Diabetic Medicine - 2022 - Speight - Impact of Glycaemic Technologies On Quality of Life and Related Outcomes in Adults

Received: 7 February 2022
| Accepted: 18 August 2022
DOI: 10.1111/dme.14944
REVIEW ARTICLE
Impact of glycaemic technologies on quality of life and

related outcomes in adults with type 1 diabetes:
A narrative review
Jane Speight1,2 | Pratik Choudhary3 | Emma G. Wilmot4,5 |

Christel Hendrieckx1,2 | Hannah Forde3 | Wai Yee Cheung6 |
Thomas Crabtree4,5 | Bekki Millar7 | Gemma Traviss-Turner8 | Andrew Hill8 |
Ramzi A. Ajjan9
1
School of Psychology, Deakin University, Geelong, Australia
2
The Australian Centre for Behavioural Research in Diabetes, Diabetes Victoria, Melbourne, Australia
3
Diabetes Research Centre, University of Leicester, Leicester, UK
4
Department of Diabetes, University Hospitals of Derby and Burton NHS Foundation Trust, Derby, UK
5
Faculty of Medicine and Health Sciences, University of Nottingham, Nottingham, UK
6
Diabetes Research Unit Cymru, Swansea University Medical School, Swansea, UK
7
Diabetes Research Steering Group, Diabetes UK, London, UK
8
Leeds Institute of Health Sciences, University of Leeds, Leeds, UK
9
Leeds Institute of Cardiovascular and Metabolic Medicine, the LIGHT Laboratories, University of Leeds, Leeds, UK
Correspondence
Ramzi A. Ajjan, Leeds Institute Abstract
of Cardiovascular and Metabolic Aims: To explore the association between the use of glycaemic technologies and
Medicine, the LIGHT Laboratories,
person-reported outcomes (PROs) in adults with type 1 diabetes (T1D).
University of Leeds, Leeds, UK.
Email: r.ajjan@leeds.ac.uk Methods: We included T1D and technology publications reporting on PROs
since 2014. Only randomised controlled trials and cohort studies that used vali-
Funding information
Abbott Diabetes Care; Avacta Life
dated PRO measures (PROMs) were considered.
Sciences; BBSRC; NIHR; Diabetes Results: T1D studies reported on a broad range of validated PROMs, mainly as
UK; Diabetes Victoria and Deakin secondary outcome measures. Most studies examined continuous glucose moni-
University
toring (CGM), intermittently scanned CGM (isCGM), and the role of continu-
ous subcutaneous insulin infusion (CSII), including sensor-augmented CSII and
closed loop systems. Generally, studies demonstrated a positive impact of tech-
nology on hypoglycaemia-specific and diabetes-specific PROs, including reduced
fear of hypoglycaemia and diabetes distress, and greater satisfaction with diabetes
treatment. In contrast, generic PROMs (including measures of health/functional
status, emotional well-being, depressive symptoms, and sleep quality) were less
likely to demonstrate improvements associated with the use of glycaemic tech-
nologies. Several studies showed contradictory findings, which may relate to
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
© 2022 The Authors. Diabetic Medicine published by John Wiley & Sons Ltd on behalf of Diabetes UK.
Diabetic Medicine. 2023;40:e14944. wileyonlinelibrary.com/journal/dme | 1 of 23

https://doi.org/10.1111/dme.14944
|
14645491, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/dme.14944 by Cochrane Romania, Wiley Online Library on [31/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 of 23 SPEIGHT et al.
study design, population and length of follow-up. Differences in PRO findings

were apparent between randomised controlled trials and cohort studies, which
may be due to different populations studied and/or disparity between trial and
real-world conditions.
Conclusions: PROs are usually assessed as secondary outcomes in glycaemic
technology studies. Hypoglycaemia-specific and diabetes-specific, but not ge-
neric, PROs show the benefits of glycaemic technologies, and deserve a more
central role in future studies as well as routine clinical care.
KEYWORDS
continuous glucose monitoring (CGM), insulin pump, intermittently scanned continuous
glucose monitoring (isCGM), person-reported outcome measure (PROM), person-reported
outcomes (PROs), quality of life (QoL), type 1 diabetes
1 | I N T RO DU CT ION
What's new?
Type 1 diabetes (T1D) is a chronic condition with a sig- • While Patient Related Outcomes Measures
nificant self-management and health burden, requiring (PROMs) are important, this review demon-
frequent insulin administration and glucose monitoring.1 strates they are mainly studied as secondary
Given the overwhelming evidence that maintaining rec- outcomes in individuals with type 1 diabetes
ommended glycaemic targets reduces long-term compli- using technology to aid diabetes management.
cations,2 numerous studies focus on reducing glycated • Generally, studies show a positive impact of
haemoglobin (HbA1c). Recently, this focus has shifted to technology on diabetes-specific PROMs, with
additional glycaemic markers such as time in range, gly- limited effects, if any, on generic PROMs.
caemic variability and hypoglycaemic exposure, given the • PROMs deserve a more central role in type 1
advances and increasing accessibility of diabetes technol- diabetes technology studies, as well as clinical
ogies, and their additional prognostic value.3 management of these individuals.
While attention to glycaemia is unquestionably key
for preventing acute and long-term complications, it does
not take into account the person's experiences, priorities
and preferences, which are equally important.4 A partic- are moving from the periphery to the centre of clinical
ular challenge in T1D management is the relative lack of diabetes care with PROMs used increasingly for bench-
adverse symptoms associated with hyperglycaemia, such marking and in clinical quality registries.8
that quality of life (QoL) can be negatively impacted more Most clinical T1D studies focus on glycaemic markers,
in the short-term by the burden of intensified therapy than with PROMs relegated to secondary outcomes, if included
by above-target glucose levels.5 For at least two decades, it at all. Therefore, if the glycaemic effects of a particular
has been recognised that successful and sustainable ap- intervention are modest or non-significant, the study is
proaches to managing T1D must include strategies that often labelled as negative even if a clear improvement in
recognise and reduce the burden of self-management.6 PROs is demonstrated. A distinguishing feature of the UK
Person-reported outcome measures (PROMs) are stan- DAFNE trial was that it recognised the burden of T1D self-
dardised, validated questionnaires completed directly by management and included QoL as a co-primary end point
the individual living with the condition, enabling them to alongside HbA1c.9 Consequently, benefits for QoL were
share their perceptions and experiences of the condition afforded equal priority to improvements in HbA1c. This
and/or its treatment. This is crucial for person-centred is a salient lesson for technology studies that have shown
clinical care.7 PROMs can be used to assess the impact of a only modest improvements in glycaemic markers (usually
management strategy on satisfaction with treatment, and HbA1c), which has meant that these devices may not have
involvement in clinical care, as well as emotional well- been funded or subsidised by health authorities, despite
being, health status, and QoL. It is now appreciated that demonstrating favourable effects on PROMs. Conversely,
“adding life to years” is as important to many people as inappropriate selection of PROMs and/or misinterpreta-
“adding years to life”. Therefore, strategies to improve QoL tion of findings can mean that relevant benefits are not
|
SPEIGHT et al. 3 of 23
T A B L E 1 Definition of key constructs used when validating

