Mixed Methods Study Exploring the Impact of Hybrid Closed-loop Systems on Patient Reported Outcomes in People With Type I Diabetes and Their Partners
Mixed Methods Study Exploring the Impact of Hybrid Closed-loop Systems on Patient Reported Outcomes in People With Type I Diabetes and Their Partners
Type 1 diabetes requires people with type 1 diabetes to constantly monitor their blood glucose and adjust their insulin doses to try and keep glucose levels in range. Hybrid closed loop systems (also called artificial pancreas) consist of a sensor that continuously measures glucose levels, an algorithm that then decides every few minutes how much insulin to deliver and an insulin pump that then delivers the insulin. These systems have been shown in observational and randomized studies to improve glucose levels and have high levels of user satisfaction.
The National Health Service (NHS) in the UK has launched a pilot for up to 1,000 people with type 1 diabetes to trial these devices for up to 1 year and the results of that pilot will be audited by clinical teams. This study will recruit people from that pilot and use well validated as well as bespoke questionnaires to assess the impact of this technology on various key patient reported outcomes such as depression, diabetes related distress and fear of hypoglycemia (low blood glucose). We will invite a small number of participants to take part in semi-structured interviews to gain a more detailed understanding of the benefits and challenges of using these devices, and the impact they have on people living with type 1 diabetes. We know that living with a partner with type 1 diabetes can also have a negative impact on quality of life, and so we will invite partners of people using the hybrid closed loops systems to complete some validated and bespoke questionnaires. A small sample will also be invited to participate in interviews.
The data from this real-world study of the impact of hybrid closed loop systems on patient reported outcomes will be of value to clinicians, people with type 1 diabetes and policy makers in understanding the value of these systems to people with type 1 diabetes and their families.
Background:
Type 1 diabetes (T1D) is a complex lifelong autoimmune condition, characterized by the destruction of the β-cells of the pancreas. There is no cure. Management of type I diabetes mandates the use of exogenous insulin and frequent (>8 times per day) monitoring of blood glucose levels to get good glycaemic control. Over 42 factors affect blood glucose levels (www.diatribe.org). Factors that affect blood glucose levels include but are not limited to: exercise/activity, dietary intake, other medications, other diseases, environment, behaviour, and decisions. The correct dosing of insulin to achieve optimal blood glucose control is complex, requiring the person to take multiple predictable and often unpredictable factors into account.
Ongoing lifelong self-management is a relentless burden, with people often finding it difficult to cope [1] - there is "no holiday". T1D is associated with significantly lower quality of life and daily functioning and has been associated with increased rates of depression[2, 3], not just in people with diabetes but also in their partners [4]. These associations with diabetes specific distress, depression, anxiety and overall well-being are reported to be worse with higher HbA1c [5] (a measure of long-term glucose control that is associated with the risk of complications). Diabetes distress as the "significant negative psychological reactions related to emotional burdens and worries specific to an individual's experience in having to manage a severe complicated and demanding chronic disease such as diabetes." (ADA) [6] Multiple factors including constant need to make important timely decisions about insulin doses, blood glucose levels, food intake, predicted/unpredicted activity and the ongoing fear of developing long-term complications of diabetes [7]. Diabetes distress is prevalent (9-month incidence 54.4% among adults with type I diabetes) [8]. High diabetes distress scores coupled with the unrelenting nature of diabetes are associated with higher HbA1c, self-management and dietary/lifestyle choices [9].
Hypoglycaemia is the commonest acute complication of insulin therapy for type 1 diabetes and is widely acknowledged as a major barrier to achieving optimal glucose control. Fear of hypoglycaemia is also a major factor contributing to anxiety and diabetes distress, and that often is a barrier for people to achieve target glucose levels.
