multiple populations and is correlated with improved glycemic control measured by A1C (Anderson et al., 2003). It ranges from 8-40 points; higher score indicates higher self-efficacy.

(Diabetes Distress Scale) (Polonsky et al., 2005), is a 17 item scale measuring distress related to emotional burden, physician-related distress, regimen-related distress, and diabetes-related interpersonal issues. It has a Cronbach's alpha of 0.88-0.93 in various studies. Higher scores are negatively correlated with several healthy behaviors. Each question is a Likert scale ranking of how serious a particular issue is from 1-6, 6 indicating a more significant problem. The scores are then averaged across all item, with final scores ranging from 1-6, with 6 indicating higher levels of distress.

11 participants decline to answer questions

Depression as measured by the PHQ-9 (Kroenke, Spitzer, Williams, & Lowe, 2010), is a widely used scale of depression used in clinical practice and research. The Cronbach's alpha ranges from 0.86-0.89, and it has been validated in multiple languages. Depression is worse with higher scores. Score ranges from 0 to 27. Higher levels of depression are associated with fewer healthy behaviors and worse glycemic control.

Fatalism is measured by the Diabetes Fatalism Scale (Egede & Ellis, 2010), which measures three subscale emotional distress, religiosity and coping and perceived self-efficacy, which are summed together. The total score ranges from 12-72, higher score indicates higher fatalism, and has a Cronbach's alpha of 0.80. The score is associated with glycemic control, depression, self-care behaviors and social factors.

Quality of life (World Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015) is a widely used measure of quality life, validated in many languages and consists of only 5 items. Each self-reported item is scored between 0 and 5, summed together and then multiplied by 4. The total range is from 0 to 100, with 0 being the lowest quality of life.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

Diabetes-specific social support is measured by the Diabetes Care Profile Support Questions,(Fitzgerald et al., 1996) with sub-scores for perceived disease specific support needs and perceived disease specific support received. Diabetes Care Profile support questions: Support wanted, and support received. Each range from 5 to 30, high scores indicating high desire for support and higher support received.

General social support is captured by the Norbeck Social Support Questionnaire Emotional and&Tangible subscales.(Norbeck, Lindsey, & Carrieri, 1981) General social support: Norbeck Social Support Questionnaire, emotional sub score (ranges 0-16, with higher scores indicating higher perceived emotional support) & tangible sub score (0-8, higher score indicating higher perceived tangible support). While the subscores are highly correlated, the authors do not recommend Cronbach's alpha as a test of internal validity.(http://eileengigliotti.com/uploads/1/1/0/2/110241155/nssq-psychometric.pdf)

14 participants decline to answer questions

Supporter diabetes-related distress is measured by the Partner Distress Scale (Polonsky, Fisher, Hessler, &Johnson, 2016). The self-report scale consists of 21 items in 4 domains: "my partner's diabetes management", "how best to help", "diabetes & me", & "hypoglycemia". Domains are summed together. Each item is score from 0 to 4, lower scores indicate less distress. The total score is an average of the 21 items. The Cronbach's alpha was 0.76 for total scores. Greater partner distress is correlated with higher A1C among patients, worse self-care & lower satisfaction with relationship

1 participant decline to answer question

Patients receive SMS-textmessage curriculum to improve self0-efficacy and self care for diabetes. A patient-identified family member receives a social support curriculum (FANS) by SMS-text-message synchronized by time and content.

TExT-MED: messages designed to inspire motivation and behavior change

FANS: SMS delivered messages to family members to improve social support

Glycemic control is measured by hemoglobin A1C collected at point-of-care from an Afinion AS100 capillary point of care machine. The Afinion machine has excellent point of care correlation with laboratory values. As a surrogate for average glycemic control over the previous 3 months and with correlation with clinical outcomes, hemoglobin A1c is a marker of overall clinical disease management. Change is 6 month value minus 12 month value.

2 intervention group patient refused blood draw at follow up.