demonstrated even when clinical experience would sug-
person-reported outcome measures (PROMS)
gest the contrary.10
The aim of this narrative review is to examine the Construct Definition
impact of diabetes technologies on person-reported out- Reliability The degree to which the measure is free
comes (assessed with validated PROMs) among adults from measurement error
with T1D, regardless of glycaemic outcomes. Internal The degree to which items in the measure
consistency are inter-related
Test–retest/ The ability to provide consistent scores
2 | S E A RC H ST R AT EGY Reproducibility over time in a stable population, i.e.
when no change in scores would be
We searched PubMed for T1D and terms synonymous expected
with PROMs and technology. Terms for PROMs included Validity The degree to which the measure assesses
commonly used measures such as the Diabetes Distress what it sets out to measure
Scale (DDS), Diabetes Treatment Satisfaction Questionnaire Content The extent to which the measure includes
(DTSQ), EuroQoL-5 Dimensions (EQ-5D), Hospital Anxiety the most relevant and important
and Depression Scale (HADS), Hypoglycaemia Fear Survey aspects of the construct(s) it sets out
to measure
(HFS), Problem Areas in Diabetes (PAID), Short Form 36
items (SF-36) and Patient Health Questionnaire (PHQ), as Structural The degree to which the relationships
among items reflect the theoretical
well as quality of life (QoL) as an umbrella term used to
framework, i.e. how well each
describe one or more PROMs. Technology terms included individual item maps to expected
insulin pumps, continuous glucose monitoring (CGM), constructs to form scales/subscales
flash glucose monitoring, intermittently scanned CGM and Construct The degree to which scores relate to
hybrid closed loop, artificial pancreas or automated insulin other measures in a manner that is
delivery systems. Our search strategy may have missed consistent with a priori hypotheses
PROMs that are rarely used or if these measures were not concerning the concepts measured
apparent in the title/abstract. Convergent The degree to which the measure is
As technology use in T1D has rapidly expanded in the related to similar measures
past 6–8 years, we limited our search to articles published Divergent Demonstration that the measure is
in English since 2014. We checked reference lists of rele- unrelated to other measures that it is
vant articles for additional studies and included older ar- not expected to have a relationship
ticles in the review if relevant. We focused on randomised with
controlled trials and longitudinal cohort studies, exclud- Known groups The degree to which scores differentiate
ing cross-sectional studies. It is beyond the scope of this between groups in the population
expected to differ on that construct
review to fully synthesise qualitative studies, though rele-
vant publications are cited. Criterion The degree to which the measure is an
adequate reflection of a ‘gold standard’
Each author performed a search for their technology
measure
section, and the last author performed an independent
Responsiveness/ The extent to which a PROM can detect
search to ensure all relevant studies were included.
Sensitivity to changes in the construct being
change measured over time
3 | P E R S O N-R E PORT E D
O U TCO M E ME ASU R E S (PROM S) the constructs to be assessed, item generation, and for de-
briefing PROMs (e.g., understandability, comprehensive-
Determining the suitability of PROMS involves assess- ness, redundancy and ease of completion).
ing how well the subjective latent constructs can be re-
ported as reliable and valid measures.11 The statistical
methods for validating PROMs have been defined in the 4 | THE IMPACT OF DIABETE S
COnsensus-based Standards for the selection of health TECHNOLOGIES ON PROs BY
Measurement Instruments (COSMIN) initiative.12 Table 1 TECHNOLOGY TYPE
summarises key constructs used in the psychometric
validation of PROMs. The development and validation of Key studies examining the impact of diabetes technologies
PROMs requires multidisciplinary collaboration. Patient using PROMs are summarised in Tables 2, 3 and discussed
and public involvement (PPI) is important for determining in the following sections.
T A B L E 2 Person-reported outcome measure (PROM) findings by technology comparison/type and study: Continuous subcutaneous insulin infusion, sensor-augmented pump and closed
4 of 23
loop studies
|
Study name Study duration &

Technologya (Citation) designb (N) Inclusion criteriac PROMsd Findings

13,14
CSII vs. MDI HypoCOMPaSS 6-month multi-centre IAH ± recurrent SH DTSQ, HFS-II, ITSQ, At 6 months: between-group differences (t-tests) in CSII group diabetes
RCT plus 18-month Gold, Clarke, treatment satisfaction (DTSQ) and satisfaction with insulin
observation (24 m HypoA-Q delivery device (ITSQ Delivery Device subscale) was higher than
total) (N = 96 at 6 m; in the MDI group (both p < 0.001). There were no between-group
N = 76 at 24 m) differences for fear of hypoglycaemia (HFS-II), IAH (Gold, Clarke,
HypoA-Q), perceived frequency of hypo-hyperglycaemia (DTSQ
items 2/3).
At 24 months (observation during which individuals continued
with their preferred technology): improvements maintained; no
significant differences between groups for any of the PROMs.
REPOSE15 24-month cluster RCT N/A SF-12, DSQOL, At 6 months: no statistically significant difference between CSII and
with data analysed WHOQOL-BREF, MDI groups.
at 6, 12 and 24 m HFS, HADS, EQ- At 12 months: CSII group reported greater treatment satisfaction
(N = 317; N = 245 at 5D, DTSQ (DTSQ) (p < 0.001) and greater improvements in diabetes-
6 m; N = 237 at 12 m specific QoL (DSQOLS) in subdomains: daily hassle or functions
and 24 m) (p = 0.019), diet restriction (p = 0.010).
At 24 months: CSII group reported greater treatment satisfaction
(DTSQ) (p < 0.001); greater improvement in diabetes-specific
QoL (DSQOLS) total (p = 0.006) and in subdomains: leisure time
restrictions and flexibility (p = 0.016), daily hassle or functions
(p = 0.006), diet restriction (p = 0.004); and reduced FoH (HFS
worry scale only) (p = 0.010).
No between-group differences at any timepoints in generic health
status (SF-12 MCS or PCS scores, EQ-5D) and quality of life
(WHOQOL-BREF) or depression and anxiety symptoms (HADS).
CSII only Oldham et al16 Baseline (start CSII) and N/A PAID At 3–6 months follow up: significant reduction in diabetes distress
(Cohort 12 m FU (N = 47) (p = 0.0002)
study) At 6–12 months follow up: significant reduction in diabetes distress
(p < 0.00001), but difference between two FU points was not
significant (p = 0.07).
SPEIGHT et al.
TABLE 2 (Continued)
SPEIGHT et al.
SAP+PLGS SMILE17 6-month open-label, HbA1c 5.8–10% (40– DTSQ(s + c), HFS, At 6 months: SAP group showed increased treatment satisfaction
vs. CSII/ multi-centre and 86 mmol/mol), Gold, Clarke (DTSQc only) (2.3 vs. 1.6; p < 0.0001) compared to CSII/SMBG
SMBG multi-country RCT IAH or SH past group. SAP group also reported reduced fear of hypoglycaemia
(N = 153) year (HFS: Total − 17.6 vs. −7.2; p = 0.0010, Worry subscale −13.5 vs.
CSII −5.9; −4.1 vs. −1.3 p = 0.0004; Behaviour subscale p = 0.043).
No between-group differences in IAH (Gold and Clarke).
HCL vs. Australian HCL 6-month RCT HbA1c DTSQ, PAID, DIDP, At 6 months: HCL group had greater diabetes-specific positive well-
current trial18 (N = 120) ≤10.5% (≤91 mmol/ PSQI, PRMQ, being (W-BQ28 subscale: p = 0.0048) and diabetes-specific quality
therapy mol) 50% MDI/ W-BQ28 subscale: of life (DIDP) (p = 0.023) compared to CSII/MDI group.
(CSII or CSII at baseline diabetes-specific No between-group differences for diabetes treatment satisfaction
MDI) positive wellbeing (DTSQ), diabetes distress (PAID), sleep quality (PSQI) or memory
(PRMQ).
HCL vs Diabeloop19 12-week multi-centre, HbA1c ≤10% DTSQ At 12 weeks: No between-group difference in treatment satisfaction
SAP open-label crossover (86 mmol/mol) (DTSQ).
RCT with 8-week CSII
washout IAH & SH excluded
(N = 63)
Kropff et al20 2-month, multi-centre, HbA1c 7.5–10% HFS-II, DTSQ(s + c) At 2 months: No between-group differences for treatment satisfaction
crossover RCT (58–86 mmol/ (DTSQ Total or DTSQ Perceived Frequency of Hyperglycaemia or
evening/night HCL mol) Hypoglycaemia items) or fear of hypoglycaemia (HFS-II Total, and
(2x8 weeks) with SH excluded Worry and Behaviour subscales).
4-week washout
(n = 32)
Bisio et al21 Single-arm 2x4‑week SAP HbA1c < 10%, no SH HFS-II, DDS, CES-D, At 4 weeks: HCL use associated with reduced diabetes distress (DDS:
and HCL (N = 15) history PSQI p = 0.046).
Age: ≥65 years No between-group differences for fear of hypoglycaemia (HFS-II),
depressive symptoms (CES-D), or sleep quality (PSQI).
(Continues)
|
5 of 23
TABLE 2 (Continued)
| 6 of 23