NICE currently recommends multiple daily injections as the therapy of choice initially for all adults with T1D. This requires them to measure their glucose 4-10 times a day and use self-management skills such as carbohydrate counting and appropriate adjustments for exercise, illness and stress etc to adjust amounts of insulin given by multiple injections through the day. The insulin usually consists of long-acting insulin taken once or twice daily (basal), and rapid acting insulin taken with each meal or to correct high glucose readings between meals (boluses). In recent years, a number of new technologies have emerged to support people with T1D to achieve better glucose control and also improve quality of life and reduce the burden of living with diabetes. Insulin pumps, also called continuous subcutaneous insulin infusion (CSII) are small pager sized devices that continuously infuse rapid acting insulin to provide the basal insulin, and the user can use buttons on the machine to take small and precise boluses to cover meals or corrections.
In terms of glucose monitoring, over the last 2 years, the use of Flash glucose monitoring has become widespread. This consists of a small disc shaped device, the size of a £2 coin that can measure glucose in the subcutaneous tissue continuously. The user has to "swipe" a reader or their phone over the device to obtain a glucose value, and also get information about the last 8 hours of glucose and which direction it is going in (up or down). Some people also have access to real-time continuous glucose monitors that provide a continuous read out on a reader or a smart watch and can also have alarms or alerts that warn the individual when glucose levels reach predetermined levels. These technologies have all been shown the reduce HbA1c as well as hypoglycaemia, and improve key patient reported outcomes such as quality of life, diabetes distress and fear of hypoglycaemia.
A hybrid closed loop (HCL) system (also known as Automated Insulin Delivery or artificial pancreas) is the next step and the most novel therapy for the management of type 1 diabetes. A HCL system combines real-time continuous glucose monitoring (rtCGM) with an insulin pump. Each HCL system has a different algorithm which adjusts the amount of insulin delivered by the pump every few minutes using real-time blood glucose data from the rtCGM - to help keep blood glucose levels in a normal range [10]. Each system requires different quantities of user interaction; all systems suggest that to achieve better glycaemic control the user inputs the quantity of carbohydrates (in their food) - so that the person can give an insulin bolus for their meal. Commercially available HCLs include Medtronic 780G, Tandem Control-IQ Hybrid Closed-Loop System and CamAPS FX. The iDCL Trial Research Group have shown that the use of hybrid closed-loop systems for six months in people with type 1 diabetes is associated with a greater percentage time in range and improves glycaemic control [11]. There are three HCL systems that are currently commercially available in the UK: the Medtronic 780G system, the Tandem ControliQ system and the CAMAPS Fx system. All have been tested in clinical trials and are approved by the MHRA [12].
Although previous studies have assessed the benefits of HCL on glucose control, there are limited data from real-world use of their impact on key patient reported outcomes. The iDCL study showed that in people with type 1 diabetes in the USA, use of HCL was associated with some improved patient reported outcomes [13] but these data may not reflect the benefits seen in real-world scenarios that include people with much higher starting HbA1c.
The NHS has recently announced a pilot roll out of commercially available hybrid closed loop systems to up to 1000 people living with type 1 diabetes who are currently have suboptimal glucose control HbA1c 8.5% (70mmol/mol) or greater despite current treatment with insulin pumps as well as flash glucose monitoring (or continuous glucose monitoring)[14]. This pilot will include 21 adult sites listed in table 1. Up to 50 people with T1D at each site will be offered the opportunity to use a commercially approved HCL system for up to a year. The glucose data (HbA1c, time in ranges (time over 10 mmol/l, time between 3.9-10mmol/l, time below 3.0 and 3.9 mmol/l) as well as outcomes such as admissions due to diabetic ketoacidosis (DKA) or severe hypoglycaemia (SH) are being collected as part of a national audit led by the Association of British Clinical Diabetologists (ABCD) who also completed audits of national roll out of other therapies.
However, this audit will not collect patient reported outcomes that are not part of routine clinical data collection.
Table 1: List of adult sites participating in the NHS HCL pilot.
