Scientific Papers

A pilot randomised controlled parallel arm trial evaluating treatment satisfaction with the Omnipod DASH® Insulin Management System compared with usual care in adults with type 1 diabetes in Australia: rationale, study design and methodologies | Pilot and Feasibility Studies

Study setting

The study will be conducted within Australia at four diabetes centres with St Vincent’s Hospital Melbourne as the lead site. Three of the study sites are metropolitan tertiary hospitals and academic centres (St Vincent’s Hospital Melbourne, Austin Health, Royal Melbourne Hospital), and one is a private regional specialist diabetes service (Geelong Endocrinology and Diabetes [Geelong]). All four participating centres have the physical infrastructure and human resources for the initiation and follow-up of patients with T1D on IPT.

Eligibility criteria

All participants will provide written, informed consent prior to any study procedures occurring (See Additional file 1: Appendix 2 for sample informed consent form).

Full participant inclusion and exclusion criteria for study enrolment are listed in Table 1. At least 60% of participants recruited will be on MDI, reflecting the fact that the majority of adults with T1D in Australia are managed with injections [7]. The age range for inclusion of 18 to 70 years reflects 97.6% of T1D adults on IPT in the country [7]. Children with T1D are excluded because their diabetes technology user experience involves their caregivers and has a different focus.

Glucose monitoring will be by SMBG given that access to continuous glucose monitoring (CGM) for people with T1D varies from country to country depending upon social advantage and wealth. Prior to the 1st of July 2022, the CGM subsidy in Australia was restricted to people with T1D under 21 years old [9], with the majority of T1D adults performing SMBG readings due to the lack of CGM affordability. For this study, CGM use for less than 25% of the time during 3 months before study commencement is permissible based on previous evidence demonstrating no benefit in glycaemic management with intermittent CGM use in T1D [10]. The use of intermittently scanned glucose monitoring (ISGM) without hypoglycaemia/hyperglycaemia alarms (Abbott Libre™ 1) is also permissible because previous data demonstrated that ISGM without alarms does not equate to real-time CGM as it is not associated with improvements in HbA1c or time in range compared to SMBG in T1D [11, 12]. The Abbott Libre™ 1 ISGM does not have the ability to show real-time glucose data unless scanned and does not have predictive high or low alert alarms, so users are not automatically prompted to change treatment decisions [11]. Participants using Libre™ 1 ISGM must be willing to perform at least four SMBG readings daily for study inclusion.

Participants are restricted to those with a maximum insulin requirement of 200 units every 2 days because the Omnipod DASH® System is worn for 2 to 3 days and can only hold 200 units of insulin [4]. Therefore, those with a higher insulin requirement than this are not suitable.


All eligible participants will receive a refresher in general diabetes education and carbohydrate counting during a 4-week run-in to minimise the impact of diabetes education as a confounder when assessing outcomes. Participants will then be randomised in equal proportions between intervention (Omnipod DASH® Insulin Management System) and usual care where they will continue to use their own insulin delivery method, for 12 weeks. Prior to the provision of the Omnipod DASH® Insulin Management System, a member of the study team will calculate and enter the insulin delivery settings in the participant’s study device, and they will be trained in all aspects of its operation. Following the initiation of Pod therapy participants will have study follow-up appointments for insulin dosing adjustment which have been scheduled to reflect clinical practice. At the end of the 12-week study, an extension phase will be implemented where those allocated to usual care will be provided with the opportunity to receive the intervention and those who are on the intervention can continue with the device for a further 12 weeks.


Discontinuation of the intervention may be initiated at the request of the participant or if there are any concerns on the part of the investigation team regarding the safety of the participant e.g. the person is unable to operate the insulin delivery device safely despite remedial action with evidence of severe and persistent hyperglycaemia and/or ketosis or recurrent hypoglycaemia. Under these circumstances, the study team will cease therapy with Omnipod DASH® System and ensure the participant’s safe transition to their usual mode of insulin delivery.