Calculated from Weight and Height. As a measure of adiposity, it correlates positively with cardiovascular disease outcomes. While imperfect, it is an easily measured and validated marker.

4 participants were not able to stand on the scale to measure weight

Calculated from Weight and Height. As a measure of adiposity, it correlates positively with cardiovascular disease outcomes. While imperfect, it is an easily measured and validated marker.

5 participants unable stand to provide weight

A measure of central adiposity, it correlates with cardiovascular outcomes.

abdominal circumference unable to be collected from 20 patients

A measure of central adiposity, it correlates with cardiovascular outcomes.

abdominal circumference not measured in 9 participants

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the systolic blood pressure for that visit. Systolic blood pressure is associated with cardiovascular complications

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available. 2 participants refused BP measurement

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the systolic blood pressure for that visit. Systolic blood pressure is associated with cardiovascular complications

11 patients with SBP unable to be obtained

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the diastolic blood pressure for that visit.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available. 2 participants refused BP measurement

Blood pressure is measured by study RAs after the patient is seated for 5 minutes, with the average of three readings used as the diastolic blood pressure for that visit.

11 patients with DBP unable to be obtained

Summary of Diabetes Self-care Activities (Toobert, Hampson, Glasgow, & RE, 2000). The Summary consists of 6 subscales representing different domains of diabetes related healthy behaviors and self-care. It has been validated in over 10 studies, with the results published in two manuscripts: one with three studies (Toobert & Glasgow, 1994), and one with seven studies(Toobert et al., 2000). Each measure ranges from Each range from 0-7, indicating number of days per week patient reports engaging in these behaviors.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Summary of Diabetes Self-care Activities (Toobert, Hampson, Glasgow, & RE, 2000). The Summary consists of 6 subscales representing different domains of diabetes related healthy behaviors and self-care. It has been validated in over 10 studies, with the results published in two manuscripts: one with three studies (Toobert & Glasgow, 1994), and one with seven studies(Toobert et al., 2000). Each measure ranges from Each range from 0-7, indicating number of days per week patient reports engaging in these behaviors.

one participant did not answer carb spacing as they did not know what that meant.

Medication adherence(Wilson et al.,2014). The 3item medication adherence scale is a self-report measure that assessing no. of days medication missed in prior month, days taken medication as advised & self-rated medication adherence. Developed in HIV research, it has been validated for non-HIV medications (Wilson, et al,2016). Each component contributes 33% of the scale. Total score ranges from 0-100. It is the combination of 3 sub-scores, (days missed dose(negative scored), self-rating, days taking medications as indicated). Higher scores indicate better medication adherence.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Medication adherence(Wilson et al.,2014). The 3item medication adherence scale is a self-report measure that assessing no. of days medication missed in prior month, days taken medication as advised & self-rated medication adherence. Developed in HIV research, it has been validated for non-HIV medications (Wilson, et al,2016). Each component contributes 33% of the scale. Total score ranges from 0-100. It is the combination of 3 sub-scores, (days missed dose(negative scored), self-rating, days taking medications as indicated). Higher scores indicate better medication adherence.

1 Participant decline to answer question

Healthcare utilization by EMR review of clinic appointments, ED visits and hospitalizations within enrollment through 6 months at each follow up visit.

166 patients who initially enrolled followed in EMR

Healthcare utilization by EMR review of clinic appointments, ED visits and hospitalizations within enrollment through 6 months at each follow up visit.

all patients initially enrolled in trial followed in EMR

Change in Self-efficacy. Diabetes Empowerment Scale Short Form (Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), is a measure of a patient's overall diabetes related self-efficacy, shortened by the original from 37 to an 8 item scale. It has reliability of alpha=0.84; and has been shown to be sensitive to change in multiple populations and is correlated with improved glycemic control measured by A1C (Anderson et al., 2003). It ranges from 8-40 points; higher score indicates higher self-efficacy.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Change in Self-efficacy. Diabetes Empowerment Scale Short Form (Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), is a measure of a patient's overall diabetes related self-efficacy, shortened by the original from 37 to an 8 item scale. It has reliability of alpha=0.84; and has been shown to be sensitive to change in multiple populations and is correlated with improved glycemic control measured by A1C (Anderson et al., 2003). It ranges from 8-40 points; higher score indicates higher self-efficacy.