Birmingham Hospital Trust County Durham and Darlington HT Derby and Burton University HT Guys and St Thomas Harrogate District Hospital Imperial College London Ipswich Hospital Trust Kings College Hospital Trust Leicester University Hospital Manchester Royal Newcastle Hospital North Bristol Hospital Trust Oxford University Hospital Portsmouth Hospital University Trust Royal Liverpool Royal Surrey Hospital Trust Sheffield Hospital Trust St Helens and Knowsley Sussex University Hospital University Hospital Dorset Warrington and Halton Hospital Trust West Hertfordshire Hospital Trust Whittington Hospital Yeovil District Foundation Trust
This pilot provides us with the ideal opportunity to explore the impact of HCL on patient reported outcomes in a real-world setting and review outcomes that are important to people with type 1 diabetes and their families. Due to the significant cost and user engagement required for excellent blood glucose outcomes, it is important that we explore the user experience (either positive or negative). Outcomes of interest include diabetes distress, depression, fear of hypoglycaemia and system usability. Should HCL result in improvements in these markers, it has the potential to influence the ongoing funding of HCL by the NHS and other health organisations internationally.
This study will differ to other published studies of HCL in the literature, as those taking part will already be using insulin pumps and flash glucose monitoring, but crucially will have much higher starting glucose levels than reported in other studies.
Internationally, there are real world studies looking at the impact of different HCL systems on glycaemic indices. However, there is limited literature on the impact of HCL on patient reported outcomes.
We propose a mixed methods study that will enrol people from the NHS HCL pilot and consist of two parts. Initially, we had planned to recruit 190 people with diabetes going on HCL therapy as part of the NHS pilot. Unfortunately, due to sudden announcement of the NHS pilot and the short window for onboarding people with diabetes, it was only possible to recruit 17 participants to our original study design (i.e. the participant completed a questionnaire before starting HCL therapy) . The majority of people were already on HCL therapy when recruitment started.
Thus, to capture changes in patient reported outcomes, we plan to collect data solely after the person has been on HCL therapy for >3 months -a cross-sectional study of people on HCL therapy and their partners. The first part will involve asking people with type I diabetes and PP1D to complete a series of validated questionnaires and a bespoke questionnaire on a variety of outcomes which have been shown to be important to people with type I diabetes and their families. We will ask participants to complete these questionnaires >3 months after starting HCL therapy. Participants will be given the opportunity to express interest in taking part in a one off semi-structured interview. A subset of these participants (and associated PP1D) will be invited to participate in qualitative interviews at 3 months following initiation of the HCL. The interviews will explore a series of questions regarding the persons experience and outcomes of the HCL.
Rationale:
HCL is an expensive technology that requires significant user input to work successfully. While we know that HCL improves glycaemic control in people with type 1 diabetes, there is limited literature regarding the impact of HCL on peoples' day-to-day lives particularly in the UK for example sleep, fear of hypoglycaemia, diabetes distress. The information obtained from answering our important research question will provide key data for policymakers regarding the impact of HCL on the lives of people living with diabetes and their partners and the importance of continued funding of HCL therapy.
From reviewing the literature, we have identified significant knowledge gaps in terms of qualitative research in HCL. We envisage that the interviews conducted as part of this research with both a person living with type I diabetes and paired PP1D will add significantly to the existing literature and provide further insight into the psychosocial impact of HCL as well as individuals daily activities and relationship with the PP1D.
Research Question:
The main research question of this study is to assess the impact of using a hybrid closed loop insulin delivery system patient reported outcomes in people with type 1 diabetes and their partners.
Objective:
To perform a mixed-methods study that will assess the impact of using a hybrid closed loop insulin delivery system patient reported outcomes in people with type 1 diabetes and their partners.
Methodology:
This is a cross-sectional mixed-method study.