HCL only Real-world HCL Observational study. N/A WHO-5, DIDS At 4 weeks: reduction in impact of diabetes on person's life (DIDS:
(Cohort Pinsker et al22 T1: at least 3 weeks p < 0.01), improvement in device-related satisfaction (DIDS:
studies) after start using HCL. p < 0.001), but a reduction in general emotional well-being (WHO-
T2: 4 weeks after T1; 5) (p < 0.001).
N = 1435. ≥14 years;
95% ≥18 years
Real-world HCL 3-month multi-centre N/A HFS, DDS, DQoL, At 3 months: reduced fear of hypoglycaemia (HFS Total p = 0.005;
Beato-Vibora et al23 prospective study DTS(Q), PSQI, Worry subscale p = 0.016; Behaviour subscale p = 0.024) and
Polonsky et al24 (N = 58; 34 adults, 22 Gold, Clarke diabetes distress (DDS scores: p = 0.002); improved diabetes-
children). At baseline, DDS, HCS, DTSQ, specific quality of life (DQoL: p < 0.001), diabetes treatment
using MDI ± CGM, IDDS, WHO-5, satisfaction (DTS(Q): p = 0.037), sleep quality (PSQI: from 49% to
CSII+SMBG or PSQI 40% ‘poor sleepers’; p = 0.004) and awareness of hypoglycaemia
SAP-PLGS (Clarke p = 0.023; Gold score not reported separately).
3-month prospective, At 3 months: reduced diabetes distress (DDS: p < 0.0001), and
single-arm study using improvements in confidence in managing hypoglycaemia (HCS:
Omnipod 5 automated p = 0.0002), satisfaction with insulin device (IDDS: p = 0.0007),
insulin delivery and satisfaction with diabetes treatment (DTSQ: p < 0.0001); but
system (N = 115 no changes observed in general emotional well-being (WHO-5:
adults) p = 0.7912) or perceived sleep quality (PSQI: p = 0.4217).
a
Technology: rtCGM, real-time continuous glucose monitoring; CSII, continuous subcutaneous insulin infusion; DBLG1, diabeloop generation 1; HCL, hybrid closed loop; isCGM, intermittently scanned CGM, MDI,
multiple daily injections; PLGS, predictive low glucose suspend; SAP, sensor-augmented pump therapy; SMBG, self-monitoring of blood glucose.
b
RCT, randomised controlled trial. In RCTs, comparisons are made between study arms at follow-up, while in single-arm studies, comparisons are made between baseline and study end.
c
Specific inclusion criteria beyond ‘adult with type 1 diabetes’: IAH, impaired awareness of hypoglycaemia; SH, severe hypoglycaemia.
d
PROMs: BGMSRQ, Blood Glucose Monitoring System Rating Questionnaire; CES-D, Centre for Epidemiological Studies Depression Scale; CIDS(−s), Confidence in Diabetes Self-Care (sensor subscale); DDS, Diabetes
Distress Scale; DIDS, Diabetes Impact and Devices Satisfaction; DIDP, Dawn2 Impact of Diabetes Profile; DQOL, Diabetes Quality Of Life; DSQOL, Diabetes-Specific Quality of Life; DTQ, Diabetes Technology
Questionnaire; DTSQ(s + c), Diabetes Treatment Satisfaction Questionnaire (status and change versions); EQ-5D, EuroQoL five dimensions; GME-Q, Glucose Monitoring Experience Questionnaire; GMSS, Glucose
Monitoring Satisfaction Scale; HADS, Hospital Anxiety and Depression Scale; HCS, Hypoglycaemic Confidence Scale; HFS, Hypoglycaemia Fear Survey; HypoA-Q, Hypoglycaemia Awareness Questionnaire; IDDS:
Insulin Device Satisfaction Survey; ITSQ, Insulin Treatment Satisfaction Questionnaire; PAID (-Peds), Problem Areas in Diabetes (Pediatric version); PRMQ, Prospective and Retrospective Memory Questionnaire;
PSQI, Perceived Sleep Quality Index; SF-12, 12-item short-form health survey (MCS, Mental health Component Summary; PCS, Physical health Component Summary); W-BQ28, 28-item Well-Being Questionnaire;
WHO-5: Well-Being Index Scale; WHOQOL-BREF, World Health Organization Quality Of Life—BREF.
SPEIGHT et al.
TABLE 3 Person-reported outcome measure (PROM) findings by technology comparison, type and study: Continuous glucose monitoring studies

SPEIGHT et al.
13,14
CGM vs. SMBG HypoCOMPaSS 6-month multi-centre IAH ± recurrent SH DTSQ, HFS-II, GME- At 6 months and at 24 months (observation during which individuals
RCT plus 18-month Q, Gold, Clarke, continued with their preferred technology): no between-group
observation HypoA-Q differences for any PROMs.
(24 m total)
(N = 96 at 6 m; N = 76 at
24 m)
DIAMOND25 24-week prospective HbA1c 7.5–10.0% WHO-5, EQ-5D-5L, At 24 weeks: increase in confidence in managing hypoglycaemia (HCS)
multicentre RCT MDI DDS, HFS-II (p = 0.03; d = 0.40) in the CGM group compared to the SMBG group;
(N = 155) (Worry subscale), and a modest reduction in diabetes distress (DDS) in the CGM group
HCS, CGM and increase in the SMBG group (p = 0.03; d = 0.44).
Satisfaction No between-group differences detected in general emotional well-being
Survey (for CGM (WHO-5), general health status (EQ-5D-5L) or fear of hypoglycaemia
group only) (HFS-II ‘Worry’).
At 24 weeks: CGM satisfaction (measured at study-end only) was
moderately associated with all PROMs except EQ-5D-5L.
GOLD26 26-week multi-centre HbA1c ≥ 7.5% WHO-5, DTSQ (status At 26 weeks: increase in treatment satisfaction (DTSQ), emotional well-
cross-over RCT: 2 x 6 m (≥58 mmol/mol) and change) HFS, being (WHO-5) and confidence in managing hypoglycaemia (HCQ)
periods plus 4 m MDI PAID, HCQ while using CGM compared to SMBG (DTSQ 30.21 vs. 26.62, p < 0.001;
washout (N = 161) WHO-5 66.1 vs. 62.7, p = 0.02; HCQ 3.40 vs. 3.27, p < 0.001).
No between-group difference for fear of hypoglycaemia (HFS).
For diabetes distress (PAID) descriptive statistics reported only (with no
significance testing).
GOLD-327 69-week RCT; extension of HCQ At 69 weeks: greater confidence in managing hypoglycaemia in CGM
GOLD compared to the SMBG group (HCQ: 3.40 [95% CI 3.32–3.47] vs. 3.27
[95% CI 3.18–3.35]).
HypoDE28 6-month multi-centre History IAH or SH T1-DDS, HFS, EQ-5D, At 6 months: rtCGM group more satisfied with method of monitoring
open-label RCT past year GMSS, Clarke (GMSS) compared to the SMBG group. In both groups, fear of
(N = 149) MDI hypoglycaemia (HFS) and diabetes distress (T1-DDS) reduced
significantly from baseline, with between-group differences at
6 months apparent only for GMSS subscales: ‘openness’ (p = 0.003) and
‘behavioural burden’ (p = 0.046); and T1-DDS subscale: ‘hypoglycaemia
distress’ (p = 0.010).
No between-group difference for EQ-5D.
IAH (Clarke) improved by ~40% but no between-group difference.
(Continues)
|
7 of 23
TABLE 3 (Continued)
| 8 of 23