To maximise adherence to the protocol, a formal checklist is used during the consenting process to ensure that the participants fully understand what will be required of them when they enter the study. The requirement to complete the general diabetes education and carbohydrate counting training during run-ins will also help to ensure that only those who are able to fulfil the requirements of the protocol are randomised. Participants allocated to intervention will be educated in the use of the device by a skilled trainer which will help to ensure that the devices are used optimally. Follow-up of all participants, including review of glucose meter and device uploads, will be done by a clinically experienced team of doctors and diabetes nurse educators ensuring that the participants feel supported during the study and meet the requirements of the protocol. In addition, prior to study appointments participants will receive an automatically generated text message reminder.

Concomitant care

In both intervention and usual care study arms, changes in insulin dosing can be made in response to the study team’s assessment of the participant’s glucose control. All other aspects of the participant’s care (e.g. antihypertensive, lipid-lowering, antidepressant medications) are to be continued throughout the study with changes to regimens decided upon according to clinical review by their usual health care professionals.

Participant timeline

The study consists of a total of 14 visits during the 4-week run-in (weeks 4 to 0), the 12-week randomised controlled study (weeks 0 to 12) and the extension period (weeks 0 to 24) (Fig. 3). Some study visits may be conducted remotely at the investigators’ discretion following experience with the COVID-19 pandemic. If required additional contact with the study team will be available. Investigators will document communications and additional reviews with participants in the participant file.

Fig. 3
figure 3

SPIRIT Figure summarising the study design

Screening and run-in period

Participants who pass screening will undergo a 4-week run-during which they will be provided with carbohydrate counting education by a specialist dietitian and trained on the study BG meter with a meal insulin dosing calculator during a face-to-face visit with the study dietitian. As proficiency in carbohydrate counting is a cornerstone of diabetes management [4], all participants, including those participants already on conventional insulin pumps, will not proceed to randomisation unless this has been demonstrated. If necessary, the run-in can be extended by 2 weeks to ensure proficiency.

Following carbohydrate counting education, all participants will then undergo 2 weeks of masked-CGM prior to randomisation. They will also receive education and written information on the use of the KeyLead Health™ mobile application used for data collection.


Participants randomised to Omnipod DASH® System will upload data from the study meter and PDM at clinic reviews in weeks 1, 2, 3 and 6. These reviews may be performed remotely at the discretion of the study team. During clinic reviews, all participants will have diabetes therapy adjustments (PDM settings for Omnipod DASH® System or insulin doses for MDI/IPT) as necessary in conjunction with study clinicians. Participants randomised to usual care will continue using their usual insulin delivery modality (MDI or IPT). They will upload data from the study meter (and insulin pump if on IPT) at clinic review in week 6.

Extension phase

After 12 weeks, all participants will be provided with the opportunity to take part in the extension phase. Those on usual care will transition to the Omnipod DASH® System and upload data from the study meter and PDM at clinic reviews in weeks 13, 14, 15 and 18 for adjustments of insulin dosing. These reviews may be performed remotely at the discretion of the study team. Those assigned to the intervention arm will continue using the Omnipod DASH® System during the extension phase and upload device data at clinic review in week 18.

Data to be collected from all participants at the end of the study (week 12) and at the end of the extension phase (week 24) will include 2 weeks of masked CGM which will be obtained from weeks 10 to 12 and weeks 22 to 24. At weeks 12 and 24, HbA1c, questionnaire responses and data from all study devices will also be collected. Semi-structured interviews will be undertaken at week 12 for participants randomised to Omnipod DASH® System and at week 24 for participants randomised to usual care.

Outcomes (Table 2)

Primary feasibility outcomes

These outcomes will inform power calculations for the future main study.

  1. (i)

    Feasibility will be determined by the proportion of the number of participants (n = 64) who complete the study per-protocol at week 12. The study design will be deemed feasible if the proportion completing the study is at least 0.80.