We will invite people with T1D who have been selected by their clinical teams to take part in the NHS England funded pilot of HCL treatment. We have developed a flyer that will provide a link email address for the PRO-HCL study. This flyer will be shared by the clinical teams and also via social media e.g. the Leicester Diabetes Centre, ABCD, DTN and Kings Facebook and Twitter pages. When potential participants email this address, we will send them a link to a secure website that has the patient information sheet, and the on-line consent form. Participants will have the option to email the study team and ask any questions they may have about the study. Once participants have completed the on-line informed consent, they will be asked to complete some personal details on the form that includes their email address, age, ethnicity, type of diabetes, relationship status, employment status, the site they are being treated at and which closed loop system they are using. We will also collect self - reported data on estimated HBA1c, time in range and % time in hypoglycaemia which patients are able to read off their libre sensor app/CGM. Participants will be asked to estimate their typical glycaemic range prior to starting HCL therapy. Participants will be asked to estimate their duration of diabetes and time since starting HCL therapy. Participants will be asked about their previous knowledge of HCL therapy, where they heard about HCL therapy, why they decided to start, any recommendations re the process of starting and any other additional comments re HCL therapy. Participants who provide optional details for us to contact their partners will be able to enter their partners contact details. They will then go onto answer baseline questionnaires
Part 1:
Questionnaires: we will use a series of validated questionnaires and bespoke questionnaire to assess key aspects of patient reported outcomes that may be affected by use of hybrid closed loop systems. We will also provide links to people with T1D to invite their partners to also complete 3 three validated questionnaires that are specifically developed for partners of people with T1D and one bespoke questionnaire..
Part 2:
Semi-structured Qualitative interviews: Using purposive sampling to ensure we get participants from a range of centres and using different devices, we will invite 15 people with T1D and 15 partners to take part in semi-structured interviews.
Part 3:
Participants will be invited to share self-reported measures of their diabetes control (time very high, time high, time in target range, time low and time very low), average glucose, glucose management indicator (GMI), glucose variability, % time CGM is active). Participants will be asked to estimate their typical glycaemic range prior to starting HCL therapy.
Questionnaires:
The questionnaires will be hosted on a secure online platform (www.onlinesurveys.ac.uk run by Jisc) that is widely used for questionnaire studies.
The questionnaires to be administered to people with type 1 diabetes who will participate in the NHS HCL pilot are:
Gold Score: This widely used in the clinical environment and is a well validated instrument to assess the risk of severe hypoglycaemia [15].
PHQ-9: This is a validated and widely used tool for screening for depression and is widely used in the literature and in the NHS [16].
DDS-T1: This 28-item questionnaire is used to measure diabetes distress in people with T1D.
Hypoglycaemia Confidence Scale: This scale was developed to measure fear of hypoglycaemia and has been used in another large trial of HCL - the iDCL [13].
Inspire Survey: The INSPIRE survey was developed specifically for use in studies using automated insulin delivery or hybrid close lips and is recommended by the FDA.
EQ5D-5L: This is a widely used questionnaire that measures the impact of a chronic health condition on daily functioning. This tool is widely used by reimbursement authorities when considering the health economic impact of an intervention.
Diabetes Treatment Satisfaction Questionnaire (DTSQs) and Diabetes Treatment Satisfaction Questionnaire change version (DTSQc) : The questionnaire, which is widely used in clinical trials, is designed to assess the change in satisfaction with a new diabetes treatment.[17] Bespoke questionnaire: bespoke questions to assess how HCL therapy has impacted patient reported outcomes using domains that have been shown to be important to people with diabetes [18, 19] System Usability Scale: this is a reliable tool for measuring usability of novel devices and consists of a tenant questionnaire with five possible response options. It allows us to evaluate a variety of products and their components.
These questionnaires will take participants 25-45 minutes to complete.