IN CONTROL29 16-week crossover, open- History IAH Gold, Clarke, PAID-5, At 16 weeks: fear of hypoglycaemia (HFS Worry subscale) was lower
label RCT with 12-week HFS, CIDS, EQ- after CGM compared to SMBG (32.5 vs. 38.9, mean difference 6.4 95%
washout 5D, WHO-5 CI1.4–11.4; p = 0.014).
(N = 52) No between-group difference for IAH (Gold, Clarke) or any other PROMs.
No change in IAH (Gold, Clarke) from baseline.
Laffel et al30 26-week multi-centre HbA1c 7.5%–<11% PAID-Peds, GMSS, At 26 weeks: rtCGM group reported greater satisfaction with glucose
open-label RCT HCS, PSQI monitoring (GMSS) than SMBG group (difference 0.27 [95% CI 0.06–
N = 153, n = 52 (34%) 0.54], p = 0.003).
(≥19 years) No between-group differences for diabetes distress (PAID-Ped), confidence
in managing hypoglycaemia (HCS) or sleep quality (PSQI).
Pratley et al31 26-week multi-centre RCT HbA1c < 10% HFS-II Worry, At 26 weeks: no between-group differences observed for any PROMs.
(N = 203) Age ≥ 60 years T1-DDS, Clarke,
PROMIS Global
Health Short form
(Emotion and
Cognition)
CONCEPTT32 At least 12-month multi- Pregnant or planning BGMSRQ, HFS-II At end of RCT (min 12 m): No between-group differences observed for any
centre / multi-country pregnancy. Age PAID, SF-12 PROMs.
open-label RCT. 18–40 years In the pregnancy group: Significant group-by-time interaction effects
Pregnant (N = 215) or favouring the CGM group observed for satisfaction with glucose
planning pregnancy monitoring (BGMSRQ Total: p = 0.043) and reduced hypoglycaemia
(N = 110) avoidance behaviours (HFS-II Behaviour subscale: p = 0.035).
In the pregnancy planning group: Significant group-by-time interaction
effects favouring the CGM group observed for satisfaction with
glucose monitoring (BGMSRQ Total: p = 0.43; BGMSRQ Impact
and Obstruction subscales: both p = 0.003) and reduced fear of
hypoglycaemia and hypoglycaemia avoidance behaviours (HFS-II
Worry and Behaviour subscales: p = 0.03 and p = 0.039 respectively).
CGM only Nefs et al33 6-month single centre HbA1c > 8% PAID, HFS Worry, At 6 months: n = 56 continued with rtCGM, of whom n = 37 completed
(Cohort Jimenez-Sahagun observational study (>64 mmol/mol) CIDS, CIDS-s, PROMS, which showed reduced diabetes distress (PAID) (p = 0.002,
study) et al34 (N = 60) Adults meeting HADS, Clarke d = −0.6), fear of hypoglycaemia (HFS Worry) (p = 0.02, d = −0.4),
3-month observational national criteria DTSQ, DQOL, DDS increased sensor-specific self-efficacy (CIDS-s) (p = 0.03, d = 0.5).
study (N = 114) for isCGM use No change from baseline in IAH, CIDS, HADS.
& committed At 3 months; significant improvements in diabetes treatment satisfaction
to completing (DTSQ: p < 0.001), and diabetes-specific quality of life (DQOL:
three educational p = 0.017); but no reduction in diabetes distress (DDS: p = 0.157).
isCGM sessions The change in DTSQ was evident regardless of HbA1c. However, the
improvement in DQOL score was driven by those with a starting HbA1c
SPEIGHT et al.
≤8.0% (p = 0.019), while those with HbA1c >8.0% demonstrating no

benefit (p = 0.728).
TABLE 3 (Continued)