  2. (ii)

    Acceptability of the Omnipod DASH® Insulin Management System in those who complete the study will be assessed by the change in DTQ-current score [13, 14] from baseline to week 12, with comparisons made with usual care. DTQ is a 30-item validated measure of satisfaction with diabetes technological tools used in T1D management. It yields a separate score for ‘current’ satisfaction (DTQ-current: How much is this a problem now?) and for ‘change’ in satisfaction (DTQ-change: How has it changed compared to before the study?), ranging from 30 to 150 for each score [13, 14]. While there is currently no single universally accepted tool for assessing treatment satisfaction with diabetes technology devices this tool is simple to implement and does not impose a significant added burden upon the participant. It has been used in other studies evaluating novel diabetes technologies [6, 15].

Secondary feasibility outcomes

  1. (i)

    User acceptance: Differences between the intervention and usual care study arms for DTQ-change score at week 12 will be analysed as will DTQ-current and DTQ-change at the completion of the extension period (week 24) [13, 14] to determine if these may offer greater resolution in delineating differences in user-perceptions (Table 3).

  2. (ii)

    Process outcomes: These will include clarity and suitability of inclusion and exclusion criteria, recruitment and retention rates, the need for additional study visits and a better understanding of data collection tools and outcomes including unanticipated responses to questionnaires. Reasons for study non-participation, exclusion and study withdrawal will be documented to provide information about the suitability of inclusion and exclusion criteria. If it is taking longer than 18 months for the completion of the recruitment of 64 participants, study inclusion and exclusion criteria may need to be reconsidered. The need for additional study visits will be documented on REDCap®. Revision to the study design will need to be considered if additional study visits are required by more than 20% of participants. The protocol does allow for remote education if a global pandemic restricts participant visits to the clinical trial centre. Data for the mode of interaction (face-to-face or virtual) will be documented on REDCap®. If relevant, the impact of face-to-face vs. remote education on outcomes will be explored.

  3. (iii)

    Resource outcomes: These will include assessing the time taken to implement study procedures (e.g. education on the use of Omnipod DASH® System) and determining whether the data collection tools selected (KeyLead™ and REDCap®) are appropriate and if so optimisation of their implementation. Data for the time taken for each study visit will be documented on REDCap®. Participants will be asked open-ended questions about their involvement with the study procedures. Although no formal qualitative analysis will be conducted, this will further inform the feasibility and acceptability of study procedures and identify possible areas of improvement. Device-related outcomes that provide information about the performance of Omnipod DASH® System will be assessed, including the length of time taken for an action to be completed such as changing insulin pump infusion sets (for participants on IPT), changing Omnipod DASH® Pods and uploading pump data for IPT and Omnipod DASH® System. The frequency of insulin infusion line failures (for those on conventional insulin pumps) and Omnipod DASH® System failures will also be recorded in real-time (Table 4). A failure is pre-defined as an unexplained BG of more than 14mmol/L that does not reduce by at least 2.8mmol/L within 1 h after a corrective insulin dose, or a BG of more than 14mmol/L with elevated ketones of at least 0.6mmol/L, or an occurrence of a non-resolvable pump occlusion/failure alarm [16]. Device-related outcomes will be recorded in real-time using the KeyLead Health™ phone application to minimise recollection bias

Table 3 Analysis of questionnaires
Table 4 KeyLead Health™ phone application buttons/modules