Diabetes related distress (Diabetes Distress Scale) (Polonsky et al., 2005), is a 17 item scale measuring distress related to emotional burden, physician-related distress, regimen-related distress, and diabetes-related interpersonal issues. It has a Cronbach's alpha of 0.88-0.93 in various studies. Higher scores are negatively correlated with several healthy behaviors. Each question is a Likert scale ranking of how serious a particular issue is from 1-6, 6 indicating a more significant problem. The scores are then averaged across all item, with final scores ranging from 1-6, with 6 indicating higher levels of distress

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Diabetes related distress (Diabetes Distress Scale) (Polonsky et al., 2005), is a 17 item scale measuring distress related to emotional burden, physician-related distress, regimen-related distress, and diabetes-related interpersonal issues. It has a Cronbach's alpha of 0.88-0.93 in various studies. Higher scores are negatively correlated with several healthy behaviors. Each question is a Likert scale ranking of how serious a particular issue is from 1-6, 6 indicating a more significant problem. The scores are then averaged across all item, with final scores ranging from 1-6, with 6 indicating higher levels of distress

Depression as measured by the PHQ-9 (Kroenke, Spitzer, Williams, & Lowe, 2010), is a widely used scale of depression used in clinical practice and research. The Cronbach's alpha ranges from 0.86-0.89, and it has been validated in multiple languages. Depression is worse with higher scores. Score ranges from 0 to 27. Higher levels of depression are associated with fewer healthy behaviors and worse glycemic control.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available. 1 participant decline to answer question

Depression as measured by the PHQ-9 (Kroenke, Spitzer, Williams, & Lowe, 2010), is a widely used scale of depression used in clinical practice and research. The Cronbach's alpha ranges from 0.86-0.89, and it has been validated in multiple languages. Depression is worse with higher scores. Score ranges from 0 to 27. Higher levels of depression are associated with fewer healthy behaviors and worse glycemic control.

2 participants decline to answer questions

Fatalism is measured by the Diabetes Fatalism Scale (Egede & Ellis, 2010), which measures three subscale emotional distress, religiosity and coping and perceived self-efficacy, which are summed together. The total score ranges from 12-72, higher score indicates higher fatalism, and has a Cronbach's alpha of 0.80. The score is associated with glycemic control, depression, self-care behaviors and social factors.

2 participants decline to answer questions. 2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Fatalism is measured by the Diabetes Fatalism Scale (Egede & Ellis, 2010), which measures three subscale emotional distress, religiosity and coping and perceived self-efficacy, which are summed together. The total score ranges from 12-72, higher score indicates higher fatalism, and has a Cronbach's alpha of 0.80. The score is associated with glycemic control, depression, self-care behaviors and social factors.

5 participants decline to answer questions

Quality of life (World Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015) is a widely used measure of quality life, validated in many languages and consists of only 5 items. Each self-reported item is scored between 0 and 5, summed together and then multiplied by 4. The total range is from 0 to 100, with 0 being the lowest quality of life.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.

Quality of life (World Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015) is a widely used measure of quality life, validated in many languages and consists of only 5 items. Each self-reported item is scored between 0 and 5, summed together and then multiplied by 4. The total range is from 0 to 100, with 0 being the lowest quality of life.

Diabetes-related supportive and obstructive family behaviors are measured by patient report on the Diabetes Family Behavior checklist (Lewin et al., 2005). Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family Behavior Checklist: supportive ranges from 4-45 (4 being lowest in family supportive behavior), non-supportive ranges from 7-35 (4 indicating least report of non-supportive behavior). Non-supportive scores are subtracted form supportive scores for a total.

2 participants in active control arm did not attend follow up visit. Hba1c taken from medical records, no other data available.