SPEIGHT et al.
rtCGM vs. Reddy et al35 8-week parallel group RCT History IAH or SH in HFS-II, PAID, Gold At 8 weeks: rtCGM group reported reduction in fear of hypoglycaemia
isCGM (N = 40) past year (HFS-II Total and Worry subscale) compared with isCGM (both
p = 0.02).
No within-or between-group differences for diabetes distress (PAID) or
HFS-II Behaviour subscale.
IAH (Gold) improved by 60% but no between-group difference.
CORRIDA36 4-week RCT No history IAH or SH WHOQOL-BREF, At 4 weeks: no changes from baseline or no between-group differences for
(N = 60) Gold generic quality of life or IAH (Gold).
isCGM vs. IMPACT37 6-month, multi-centre HbA1c < 7.5% DDS, DQOL, DTSQ, At 6 months: isCGM group showed improved treatment satisfaction
SMBG RCT (<58 mmol/mol), HFS-II (DTSQ) compared with SMBG, and in perceived frequency of
(N = 241) no IAH hyperglycaemia (DTSQ item 2) (both p < 0.0001).
No between-group differences for diabetes distress (DDS), diabetes-specific
quality of life (DQoL) or fear of hypoglycaemia (HFS-II).
isCGM only FLARE-NL438 6- and 12-month N/A SF-12, EQ-5D-3L, At 6 and 12 months: improved general health status (EQ-5D-3L and EQ-
(Cohort nationwide, single-arm EQ-VAS VAS) and mental health status (SF-12 MCS) compared with baseline.
studies) observational No change in physical health status (SF-12 PCS).
(N = 1365, including 1.054
(77%) with T1D)
Al-Hayek et al39 3-month prospective Age: 18–40 years DDS, PSQI At 3 months: reduced diabetes distress (DDS: 3.8 vs. 2.5; p < 0.001) and
cohort study (N = 95) improved sleep quality (PSQI: 8.7 vs. 3.9; p < 0.001).
Deshmukh et al40 National audit N/A Gold, DDS2 At follow-up: improved awareness of hypoglycaemia (Gold: 2.7 to 2.4,
At follow up 98% T1DM, p < 0.0001) and reduced diabetes distress (DDS2: item 1, 2.9 to 2.2; item
(N = 3182) 2, 3.0 to 2.2; p < 0.0001).
a
Technology: rtCGM, real-time continuous glucose monitoring; CSII, continuous subcutaneous insulin infusion; DBLG1, diabeloop generation 1; HCL, hybrid closed loop; isCGM, intermittently scanned CGM, MDI,
multiple daily injections; PLGS, predictive low glucose suspend; SAP, sensor-augmented pump therapy; SMBG, self-monitoring of blood glucose.
b
RCT, randomised controlled trial. In RCTs, comparisons are made between study arms at follow-up, while in single-arm studies, comparisons are made between baseline and study end.
c
Specific inclusion criteria beyond ‘adult with type 1 diabetes’: IAH, impaired awareness of hypoglycaemia; SH, severe hypoglycaemia.
d
PROMs: BGMSRQ, Blood Glucose Monitoring System Rating Questionnaire; CES-D, Centre for Epidemiological Studies Depression Scale; CIDS(−s), Confidence in Diabetes Self-Care (sensor subscale); DDS, Diabetes
Distress Scale; DIDS, Diabetes Impact and Devices Satisfaction; DIDP, Dawn2 Impact of Diabetes Profile; DQOL, Diabetes Quality Of Life; DSQOL, Diabetes-Specific Quality of Life; DTQ, Diabetes Technology
Questionnaire; DTSQ(s + c), Diabetes Treatment Satisfaction Questionnaire (status and change versions); EQ-5D, EuroQoL five dimensions; GME-Q, Glucose Monitoring Experience Questionnaire; GMSS, Glucose
Monitoring Satisfaction Scale; HADS, Hospital Anxiety and Depression Scale; HCS, Hypoglycaemic Confidence Scale; HFS, Hypoglycaemia Fear Survey; HypoA-Q, Hypoglycaemia Awareness Questionnaire; ITSQ,
Insulin Treatment Satisfaction Questionnaire; PAID (-Peds), Problem Areas in Diabetes (Pediatric version); PRMQ, Prospective and Retrospective Memory Questionnaire; PSQI, Perceived Sleep Quality Index; SF-12, 12
item short form health survey (MCS, Mental health Component Summary; PCS, Physical health Component Summary); W-BQ28, 28-item Well-Being Questionnaire; WHO-5: Well-Being Index Scale; WHOQOL-BREF,
World Health Organization Quality Of Life—BREF.
|
9 of 23
|
4.1 | Insulin pumps (CSII) compared with baseline, which was also evident at
6–12 months follow-up.16
A systematic review published in 2007 reported equivocal
evidence for continuous subcutaneous insulin therapy
(CSII), also known as insulin pumps, on QoL and other 4.2 | Continuous glucose monitoring
PROs.5 It is likely that differences in results are due to (CGM)
heterogeneity in study design, sample size and selection,
as well as variation in PROMs (very few of the studies Several trials have assessed the impact of CGM on
actually assessed QoL). Indeed, many studies examining PROs. The DIAMOND RCT compared CGM with self-
the benefits of CSII use generic measures, which may not monitoring of blood glucose (SMBG) in adults with T1D
be sensitive to subtle differences between insulin delivery using multiple daily injections (MDI) and demonstrated a
devices.13 greater increase in confidence in managing hypoglycaemia
However, more recent studies show that people with in the CGM arm and moderate improvement in diabetes
T1D using CSII report greater treatment satisfaction and distress compared with the SMBG group over 24 weeks.
diabetes-specific QoL than those using multiple daily in- No between- group differences were observed in
jection (MDI), with low discontinuation rates for CSII. general emotional well-being, health status, or fear of
An important example is the REPOSE cluster randomised hypoglycaemia.25 Additionally, participants in the CGM
controlled trial (RCT), which compared CSII with MDI arm scored high on CGM satisfaction, primarily related
in the presence of equivalent structured education.15 to “benefits” and “loss of hassles”.44 Importantly, CGM
REPOSE had a large sample (N = 317) and longer fol- satisfaction was not related to glycaemic changes and
low-up period (2 years) than most previous studies and it is worth noting that because the measure was CGM-
benefitted from a high PROM completion rate (90%). Both specific, there was no comparison with baseline, previous
groups experienced improvements in psychosocial out- monitoring, or the SMBG group.
comes, but there were some notable differences between The GOLD study, a crossover RCT of CGM versus
arms. At 24 months, those allocated to CSII reported SMBG in those on MDI, demonstrated improved gen-
greater diabetes-specific QoL in three domains (i.e., lei- eral emotional well-being and confidence in managing
sure, dietary freedom and daily hassles), as well as greater hypoglycaemia in the CGM group at 6 months, but no
diabetes treatment satisfaction and less worry about hypo- between-group differences for fear of hypoglycaemia.26,27
glycaemia compared to the MDI group. Some differences A RCT assessed CGM versus SMBG in 153 adolescents
were also evident earlier at 12 but not 6 months. Of note, and young adults with diabetes (only a third were older
no differences were detected in generic health status or than 19 years).30 At 26 weeks, the CGM group reported
depression/anxiety assessments at any of the time points greater glucose monitoring satisfaction, but there were no
(Table 2). between-group differences for diabetes distress, hypogly-
Similarly, the HypoCOMPaSS RCT compared CSII to caemia confidence, sleep quality or IAH. A RCT of CGM
MDI in 96 adults with long-standing T1D and impaired in 203 older adults (>60 years) found no differences at
awareness of hypoglycaemia (IAH) and provided equiva- 26 weeks in any PROMs.31
lent psycho-education and attention to both groups.13 At The CONCEPTT RCT compared CGM with SMBG
6 months, between-group analyses showed comparable in women (18–40 years) with T1D who were pregnant
reductions in severe hypoglycaemia, fear of hypoglycae- or planning pregnancy.32 While there were no between-
mia, and insulin doses, with equivalent HbA1c. However, group differences in any PROMs at the study end, there
diabetes treatment satisfaction and satisfaction with insu- were group-by-time interactions favouring CGM for sat-
lin “delivery device” was higher with CSII than MDI at isfaction with glucose monitoring and fear of hypoglycae-
6 months.14,41 These differences were no longer apparent mia both during pregnancy and pregnancy planning.
at 24 months, following an 18-month observational phase A single-arm observational study of 60 adults with T1D
during which individuals used their preferred insulin de- (36 of whom completed PROs) showed that CGM use is
livery system. associated with reduced diabetes distress and fear of hy-
While this review focuses on RCTs, it is worth men- poglycaemia together with increased sensor-specific self-
tioning that observational and qualitative studies also efficacy 6 months after starting CGM.33
report improved QoL and related outcomes with CSII, People with T1D are often excluded from RCTs if
demonstrating enhanced lifestyle flexibility and improved they have a history of problematic hypoglycaemia. The
diabetes self-management among CSII users.42,43 A small HypoDE study is the largest RCT to date (N = 149) assess-
cohort study involving 47 individuals with T1D starting ing the impact of CGM in adults with a history of IAH or
on CSII showed reduced diabetes distress at 3–6 months severe hypoglycaemia.28 CGM use in adults using MDI led
|
to a 72% reduction in subsequent hypoglycaemic events. study of 114 individuals with T1D showed improved treat-
Although there was a trend towards superior improvement satisfaction and diabetes-specific quality of life but no
ments in each score with CGM versus SMBG at 6 months, reduction in diabetes distress with the use of isCGM.34 A UK
between-group differences were detected only for satisfac- national audit collected diabetes distress data at baseline and
tion with the monitoring method and diabetes distress. follow-up (median 7.5 months) in 2532 individuals with di-
Similarly, in the HypoCOMPaSS study (which also as- abetes (97% with T1D) starting isCGM. It showed reduced
sessed the impact of CGM versus SMBG among 96 adults diabetes distress and improved awareness of hypoglycaemia.
with problematic hypoglycaemia), treatment satisfaction However, follow-up data were unavailable for two-thirds of
improved and fear of hypoglycaemia reduced across the the 8320 participants originally approached.40
whole cohort at 6-month follow-up, and was maintained The CORRIDA RCT compared isCGM with real-time
at 24 months.14 However, there were no between-group CGM in 60 individuals with T1D and showed no between-
differences in these outcomes, suggesting that equivalent group differences in IAH or general QoL at 4 weeks.36
clinical attention and psycho-education is as important Another study involving 40 individuals with T1D and IAH
as the technology. In a 16-week crossover trial (with a 12- showed CGM is superior to isCGM for reducing fear of hy-
week washout period), of 52 individuals with T1D and poglycaemia but without effect on diabetes distress.35 IAH
IAH, fear of hypoglycaemia was lower in the CGM than in improved by 60% irrespective of device allocation.
the SMBG group. However, no between-group differences
were detected in IAH, diabetes self-care, diabetes distress,
general emotional well-being or health status.29 4.4 | Sensor-augmented pumps
In summary, it appears CGM may have an important (SAP) and hybrid closed loop (HCL)
role in improving PROs, which may be mediated by the
prevention or pro-active management of hypoglycaemia, Sensor-augmented pumps (SAP) combine CGM with CSII.
but this needs further investigation in future studies. The first iterations could only suspend insulin delivery if
glucose was too low (threshold suspend) or predicted to
be too low (predictive suspend). The latest versions, so-
4.3 | Intermittently scanned CGM called hybrid closed loops (HCL; also known as “artificial
pancreas”), can also deliver insulin as either changes to basal
Intermittently scanned CGM (isCGM or Flash glucose rates or small boluses to prevent high glucose excursions.
monitoring) has the advantage of long sensor life and A network meta-analysis and narrative synthesis of 52
factory calibration, thus eliminating the need for capillary T1D studies compared the effects of various technologies
glucose monitoring, except in cases of extreme glucose on HbA1c, hypoglycaemia and PROs.45 The work con-
levels. Unlike CGM, the first iteration of isCGM did cluded that, although risk of bias was moderate-to-high
not have low/high glucose alarms but the latest version and certainty of evidence was low, SAP therapy may be su-
(FreeStyle Libre 2) has optional glucose alarms. perior to other diabetes technologies for improving PROs.
The IMPACT RCT, investigating isCGM in 241 adults However, incremental advances in SAP, from suspend on
with T1D and baseline HbA1c <7.5%, found less hypogly- low to predictive suspend and HCL, were not compared.
caemia in the isCGM group compared to the SMBG group Importantly, CGM was consistently associated with im-
after 6-month follow-up.37 Diabetes treatment satisfaction proved PROs irrespective of how insulin was delivered.
was greater, and perceived frequency of hyperglycaemia SMILE was an open-label RCT comparing SAP with
was lower, in the isCGM compared with the SMBG group. predictive low glucose suspend (PLGS) to CSII/SMBG
Although the difference in diabetes-specific QoL did not (control) in adults with long-standing T1D at high risk of
reach statistical significance in the full analysis, there was hypoglycaemia and who used CSII prior to enrolment.17
a trend favouring isCGM, but no between-group differ- SAP-PLGS improved diabetes treatment satisfaction and
ences for diabetes distress or fear of hypoglycaemia. reduced fear of hypoglycaemia compared to CSII/SMBG,
An observational study, involving 1365 individuals but there were no between-group differences for IAH.
with diabetes (1054 with T1D), showed improved gen- A RCT of the “Diabeloop” HCL system compared to SAP,
eral and mental health status at 6 and 12 months (but not in 63 adults with T1D, found no between-group differences
physical health status) compared with baseline.38 Given at 12 weeks in diabetes treatment satisfaction.19 A 6-month
most had T1D (77%), it may be reasonably assumed that RCT comparing HCL with standard care (without CGM) in
these benefits apply to this subgroup. 120 adults with T1D found improved diabetes-specific pos-
In a 3-month prospective cohort study (95 adults with T1D) itive well-being and diabetes-specific QoL at 6 months but
isCGM use was associated with reduced diabetes distress no between-group differences in diabetes treatment satisfac-
and improved sleep quality.39 Another 3-month single-arm tion, diabetes distress, subjective sleep quality or cognition.18
|
Beyond RCTs, observational studies on HCL reported by the #WeAreNotWaiting movement.49 These “user-led”
improved sleep and general well-being and reduced diabe- or “Do-It-Yourself (DIY)” systems are also referred to as
tes burden.46 A real-world evaluation of the Medtronic 670G “OpenAPS”, “DIYAPS” or “Looping”. They are built with
in 92 youth (including n = 27 aged 18+) found no changes ease of use, automation, communication and the user
in fear of hypoglycaemia or diabetes distress across time, interface in mind. At present, there is a relative paucity
with 30% of youth discontinuing HCL in the first 6 months. of evidence for such systems using validated PROMs. A
The authors report this may be related to challenges with large, multi-country quantitative survey (employing study-
calibration and the high workload required to maintain specific items) of 722 adults using OpenAPS showed self-
the system in automated mode.47 These data contrast with reported benefits of putting diabetes on “auto-pilot” (81%
a 3-month observational, single-arm study (34 adults and of users) and for subjective sleep quality (72% of users).50
22 children), showing HCL use is associated with improved A large qualitative (ethnographic) study identified a range
diabetes- specific QoL, diabetes treatment satisfaction, of QoL benefits by extracting user experiences from Twitter
subjective sleep quality, and awareness of hypoglycaemia, posts.51 Further qualitative thematic analysis reported in
and reduced diabetes distress and fear of hypoglycaemia.23 the same paper illustrated the quantitative findings showing
These differences may reflect different study designs, popu- that improved QoL was due largely to reducing the burden
lations, expectations and reimbursement criteria. of diabetes self-management, improving sleep, reducing
A crossover RCT of 32 individuals with T1D compared diabetes distress and burnout, and increasing autonomy/
HCL and SAP over a 2-month period (with 4-week wash- personal control. It should be noted that the current evidence
out). No between-group differences were detected in fear base has been led largely by the OpenAPS community, and
of hypoglycaemia or treatment satisfaction.20 A small, is characterised by cohort studies and surveys (as opposed
single-arm 4-week pilot of SAP in 15 older adults with to RCTs). Robust, independent evidence is needed and may,
diabetes (mean age 69 ± 3 years), followed by 4 weeks of in part, be provided by the upcoming ABCD nationwide
automated insulin delivery (Control IQ), showed the latter DIYAPS audit launched in 2020 (http://abcd.care/diyaps).
is associated with reduced diabetes distress. There were no
changes in fear of hypoglycaemia, depressive symptoms
or subjective sleep quality.21 5 | THE IMPACT OF DIABETE S
Real-world follow-up of 967 users of the Tandem TECHNOLOGIES ON QUALITY O F
Control-IQ HCL reported improved satisfaction with de- LIFE AND RELATED OUTCOME S BY
vice use over time and reduced diabetes impact.48 Another TYPE OF PROs
real-world study invited 9085 Tandem control IQ HCL
users to complete several PROMs at two timepoints: the Table 4 summarises study outcomes by psychological
first at least 3 weeks after starting the pump and the sec- construct and the PROMs used for assessment. The PROMs
ond 4 weeks later. A total of 1435 users completed study identified in this review assessed several psychological
questionnaires at both timepoints, showing improved constructs, including:
device-related satisfaction and emotional well-being at
the second timepoint.22 A recent single-arm study using • Hypoglycaemia-specific: fear and confidence in manag-
the Omnipod 5 automated delivery system in 115 adults ing hypoglycaemia;
with T1D has shown improvement in diabetes-specific • Diabetes-specific: QoL, well-being and distress, satisfac-
PROMs at 3 months of device use, including reduced dia- tion with treatment;
betes distress, improved confidence in managing hypogly- • Generic: health or functional status, emotional well-
caemia and satisfaction with diabetes treatment.24 being, depressive symptoms, subjective sleep quality.
In summary, although evidence regarding the effect of
HCL on PROs is limited by small, short-duration studies
and few RCTs, evidence is accumulating to suggest that 5.1 | Hypoglycaemia-specific PROMs
this approach has considerable benefits for some PROs.
Fear of hypoglycaemia was most commonly assessed
in RCTs. Three of eight RCTs showed reduced fear of
4.5 | Open-source automated insulin hypoglycaemia and/or improved confidence in managing
delivery systems hypoglycaemia with CGM. Of the two RCTs comparing
real-
time CGM (rtCGM) with isCGM, one reported a
Open-source automated insulin delivery systems are between-group difference in fear of hypoglycaemia,
designed and built by people with diabetes for their own favouring rtCGM, suggesting that low glucose alarms
personal use, based on open-source algorithms, developed are beneficial. Two RCTs comparing MDI with CSII
TABLE 4 Summary of study findings by psychological construct and person-reported outcome measure (PROM)
Construct and PROM Study name: duration & design (ref) SMBG rtCGM isCGM MDI CSII SAP HCL
Confidence in diabetes self-care
SPEIGHT et al.
Confidence in Diabetes Self-care 6-month Cohort (Nefs et al, 2020)33 +a