Secondary participant-centred outcomes

Participant-reported outcomes for diabetes distress, sleep quality, fear of hypoglycaemia and diabetes treatment experience will be assessed through six validated questionnaires administered which have been employed in prior diabetes technology studies [6, 15, 17,18,19] at 12 and 24 weeks (Table 3). These include the Problem Areas in Diabetes (PAID) questionnaire [6], Pittsburgh Sleep Quality Index (PSQI) [15], Hypoglycaemia Fear Survey Short-Form (HFS-II-SF) [15], Diabetes Medication System Rating Questionnaire Short-Form (DMSRQ-SF) [17], User Evaluation Questionnaire (UEQ) [18] and System Usability Scale (SUS) [19]. The short forms of HFS-II and DMSRQ will be used to reduce participant burden, which will help to ensure that data collection is complete. Participants allocated to intervention will be interviewed regarding their experience with and their recommendations for further refinement of the Omnipod DASH® System 12 weeks post-randomisation. The interview topic guide for participant interviews will be informed by literature regarding barriers, suboptimal use and discontinuation of IPT [4]. The interviews will be implemented by study nurses and audio-recorded at week 12 for participants randomised to Omnipod DASH® System and at week 24 following the extension for those initially randomised to usual care. The interviews are expected to take approximately 20 min. The audio recordings will be professionally transcribed by a native English speaker for later thematic analysis using Braun and Clarke’s six-phase approach [20].

Secondary healthcare professional perceptions

The perceptions of the physicians and diabetes nurse educators involved in the conduct of the study will be assessed through interviews. Two healthcare professionals (one doctor and one nurse) from each of the four study sites will be interviewed about their perceptions of the Omnipod DASH® System and their recommendations for further refinement at study completion. The topic guide for the healthcare professionals will be informed by analysis of the participant interviews and research that has explored clinician attitudes towards technology use [21, 22]. The interviews will be audio-recorded and transcribed for later thematic analysis [20].

Secondary glycaemic outcomes

Masked-CGM outcomes will be assessed using international consensus metrics [23, 24] allowing comparison with other studies. All participants will wear a standardised masked-CGM system (Abbott FreeStyle® Libre™ Pro Flash Glucose Monitoring device [Abbott Diabetes Care, Alameda, California]) for 2 weeks pre-randomisation (week -2 to 0), at weeks 10 to 12 and at weeks 22 to 24. HbA1c will be measured pre-randomisation, week 12, and week 24 in alignment with CGM data collection and assayed by a single DCCT-aligned laboratory. Comparisons will be made between study arms in week 12. A change in CGM time in the target range and change in HbA1c of 5% and 0.3%, respectively, are clinically significant [6].

Hypoglycaemic events, defined as an event where typical symptoms of hypoglycaemia are accompanied by a finger-prick blood glucose (BG) measurement of less than 3.9mmol/L [16], will be recorded in real-time using the KeyLead Health™ (Melbourne, Australia) mobile application (Table 4). The replacement of paper diaries with an electronic platform is expected to improve adherence and accuracy (refer to Methods: Data Collection for details regarding KeyLead Health™).

Sample size

The exploratory nature of the study precludes a power calculation because there is no prior randomised controlled study using Diabetes Technology Questionnaire ‘current’ (DTQ-current) score to assess treatment satisfaction with insulin patch pump devices. The sample size for this pilot study was determined as per time and resources. We anticipate that 64 participants with an estimated dropout of 10%, based upon prior experience with insulin pump studies, will provide exploratory data for assessment of primary feasibility outcomes for study completion rates and acceptability of the Omnipod DASH® System device, and protocol implementation across clinical sites and insulin delivery modalities (MDI and IPT). Data collected in this pilot study will allow for the determination of sample size for a future full-scale randomised controlled study with an effect size of 80% power.


Participants will be recruited from four major adult diabetes centres. Each clinical centre was chosen based on documentation for patient availability. The total pool of people with T1D attending these centres is approximately 3000. CGM uptake in the T1D population attending these centres is greater than in the general T1D population and estimated at approximately 50%. Nevertheless, this will provide a pool of about 1500 individuals from whom 64 will be recruited for the study. As insulin patch pumps are not currently in Australia as of study commencement and with the study extension all participants recruited will be provided with the opportunity to experience a novel insulin delivery platform, we expect that involvement in the study will appeal to the person living with T1D. With the uncertainties associated with the COVID-19 pandemic, we expect recruitment of the 64 participants to take approximately 18 months.

Source link