A Nonrandomized Pilot of a Group Video-Based Telehealth Intervention to Reduce Diabetes Distress in Parents of Youth with Type 1 Diabetes Mellitus

Abstract

Objectives:

To refine and pilot a video-based telehealth intervention to reduce diabetes distress, depressive symptoms, and hypoglycemia fear in parents of school-age children with type 1 diabetes and to explore for changes in child glycated hemoglobin (HbA1c).

Methods:

We recruited 41 parents of children (5–12 years) to participate in a manualized, video-based telehealth intervention (CARES). Of these, 29 parents completed either a 12- (n=13) or 8-week (n=16) version of CARES based on the timing of their recruitment. We assessed feasibility (i.e., attrition, attendance) and parent satisfaction with CARES. We used repeated measures ANOVA with parent group (8- versus 12-sessions) as a between-subject variable and time as a within-subject variable to measure change in our dependent variables.

Results:

Mostly mothers participated (97.3%). Parents’ mean age was 39.65±6.84 years and child mean age was 9.86±1.57 years at pre-treatment. CARES had low attrition (20–25%) and good attendance (96–98%). Parents also reported high levels of treatment satisfaction (>85%). There were significant main effects for time for parent-reported diabetes distress and depressive symptoms at post-treatment and three-months follow-up. There was a statistical trend suggesting a time × group interaction for parent depressive symptoms at post-treatment. There was a significant main effect for time for hypoglycemia fear at the three-months follow-up but no change at post-treatment. There was no change in child HbA1c at post-treatment.

Discussion:

CARES showed high parent satisfaction, good feasibility, and promising results for reducing diabetes distress in parents of school-age children with type 1 diabetes.

Introduction

The daily management of type 1 diabetes is arduous and time-consuming [1]. Parents of school-age children typically play a major role in their child’s type 1 diabetes care [1]. This level of responsibility leaves parents vulnerable to negative impacts on their psychosocial well-being, such as diabetes distress [1, 2]. Diabetes distress comprises a constellation of negative psychological symptoms that can overlap with symptoms of fear and depression, but specifically relate to living with diabetes or caring for someone with diabetes [3]. Diabetes distress can also include diabetes-specific behavioral symptoms such as suboptimal treatment engagement or maladaptive coping strategies (e.g., guessing an insulin dose to avoid checking glucose) that can directly relate to higher glucose levels and increase the risk for future health complications [3–5].

There are published examples of promising treatments to reduce diabetes distress in adults with type 1 diabetes [6–8], and one study that intervened in mothers of adolescents with type 1 diabetes [9]. Most of the existing interventions use cognitive behavioral principles, problem-solving, and/or education to reduce diabetes distress [7–9]. For example, to address diabetes distress in mothers of adolescents with type 1 diabetes, Jaser et al. [9] piloted an individual, seven session in-person/telephone intervention targeting positive parenting and coping behaviors. In an initial sample of mothers randomized to either the treatment (n=15) or usual care control arm (n=15), researchers reported a significant reduction in maternal reported diabetes distress for the intervention versus usual care [9]. Yet on average, mothers completed only 57% of sessions [9], revealing an opportunity to pilot other treatment delivery methods that might enable greater session attendance. In addition, to date there are no interventions specifically targeting diabetes distress in parents of younger children with type 1 diabetes, despite recent data suggesting between 20–33% of parents of school-age youth with type 1 diabetes endorse diabetes distress [4, 5]. Finally, targeting diabetes distress in parents of school-age children may serve as a preventative measure. Adolescence is a period when sub-optimal type 1 diabetes treatment engagement and glycemia can begin to occur and low parental involvement can exacerbate these outcomes [10]. It is possible providing parents with coping strategies before these potential changes happen could help parents to better regulate any feelings of diabetes distress and to maintain supportive engagement in their child’s type 1 diabetes into adolescence.