(CIDS)
Confidence in Diabetes Self-care 6-month Cohort (Nefs et al, 2020)33 +a
sensor-specific (CIDS-s)
Diabetes distress
Diabetes Distress Scale (DDS) DIAMOND: 24-week RCT (Polonsky et al, 2017)25 − +
26-week RCT (Pratley et al, 2020)31 − −
21
4-week Pilot crossover (Bisio et al, 2021) − +
24
3-month Cohort (Polonsky et al, 2022) +
3-month Cohort (Jimenez-Sahagun et al, 2022)34 +
3-month Cohort (Beato-Vibara et al, 2020)23 +
3-month Cohort (Hayek et al, 2020)39 +
Diabetes Distress Scale, 2-item Audit (Deshmukh et al, 2020)40 +
short-form (DDS2)
Diabetes Distress Scale for Type 1 HypoDE: 6-month RCT (Heineman et al, 2018)28 − +
diabetes (T1-DDS) 37
IMPACT: 6-month RCT (Bolinder et al, 2016) − −
18
Problem Areas In Diabetes Australian HCL: 6-month RCT (McAuley et al, 2020) − − −
(PAID) scale CONCEPTT (Pregnancy and Pregnancy planning): − −
12-week RCT (Feig et al, 2017)32
GOLD: 26-week Crossover RCT (Lind et al, 2017)26 − −
IN CONTROL: 16-week RCT (van Beers et al, 2016)29 − −
35
8-week RCT (Reddy et al, 2017) − −
30
26-week RCT (Laffel et al, 2020) − −
6-month Cohort (Nefs et al, 2020)33 +
12-month Cohort (Oldham et al, 2020)16 +
Diabetes-specific positive well-being
4-item subscale of the W-BQ28 Australian HCL: 6-month RCT (McAuley et al, 2020)18 − − +
Diabetes-specific quality of life
DAWN Impact of Diabetes Profile Australian HCL: 6-month RCT (McAuley et al, 2020)18 − − +