Partners of people with T1D (PPT1D) will be invited to complete three validated questionnaires that have been specifically developed for partners and one bespoke questionnaire:
The participant details will initially be stored on www.onlinesurveys.ac.uk (run by Jisc) and then transferred to secure servers at the University of Leicester. Data from all questionnaires will then be stored on a secure server in the University of Leicester. All personal data on www.onlinesurveys.ac.uk (run by Jisc) will be deleted after 15 months. Each participant with type 1 diabetes will be assigned a unique anonymous identifier. The PP1D will be assigned an associated anonymous unique identifier. The code for anonymization will be stored separately on an excel file in a password protected computer in the University of Leicester.
Semi Structured Interviews:
Approximately 15 participants with type 1 diabetes and 15 PP1D will be asked to participate in semi-structured qualitative interviews at >3 months after the person with type 1 diabetes starts the HCL. Informed consent will be taken before commencement of the interviews. All participants will have the option to withdraw from the study at any time without their usual care being affected. Interviews will be carried out by telephone or by video call (Microsoft teams). This will depend on the preferences of the participants and on COVID-19 regulations at the time of interview. The interview will take place in a closed-door office for confidentiality and privacy purposes.
At the start of the interview, the interviewer will state the participants ID number to ensure the participant remains anonymous. The interviews will be audio recorded on an encrypted digital voice recorder and transcribed verbatim by an independent transcriber. Audio recordings will be deleted once transcribed. The transcriptions will be stored securely in a password secured University of Leicester drive. Details will be anonymised.
Topics that will be covered in the interview include but are not limited to the effect of the HCL on glucose control, time spent managing diabetes, device specific issues, life outside diabetes (relationships, work, and leisure), mood, well-being, sleep, drawbacks and challenges of the systems. It is estimated that the interview will take between 30 and 60 minutes. Each interview will be facilitated by a trained qualitative researcher. The interviewer will be based at University of Leicester. The researcher will be supervised by a qualitative lead, both based at the Leicester diabetes research Centre.
The interviews will not include topics that are sensitive or embarrassing to the individual and we do not anticipate that the participants will be upset by the questions. If a participant becomes upset, the researcher will stop the interview, and one of the members of the research team will offer immediate support to assess if they are feeling okay and safe. If the participant can carry on, then the interview will resume. If they are unable to carry on, the interview will be discontinued (or participant is removed from discussion). They will be encouraged to contact their GP if needed, or to contact a member of the healthcare team for further advice/support.
Data Analyses for questionnaires:
Dr Tomás Griffin will lead on data analyses. In brief, all analyses will be performed using R V3.2.0/SPSS/Minitab/GraphPad. Summary statistics of continuous variables will be given by mean (standard deviation) and frequencies (percentages) for categorical variables. Summary statistics of data which is not normally distributed will be provided by medians (interquartile range and minimum/maximum).
Analysis of semi-structured interviews
Data from the qualitative interviews will be analysed using thematic analyses, informed by the method of constant comparison. Using a qualitative data indexing software (NVivo12), data will be analysed, and key themes will be generated.
Study Setting:
This is a phenomenological study with remote data collection. People with type 1 diabetes who are participating in the NHS England supported pilot roll out of hybrid closed loop insulin pump therapy at 21 adult sites will be invited to participate. Clinicians at the clinical sites will provide people who are participating in the HCL pilot or who are on HCL therapy with a paper flyer for the study at clinic or email the flyer to the participant. The flyer will include an email address that participants in the HCL pilot will be invited to email to receive more details on the study. This flyer will also be shared via social media e.g. the Leicester Diabetes Centre, ABCD, DTN and Kings Facebook and Twitter pages, patient advocacy groups and national/international organisations such as DTN. Once the person on HCL therapy contacts the email address, they will be sent a link to the patient information leaflet and consent form. The research team will be based within the Diabetes Research Center at University of Leicester. All data will be collected and stored at the Leicester site.
Inclusion Criteria:
Exclusion Criteria:
1. Diagnosis Type II or other form of diabetes