Building from existing interventions [7–9], we designed an intervention to reduce diabetes distress in parents of school-age children with type 1 diabetes (ages 5–12 years-old). To address attendance concerns from the previous maternal-focused diabetes distress intervention [9], we adopted a delivery approach that we found to be successful in a previous group-based telehealth intervention targeting hypoglycemia fear in parents of preschoolers with type 1 diabetes (REDCHiP) [11, 12]. Specifically, with a telehealth treatment approach in our previous study parents attended 94% of treatment sessions and noted the convenience of participating from home, work, or school [11]. Parents also noted that the group format facilitated a supportive environment that empowered parents to cope with the stressors of raising a child with type 1 diabetes [11].

To refine and pilot our intervention, we applied the ORBIT model for behavioral intervention development [13]. The ORBIT model was first introduced in 2015 and provides a progressive and iterative approach for developing and testing new interventions. A key difference of the ORBIT model versus other approaches is its focus on using small, fluid, and cost-effective formative studies to develop and refine interventions ahead of launching expensive efficacy trials [13]. The ORBIT model defines four phases of treatment development. Phase 1 includes defining the basic elements of the intervention (Phase 1a: Define) and refining the intervention to determine its delivery mode, target population, and treatment length (Phase 1b: Refine). Phase 2 involves proof of concept testing (Phase 2a) and randomized pilot testing (Phase 2b), while Phases 3 and 4 use larger samples, robust designs (e.g., blinding, placebo or an attention-only condition) or implementation science to explore an intervention’s efficacy and effectiveness, respectively [13].

The goal of the current study was to conduct a Phase 1b evidentiary study examining feasibility, parent satisfaction, and initial treatment response of our telehealth intervention in the context of a small, nonrandomized trial. Our primary hypothesis was that parents who participated in our intervention would report significantly lower diabetes distress post-treatment and at three-months follow-up. We also hypothesized that our intervention would be feasible, meaning that it would exhibit an attrition rate of <25% of families and a group attendance rate of >85% of all telehealth sessions. Finally, we tested for possible secondary changes in parent-reported depressive symptoms and hypoglycemia fear at post-treatment and three-months follow-up, as parents may experience symptom overlap between diabetes distress, depression, and fear [3, 5, 12]. We also explored for any changes in child glycated hemoglobin (HbA1c) at post-treatment because of our intervention.

Methods

Participants and Procedures

We recruited all families from an 11-site Pediatric Diabetes Clinic network in the Mid-Western region of the United States. Eligible families had a child between 5–12 years-old with a diagnosis of type 1 diabetes who followed an intensive insulin regimen. The parent who participated had to be the child’s legal guardian and English-speaking. Further, we did not recruit children who evidenced type 2 diabetes or monogenic diabetes, other co-morbid chronic illnesses (e.g., renal disease), or the chronic use of medications that could impact glycemic control (e.g., systemic steroids). Parents provided written informed consent to participate and permission to measure their child’s HbA1c. We obtained institutional review board approval prior to study initiation and registered the pilot at ClinicalTrials.gov (NCT03698708).

The parent who identified as the primary caregiver of their child’s type 1 diabetes completed all study assessments and attended each intervention session. Additional caregivers were welcome to attend intervention sessions, but to examine the study hypotheses, we only used data from the primary caregiver. At pre-treatment, parents provided demographic information and reported on their child’s medical history. We collected outcome variables at pre-treatment, post-treatment, and three-months follow-up using REDCap software [14] for parent surveys and a validated fingerstick blood sample and mail-in kit for children’s HbA1c levels. We paid families for completing each study assessment, but we did not reimburse parents for attending intervention sessions. We recorded recruitment data (e.g., number approached, enrolled, completed) and telehealth session attendance to assess feasibility. We collected parent feedback about ways to improve our intervention through qualitative group interviews at the end of each cohort.