(Continues)
| 13 of 23
TABLE 4 (Continued)
14 of 23
|
37
Diabetes Quality of Life (DQoL) IMPACT: 6-month RCT (Bolinder et al, 2016) − −

questionnaire 34
3-month cohort (Jimenez-Sahagun et al, 2022) +
24
15
Diabetes-Specific Quality Of Life REPOSE: 12/24-month RCT (Heller et al, 2017) − +
Scale (DSQOLS)
Fear of hypoglycaemia / Confidence in managing hypoglycaemia
Hypoglycaemic Confidence Scale DIAMOND: 24-week RCT (Polonsky et al, 2017)25 − +
(HCS) 26
GOLD: 26-week Crossover RCT (Lind et al, 2017) − +
GOLD-3: 69-week Crossover RCT (Olafsdottir et al, 2018)27 − +
24
26-week RCT (Laffel et al, 2020)30 − −
Hypoglycaemia Fear Survey CONCEPTT (Pregnancy and Pregnancy planning): − −
(HFS / HFS-II) 12-month RCT (Feig et al, 2017)32
HypoCOMPaSS: 6-month RCT (Little et al, 2014)41 * − − − −
HypoDE: 6-month RCT (Heineman et al, 2018)28 * − −
IMPACT: 6-month RCT (Bolinder et al, 2016)37 − −
IN CONTROL: 16-week RCT (van Beers et al, 2016)29 − +b
REPOSE: 6/12-month RCT (Heller et al, 2017)13 − −
13
REPOSE: 24-month RCT (Heller et al, 2017) − +
SMILE: 6-month RCT (Bosi et al, 2019)17 − +c
Crossover RCT (Kropff et al, 2016)20 − −
8-week RCT (Reddy et al, 2017)35 +d −
26-week RCT (Pratley et al, 2020)31 − −
21
4-week Pilot crossover (Bisio et al, 2021) − −
3-month Cohort (Beato-Vibara et al, 2020)23 +c
Hypoglycaemia Fear Survey CONCEPTT (Pregnancy): 12-month RCT (Feig et al, − +
(HFS / HFS-II), Behaviour 2017)32
subscale only CONCEPTT (Pregnancy planning): 12-month RCT (Feig − +
et al, 2017)32
SPEIGHT et al.
TABLE 4 (Continued)
Hypoglycaemia Fear Survey CONCEPTT (Pregnancy): 12-month RCT (Feig et al, − −
SPEIGHT et al.
(HFS / HFS-II), Worry 2017)32

subscale only CONCEPTT (Pregnancy planning): 12-month RCT (Feig − +
et al, 2017)32
DIAMOND: 24-week RCT (Polonsky et al, 2017)25 − −
33
6-month Cohort (Nefs et al, 2020) +
General emotional well-being, depressive symptoms, anxiety symptoms
Centre for Epidemiological Pilot crossover (Bisio et al, 2021)21 − −
Studies Depression Scale
(CES-D)
Hospital Anxiety and Depression REPOSE: 6/12/24-month RCT (Heller et al, 2017)12 − −
Scale (HADS) 32
6-month Cohort (Nefs et al, 2020) −
25
WHO-5 Well-being Index DIAMOND: 24-week RCT (Polonsky et al, 2017) − −
GOLD: 26-week Crossover RCT (Lind et al, 2017)26 − +
24
3-month Cohort (Polonsky et al, 2022) −
29
IN CONTROL: 16-week RCT (van Beers et al, 2016) − −
4-week Cohort (Pinkser et al, 2021)22 +
Generic health status
EuroQol-5D (EQ-5D) DIAMOND: 24-week RCT (Polonsky et al, 2017)25 − −
37
FLARE-NL4: 6/12-month cohort (Fokkert et al, 2019) +
13
REPOSE: 6/12/24-month RCT (Heller et al, 2017) − −
Short-Form Health Survey, 12 CONCEPTT (Pregnancy and Pregnancy planning): 12- − −
items (SF-12) month RCT (Feig et al, 2017)32
Short-Form Health Survey, FLARE-NL4: 6/12-month cohort (Fokkert et al, 2019)38 +
12 items (SF-12): Mental REPOSE: 6/12/24-month RCT (Heller et al, 2017)15 − −
Component Summary
Short-Form Health Survey, FLARE-NL4: 6/12-month cohort (Fokkert et al, 2019)38 −
12 items (SF-12): Physical 15
Component Summary
Generic quality of life
WHOQOL-BREF CORRIDA: 4-week RCT (Haskova et al, 2020)36 − −
15


|
Impaired awareness of hypoglycaemia

Clarke score HypoCOMPaSS: 6-month RCT (Little et al, 2014)41 − − − −
15 of 23
(Continues)
TABLE 4 (Continued)
16 of 23
|
28 *
HypoDE: 6-month RCT (Heineman et al, 2018) − −


17
SMILE: 6-month RCT (Bosi et al, 2019) − −
31
26-week RCT (Pratley et al, 2020) − −
33
6-month Cohort (Nefs et al, 2020) −
Gold score Audit (Deshmukh et al, 2020)40 +
36
CORRIDA: 4-week RCT (Haskova et al, 2020) − −
41
HypoCOMPaSS: 6-month RCT (Little et al, 2014) − − − −
SMILE: 6-month RCT (Bosi et al, 2019)17 − −
35 *
8-week RCT (Reddy et al, 2017) − −
HypoA-Q HypoCOMPaSS: 6-month RCT (Little et al, 2014)41 − − − −
Memory
Prospective and Retrospective Australian HCL: 6-month RCT (McAuley et al, 2020)18 − − −
Memory Questionnaire
(PRMQ)
Satisfaction with diabetes treatment / glucose monitoring
Blood Glucose Monitoring CONCEPTT (Pregnancy): 12-week RCT (Feig et al, 2017)32 − +
Satisfaction Rating CONCEPTT (Pregnancy planning): 12-week RCT (Feig − +
Questionnaire (BGMSRQ) et al, 2017)32
Diabetes Treatment Satisfaction Australian HCL: 6-month RCT (McAuley et al, 2020)18 − − −
Questionnaire (DTSQ) Diabeloop: Crossover RCT (Benhamou et al, 2019)19 − −
(DTSQ-s = status version;
GOLD: 26-week Crossover RCT (Lind et al, 2017)26 − +
DTSQ-c = change version)
HypoCOMPaSS: 6-month RCT (Little et al, 2014)41 − − − +
13
HypoCOMPaSS: 24-month (preference) (Little et al, 2018) − − − −
29
IMPACT: 6-month RCT (Bolinder et al, 2016) − +
REPOSE: 6-month RCT (Heller et al, 2017)15 − −
15
REPOSE: 12/24-month RCT (Heller et al, 2017) − +
SMILE: 6-month RCT (Bosi et al, 2019)17 − +
20
Crossover RCT (Kropff et al, 2016) − −
34
3-month cohort (Jimenez-Sahagun et al, 2022) +
3-month Cohort (Polonsky et al, 2022)24 +
SPEIGHT et al.
23
3-month Cohort (Beato-Vibara et al, 2020) +
TABLE 4 (Continued)
14
Glucose Monitoring Experiences HypoCOMPaSS: 6-month RCT (Speight et al, 2019) − +
SPEIGHT et al.
Questionnaire (GME-Q) HypoCOMPaSS: 24-month (preference) (Speight et al, − −

2019)14
Glucose Monitoring Satisfaction HypoDE: 6-month RCT (Heineman et al, 2018)28 − +
Scale (GMSS) 26-week RCT (Laffel et al, 2020)30 − +
14
Insulin Treatment Satisfaction HypoCOMPaSS: 6-month RCT (Speight et al, 2019) − +
Questionnaire (ITSQ) HypoCOMPaSS: 24-month (preference) (Speight et al, − −
2019)14
Sleep quality
Pittsburgh Sleep Quality Australian HCL: 6-month RCT (McAuley et al, 2020)18 − − −
Inventory (PSQI) 26-week RCT (Laffel et al, 2020)30 − −
3-month Cohort (Polonsky et al, 2022)24 −
21
4-week Pilot crossover (Bisio et al, 2021) − −
39
3-month Cohort (Al-Hayek et al, 2020) +
Technology: CSII, continuous subcutaneous insulin infusion; HCL, hybrid closed loop; isCGM, intermittently scanned CGM; MDI, multiple daily injections; rtCGM, real-time continuous glucose monitoring; SAP,
sensor-augmented pump therapy (±PLGS, predictive low glucose suspend); SMBG, self-monitoring of blood glucose. (−) indicates no effect on the outcome measured, (+) indicates an improvement in the outcome
measured.
a
Improvement in CIDS-s only.
b
Reduction only for HFS Worry subscale.
c
Reduction HFS Total, Worry and Behaviour subscales.
d
Reduction only for HFS Total score and Worry subscale.
*Within-group reduction.
|
17 of 23
|
(REPOSE and HypoCOMPaSS) reported no between-group related outcomes, particularly in reducing the negative
differences in fear of hypoglycaemia at 6 or 12 months. Both impact of diabetes and hypoglycaemia, while there appear
trials included equivalent psycho-education, attention and to be fewer benefits for generic PROs. While technology
clinical support suggesting that equivalent benefits can be can benefit people with diabetes, there can also be
achieved regardless of technology use. subjective burdens and barriers to uptake (Figure 1),
Three RCTs assessed the impact of more advanced, which can only be assessed using PROMs. PROMs offer a
automated insulin delivery technologies, two of which systematic, valid and reliable approach to understanding
reported positive impacts on fear of hypoglycaemia com- a person's experiences (e.g., satisfaction, confidence, well-
pared to MDI/CSII. being, impact on QoL) regarding the management of their
None of the trials showed a difference in IAH between diabetes. It is important that PROMs analyse constructs
rt/isCGM and SMBG, suggesting that awareness of hypo- that are affected by, and sensitive to, the condition and/or
glycaemia is not necessarily improved with CGM. the technology.
5.2 | Diabetes-specific PROMs 6.1 | Implication of study design in