Intervention

Cognitive Adaptions to Reduce Emotional Stress (CARES), is a manualized, closed-group intervention specifically designed for parents of school-age children with type 1 diabetes. CARES builds on an existing intervention to reduce diabetes distress in mothers of adolescents with type 1 diabetes by using cognitive-behavioral therapy (CBT) strategies to help parents to identify unhelpful thoughts and feelings related to caring for a child with diabetes [9, 15, 16]. However, new treatment elements specific to CARES include its use of acceptance and commitment therapy strategies to help parents to manage concerns that may not be modifiable (e.g., distress related to long-term complications of diabetes) [17, 18] and its focus on behavioral activation and mindfulness strategies to help parents reduce distress through healthy coping [19, 20]. Each CARES session lasts between 45–60 minutes and a group leader facilitates the progression of each treatment session. In sessions, the group leader teaches parents new treatment content and parents can reflect on their own thoughts and feelings about diabetes, practice treatment and coping strategies, and give and receive support from other parents. This pilot trial used masters-level clinical psychology students to deliver the intervention under the supervision of a licensed clinical psychologist.

In this pilot trial, parents participated in CARES via the HIPAA-compliant Zoom application. The first four cohorts of parents participated in a 12-session CARES version. Then, based on parent feedback, we pivoted to an 8-session CARES version to reduce participant burden. In our 8-sesson CARES format, we retained information targeting parents’ unhelpful thoughts and feelings about diabetes and teaching parents new coping strategies for diabetes distress because parents universally reported finding this information most helpful in qualitative interviews (e.g., “The coping skills, the deep breathing, and being in the present, just everything like I said, I learned something new. But I did like the coping skills the best.”). However, we removed sessions that reviewed diabetes education because most parents identified these sessions as less helpful (e.g., “I think, probably, the least helpful for me was the basic stuff. The nuts and bolts of diabetes was for me, because we’ve been doing this for a while.”) and we reduced the number of opportunities parents had to practice treatment strategies in session (See Figure 1).

Figure 1.

CARES 12-Session and 8-Session Intervention Content

Measures

Problem Areas in Diabetes (PAID-PR) [21]

Problem Areas in Diabetes (PAID-PR) [21] is a validated 18-item measure of diabetes distress, with higher scores suggesting greater distress. We scored the PAID-PR based on a Total score (range: 0–72). The present sample had a Cronbach’s alpha level of 0.87. We used parents’ PAID-PR score to test our primary hypothesis.

Center for Epidemiological Studies Depression Scale-Revised (CESD-R) [22]

Center for Epidemiological Studies Depression Scale-Revised (CESD-R) [22] is a validated 20-item measure of depressive symptoms occurring within the last two weeks. We scored the CESD-R based on a continuous Total score (range: 0–60), with higher scores reflecting more symptoms. The present sample had a Cronbach’s alpha level of 0.92. We used parents’ CESD-R score as a secondary outcome measure.

Hypoglycemia Fear Survey for Parents (HFS-P) [23].

The HFS-P is a validated 25-item parent-report measure of hypoglycemia fear. The HFS-P uses a 5-point Likert scale response set (1=Never to 5=Very Often) with higher scores suggesting greater hypoglycemia fear. We scored the HFS-P based on a continuous Total Score (range: 25–125) and parents’ Cronbach’s alpha level was 0.90 for the study. We used parents’ HFS-P score as a secondary outcome measure.

Child glycemic levels (HbA1c).

We used a validated fingerstick mail-in kit at pre- and post-treatment to measure children’s HbA1c. We analyzed children’s HbA1c levels in a central laboratory using automated high performance liquid chromatography with methods reliable to Diabetes Control and Complications Trial (DCCT) standards (Reference range 4.0–6.0%; Tosoh 2.2, Tosoh Corporation, San Francisco, CA) [24]. We used children’s HbA1c to explore for any changes resulting from CARES.

Treatment Satisfaction Survey. We used a 15-item Treatment Satisfaction Survey to rate satisfaction with the intervention on a 5-point Likert scale (1= not at all to 5= very or a lot). Participants competed this survey at the end of the intervention period. We scored parents’ treatment satisfaction based on a total score (range: 15–75).

Data Analyses

We calculated descriptive statistics to examine feasibility and parents’ satisfaction with CARES. We completed these calculations separately for parents who completed the 8- (n=16) and 12- (n=13) session CARES versions, respectively. To test our primary hypothesis, we used a series of repeated measures analysis of variance (ANOVA) models with parent group (8- versus 12-sessions) as a between-subject variable, time as a within-subject variable, and parents’ PAID-PR score as the dependent variable. We used separate models to test for treatment effects between parents’ pre- and post-treatment PAID-PR scores and between parents’ pre- and 3-months follow-up PAID-PR scores (maintenance effect). To test for treatment effects in our secondary outcomes (i.e., parents’ CESD-R and HFS-P scores), we used a similar analytical approach. We also used repeated measures ANOVA to explore for changes in child HbA1c between pre- and post-treatment. Power calculations suggested 80% power for large effect sizes (0.80) and n=15 at a two-tailed alpha of 0.05. To assist with interpreting our results, we report Cohen’s d effect sizes, generally interpreted as small (0.2), medium (0.5), or large (0.8) [25, 26] in addition to the alpha levels.

Results

Table 1 lists pre-treatment family demographic and child medical history data for the full sample, as well as separately for families who participated in the 8- or 12-session CARES versions. Overall, we approached 41 families, 37 families enrolled, and 29 completed treatment (21% attrition). Most participants were mothers. Families who were lost to attrition reported scheduling difficulties, no longer having time to participate in the study, or they became unreachable by study staff. The average number of group members for the 8- and 12-session CARES was 3.2 and 3.5, respectively (overall mean= 3.33). There were no differences in demographic characteristics for the families who completed treatment versus the families who did not complete treatment (n=8; p’s>0.08). Similarly, there were no differences in demographic characteristics for our sample versus the larger clinic population from which it was drawn (p’s>0.06). Since we pivoted to an 8-session CARES midway through our pilot trial and did not randomize parents to either the 8- or 12-session CARES, we also compared parents who completed either the 8- or 12-session CARES on their pre-treatment outcome variables and found no significant differences (p’s >0.15). See Table 2 for descriptive statistics of outcome measures at each time point for all parents and subdivided for parents completing either the 8- or 12-session CARES.

Table 1.

Sample Characteristics

	Sample Enrolled n=37	12-session n=17	8-session n=20	Drop-out n=8
Child age, yr (M±SD)	9.86±1.57	9.59±1.62	10.10±1.52	9.25±1.83
T1D duration, yr (M±SD)	3.00±2.55	2.64±1.95	3.28±2.97	2.14±1.21
HbA1c, % (M±SD)	8.24±1.26%	8.25±0.84	8.24±1.56	7.88.24±1.02%
HbA1c, mmol/mol (M±SD)	67±13.8	67±9.2	67±17.1	63±11.1
HbA1c >7.5%(58 mmol/mol) at pre-treamtent (%)	25 (68)	14 (82)	11 (55)	5 (62)
Boys, n (%)	21 (56.8)	7 (41.2)	14 (70)	5 (62.5)
Non-Hispanic White, n (%)	32 (86.5)	15 (88.2)	17 (85.0)	8 (100)
Insulin pump, n (%)	27 (75)	12 (70.6)	16 (80)	6 (75)
Continuous glucose monitor, n (%)	21 (56.8)	11 (64.7)	10 (50)	4 (50)
Parent age, yr (M±SD)	39.65±6.84	41.29±7.49	38.25±6.07	35.75±6.56
Mother, n (%)	36 (97.3)	16 (94.1)	20 (100)	7 (87.5)

Table 2:

Descriptive Statistics for Outcomes Measures at each Timepoint

		All participants					8 Sessions					12 Sessions
Measure	Time	N	Mean	SD	Min	Max	N	Mean	SD	Min	Max	N	Mean	SD	Min	Max
Child HbA1c (%)	Pre-Tx	37	8.24	1.26	5.9	11.4	20	8.24	1.56	5.9	11.4	17	8.25	0.84	6.9	10.2
	Post-Tx	29	8.19	1.15	5.8	10.3	16	8.17	1.34	5.8	10.3	13	8.21	0.92	7.1	10
	Follow-up	28	8.35	1.28	5.9	10.8	15	8.47	1.50	5.9	10.8	13	8.20	1.00	6.7	9.9
Parent PAID-PR Total	Pre-Tx	37	36.45	13.77	8.75	61.25	20	39.69	14.72	12.5	61.25	17	32.65	11.87	8.75	51.25
	Post-Tx	29	24.40	11.32	5	50	16	27.50	12.76	7.5	50	13	20.58	8.16	5	38.75
	Follow-up	28	24.24	12.21	3.75	51.25	15	25.83	14.98	3.75	51.25	13	22.40	8.17	8.75	40
Parent CESD-R	Pre-Tx	37	13.49	10.17	1	34	20	11.25	8.87	1	32	17	16.12	11.21	1	34
	Post-Tx	29	10.79	9.66	0	38	16	11.56	11.08	0	38	13	9.85	7.91	1	27
	Follow-up	28	10.11	9.63	0	38	15	9.47	8.90	0	31	13	10.85	10.72	1	38
Parent HFS-P Total	Pre-Tx	37	67.27	15.13	39	105	20	67.95	16.27	39	105	17	66.47	14.12	40	91
	Post-Tx	29	61.31	9.70	44	81	16	63.56	9.56	52	81	13	58.54	9.49	44	72
	Follow-up	28	57.89	9.55	42	77	15	59.73	9.48	46	77	13	55.77	9.55	42	74

Note. Pre-Tx, pre-treatment assessment; Post-Tx, post-treatment assessment; Follow-up, three-months follow-up assessment

Feasibility

Of the 37 families who enrolled, we initially assigned 17 families to complete our 12-session CARES and of these, 13 families completed treatment and the three-months follow-up (24% attrition rate). We initially assigned 20 families to complete our 8-session CARES and of these, 16 families completed treatment (20% attrition rate) and 15 completed the three-months follow-up (25% attrition). For parents who started treatment, attendance at group sessions ranged from 96–98% with no difference for parents participating in the 8- or 12-session CARES.

At post-treatment parents completed a Treatment Satisfaction Survey. Parents who completed the 12-session CARES reported a total score of 68.54±5.8, or 91.4% satisfaction, while parents who completed the 8-session CARES reported a total score of 66.38±6.6, or 88.5% satisfaction.

Pre-Post Treatment Outcomes

Repeated-measures ANOVA model results indicated main effects for time for parents’ PAID-PR total score and CESD-R score. Consistent with our primary hypothesis, parents reported significant reductions in diabetes distress (d=0.88, p<001) pre- to post-treatment. Parents also reported significant reductions in depressive symptoms (d=0.22, p=.0.02) pre- to post-treatment, though the time × group interaction showed a statistical trend (d=0.15, p=0.052) suggesting a greater reduction in depressive symptoms for parents who participated in the 12- versus 8-session CARES. There was no significant time × group interaction for parents’ PAID-PR scores pre- to post-treatment. We also did not find a significant main effect for time or a time × group interaction for parents’ HFS-P total scores or child HbA1c.

Maintenance of Treatment Outcomes

Model results assessing maintenance of treatment effects found main effects for time for parents’ PAID-PR total score, CESD-R score, and HFS-P score but no time × group interaction for these dependent variables. Again, consistent with our primary hypothesis, parents reported significant reductions in diabetes distress (d= 1.13, p<0.001) pre-treatment to three-months follow-up. Similarly, we observed significant reductions in parents’ depressive symptoms (d= 0.40, p=0.003) and hypoglycemia fear (d=0.29, p=0.015) at three-months follow-up compared to pre-treatment levels.

Discussion

In this study, we set out to refine and pilot a group, video-based telehealth intervention specifically targeting diabetes distress in parents of school-age children with type 1 diabetes (ages 5–12 years-old). Overall, our findings suggest that a telehealth approach is feasible; we observed relatively low attrition rates during the treatment period (20–24%) and families attended 96–98% of virtual sessions. Parents also reported high levels of treatment satisfaction (>85% satisfaction) with CARES and their qualitative input helped us to streamline CARES by shortening the treatment length from 12- to 8-sessions without removing the content that parents found most helpful.

Our findings also offer promising results for CARES. In support of our primary hypothesis, parents who received CARES reported significant reductions in diabetes distress post-treatment and a maintenance of these treatment effects up to three months later. Similarly, in support of one of our secondary hypotheses, parents reported significant reductions in depressive symptoms at post-treatment and a maintenance of reduced parent depressive symptoms at three-months follow-up. As an unexpected outcome, we observed a statistical trend for a time × group interaction for depressive symptoms at post-treatment, potentially signaling a greater reduction in depressive symptoms among parents in the 12- versus 8-sessions CARES. In partial support of one of our secondary hypotheses, we observed a significant reduction in parents’ hypoglycemia fear when examining their scores pre-treatment to three-months follow-up but no change in their scores pre- to post-treatment. Finally, there was no change in child HbA1c pre- to post-treatment.

We applied the ORBIT model to guide our study design and specifically conceptualized this as a Phase 1b study to refine the CARES intervention [13]. Based on this design, we did not randomized parents to a treatment arm and we did not include a control. In the absence of a control group, we cannot rule out the possibility that parents’ symptoms improved on their own and not because of CARES. Yet evidence from a longitudinal study that recruited adults with type 1 diabetes [27] and a longitudinal study that recruited parents of youth with type 1 diabetes [28] indicate that diabetes distress levels can remain relatively stable for up to nine months, offering some support for the notion that parents may have indeed benefited from receiving CARES. Likewise, because we did not randomize parents to receive either the 8- or 12-session CARES, we cannot to rule out the possibility that individual differences across groups affected our study results. Prior to analyzing our results, we conducted comparisons of the 8- and 12-session CARES groups and found no significant differences in parent demographics or pre-treatment scores, which may help to strengthen our pilot results. Nonetheless, to confirm our study results, we would advocate for a future, larger randomized controlled trial to mitigate the potential impact of parent confounds. Similarly, a larger trial could offer greater power to explore differences between the 8- and 12-sessions CARES on parent outcomes and to replicate the trend time × group interaction we observed for parent depressive symptoms.

To deliver CARES, we used a group, video-based telehealth approach. The primary reason we selected a group approach was based on feedback from a previous study, in which parents noted the importance of a supportive, empowering group environment with other parents and group leaders in their overall treatment experience [11]. In CARES, given its focus on reducing parents’ diabetes distress, we wanted the group environment to help reduce any feelings of isolation parents might have and to normalize the challenges they experience caring for a child with type 1 diabetes. Moreover, parents’ qualitative report supported this design decision (“The group really helped me and I learned so much from all of you guys.” “I think sometimes we kind of get siloed in our own brain and then somebody goes, ‘why don’t you do this?’ And you go, gosh that was a really easy solution that I didn’t think of.”) The primary reason we selected to deliver CARES using video-based telehealth was to eliminate the necessity for parents to leave work or home and travel to the treatment site for groups, which we hoped would facilitate their attendance. There are other potential benefits to using video-based telehealth. For example, this delivery approach could make it easier for families who live a long distance from the clinic and/or families with limited access to transportation to receive treatment. It is possible a telehealth approach may be highly cost-effective and scalable [29]. Also, given recent changes in the guidance surrounding telehealth care during the COVID-19 pandemic, it is possible that it could be fairly resistant to disruptions due to public health events (e.g., sheltering at home, social distancing). Nonetheless, because we did not pilot an alternative delivery approach (e.g., in-person groups), we cannot definitively determine the benefits of using video-based telehealth to deliver CARES. Thus, future research should include distance and perceived burden metrics to compare whether parents prefer telehealth participation in CARES versus an alternative delivery approach, or whether family characteristics could be used to tailor preferred treatment delivery method. Related, we would also recommend that future research examine for any social or health inequities that could be inadvertently created through a video-based telehealth approach and attempt to mitigate these inequities.

Consistent with the previous intervention to reduce distress in mothers of adolescents with type 1 diabetes [9], our exploratory analyses did not find any changes in child HbA1c related to parents’ participation in CARES. It is possible that we were underpowered to detect a significant change in child HbA1c due to our small sample size and because we did not restrict CARES participation only to parents of children with elevated HbA1c levels. However, this negative result also aligns with past research which has sometimes shown a more robust response for psychosocial versus glycemic outcomes following behavioral interventions in persons with type 1 diabetes [30]. Future research is needed to explore for changes in child HbA1c after parents’ participate in CARES in a larger, randomized trial.

Some potential strengths of this study include our use of the ORBIT model for behavioral intervention development to standardize study methods and help refine the intervention [13], our recruitment of a sample that is representative of the larger clinic population, our use of robust statistical tests to analyze two manualized intervention formats that incorporate evidence-based strategies to reduce internalizing symptoms and distress, our use of validated outcome measures, including a validated home HbA1c kit and a single laboratory to analyze children’s HbA1c levels using a gold-standard method (HPLC) [24], and our assessment of any maintenance of treatment effects.

In addition to previously mentioned limitations (i.e., nonrandomized design, no control group), we recognize that this study recruited a relatively small sample size, which introduces the possibility that we did not observe some changes in outcomes because of limited power. There was a lack of participation by fathers or other primary caregivers (e.g., grandparents, legal guardians, etc.), which could have been a negative consequence on our decision to collect data only from the primary caregiver of the child’s type 1 diabetes. In a future larger trial, we may need to use an alternative recruitment approach to promote higher participation by fathers and other caregivers. Lastly, even though our sample demographics matched the diversity of the larger clinic population, we acknowledge that there was limited racial and ethnic diversity in our sample, which could affect the generalizability of our results.

Brief Conclusion

We applied methods from the ORBIT model for behavioral intervention development conduct a small, nonrandomized study to refine and pilot a group video-based telehealth intervention for parents of school-age children with type 1 diabetes. Our results suggest initial feasibility for the CARES intervention and high levels of parent satisfaction. Our results also suggest that it is plausible CARES can have a significant and beneficial effect in reducing parents’ report of diabetes distress and depressive symptoms. Per the ORBIT model, evidence of achieving these milestones indicate a treatment may be appropriate for Phase 2 proof of concept or randomized pilot testing. [13] Therefore, as a next step, we would support a randomized, controlled trial of CARES against a usual care control group to examine the source of treatment effects (i.e., the intervention or spontaneous remission). We would encourage future trials recruit greater participation from fathers and other primary caregivers [31] as well as with greater participation of families from minoritized backgrounds for whom known health inequities in type 1 diabetes care exist [32]. Finally, we would encourage a formal comparative effectiveness trial to determine the optimal treatment dose of CARES, as these data may help to inform future implementation of the intervention in clinics.

Key Messages:

• Diabetes distress is common among parents of school-age children with type 1 diabetes yet there are no targeted interventions for these families.

• We designed and piloted tested CARES, a video-based telehealth intervention specifically targeting diabetes distress in parents of school-age children.

• Study results show promising evidence that CARES is feasible and can reduce parents’ symptoms of diabetes distress.