technology studies
Two of nine RCTs reported reduced diabetes distress among
those allocated to CGM compared to SMBG. Two studies The studies described here are mostly, though not ex-
reported reduced diabetes distress and one improved clusively, RCTs. While RCTs have high internal valid-
confidence with glucose sensor use. One cohort study ity, other study designs are stronger on external validity
showed CSII was associated with reduced diabetes distress. (Figure 2). Moreover, some RCT protocols are demanding,
CGM use resulted in greater satisfaction with diabetes which may disrupt a person's routine, sleep or QoL. This
treatment in general and specifically with the monitoring may be a reason why some cohort studies show greater
device, compared to SMBG. One crossover SAP/HCL trial benefits than RCTs for (generic) PROs. Therefore, RCT
found no between-group differences for diabetes treat- evidence needs to be complemented by real-world cohort
ment satisfaction. Overall, CSII appeared to lead to greater studies to fully understand the impact of technology on
treatment satisfaction compared to MDI. the PROs. A key challenge for all study designs is how
Three RCTs assessed the impact of more advanced, quickly diabetes technologies are evolving—by the time
automated insulin delivery technologies, two of which findings are published, the technology has advanced and
reported positive impacts on diabetes-specific-well-being the findings may lack relevance. There is a need to apply
and QoL compared to MDI/CSII, but not for diabetes dis-
tress. Two cohort studies observed improvements after
3 months of HCL use for several diabetes-specific PROMs.
Flexibility
Reassurance
Reduced hypoglycaemia
5.3 | Generic PROMs Improved glycaemia
Benefits
Eight generic PROMs were assessed in 10 studies. General
emotional well-being improved in one RCT (between-
group difference favouring CGM over SMBG) and one Clinician Sleep quality
cohort study of HCL. One cohort study of isCGM showed support
Data
& quantity
Visibility
improved general and mental health (but not physical Alarms
Sensors
health). No benefits were shown for general anxiety or
depressive symptoms or generic QoL.
Barriers Cost Burdens
Improved subjective sleep quality was reported for Skin reaction
Perception as ‘miracle’ Wearing a device
both isCGM and HCL in two cohort studies but none of Hype vs hope Engagement
the RCTs measured it. Discipline
Lot of work with no break
6 | DI S C USSION
F I G U R E 1 Summary of the potential benefits, burdens and
This review demonstrates that diabetes technologies are barriers associated with using diabetes technologies from the
often associated with considerable benefits for QoL and perspective of the person with type 1 diabetes.
|
Implications of study design
Randomised controlled trials Cohort studies
Advantages Disadvantages Advantages Disadvantages

Considered ‘gold standard’ due to Difficult to blind to technology allocation. Understand whether a particular technology Inherent bias (there is a reason for adopting a
randomisation, prospective nature, comparison Participants may have preferences for novel suits an individual in day-to-day life. particular technology) and lack of a control
to a control group, blinding and controlling for technologies, leading to selection bias or Can overcome some limitations of RCT designs group (causality cannot be determined).
different variables differential dropout. Limited generalisability (including selection bias) Limited participant details and variable follow-up
Partially-randomised preference trials
Advantages Disadvantages
People with strong preferences are given their These studies are often perceived as having too
choice of technology, and those without distinct many uncertainties, making sample size
preferences are randomised calculation problematic
FIGURE 2 Implications of study design when assessing PROMs. RCT: randomised controlled trial, FU: follow-up.
adaptive trial designs that can keep up with this fast-paced ensure that participants received equivalent education,
area.52 attention and support regardless of allocation to interven-
As for all studies, participant selection is crucial. tion or control group. In both RCTs, there were few dif-
Historically, adults with problematic hypoglycaemia, ferences (between CSII and MDI, or CGM and SMBG) in
pregnant women and older adults have been excluded biomedical or psychological outcomes, with the notable
from technology RCTs. It is only recently that such groups exception of treatment satisfaction being greater among
have been included, yet arguably these people could ben- those allocated to pump.
efit the most. Other groups with limited representation Finally, interpretation of PROM findings needs to
include people with higher HbA1c, people from minority consider the statistical analysis. RCTs typically report
ethnic groups, people with lower socioeconomic status, between- group differences at follow- up, rather than
and those who do not speak English. Moreover, RCT within-group differences over time. Comparative effec-
participants are frequently well-educated, motivated and tiveness trials are becoming more common, such that im-
often have low levels of depressive symptoms or impaired provements in both groups (despite lack of between-group
diabetes-specific of generic well-being. While cohort stud- difference) could be viewed positively.
ies may be less restrictive, they can only include those
who have routine access to technologies. Thus, technology
studies can exclude large proportions of individuals with 6.2 | Implication of PROM selection in
diabetes who may benefit from glycaemic technologies. technology studies
Very few studies include details of the extent to which
participants have used the technology as intended (e.g., This review has highlighted the numerous PROMs that
wearing sensors at least 80% of the time). Consequently, exist and may be suitable for the evaluation of glycaemic
many studies offer relatively limited insights into the real- technologies. The psychological construct that the PROM
world experiences of people using these technologies. This is assessing needs to be considered in the context of the
is where qualitative studies are particularly beneficial,53,54 technology. Questions need to be asked whether, and
as they provide evidence of how the technology works in in study time frame, the technology used could lead to
real life, for whom and how well. significant changes in the PRO of interest. For example,
Another important consideration is the issue of edu- where fear of hypoglycaemia is low at baseline, it is
cation and attention. When people adopt a new diabetes unlikely that a significant difference will be observed,
technology, they are often seen by specialist teams and whereas confidence in managing hypoglycaemia likely
receive intensive education/support, resulting in overall has room for improvement.
improved clinical care, which may improve PROs.55 Both It is crucial that PROs are valued by all stakeholders
the REPOSE and HypoCOMPaSS trials were designed to and selected judiciously. RCTs largely relegate PROs to
|
secondary outcomes and perhaps it is time to move these diabetes understand how to complete the assessments
to a more central role and enable studies to be designed and that their feedback will be valued.69 Some health care
and powered appropriately for PROMs. It is also important professionals may be concerned about “response bias”,
that PROMs are not too lengthy or burdensome to com- whereby individuals respond in a certain way if they per-
plete.56 Online assessments are efficient and can improve ceive this affects recommendation or management (e.g.,
completion rates, eliminate data entry errors, fast-track for their suitability to drive, whether they are ‘deserving’
data analysis, and are, overall, a cost-effective approach.57 access to a certain technology). The main counter to this
Ecological momentary assessments offer a convenient phenomenon is for health care professionals to ensure
method for study participants to provide real-time com- that their relationship with the person with T1D is built
pletion of PROMs to demonstrate day-to-day impacts.58 on trust and open communication. There are also chal-
Psychological constructs examined less frequently lenges in identifying how to collect and incorporate suffi-
in the technology RCTs included confidence in manag- cient PROM data into clinical records for easy access and
ing hypoglycaemia, diabetes-specific positive well-being monitoring over time.68
and generic PROs, such as emotional well-being, sleep,
memory, and QoL. These are all of interest because qual-
itative research suggests improvements in most of these 7 | CONCLUSIONS
constructs following technology,59,60 and therefore these
constructs may require greater attention in future quan- This review has demonstrated that PROs are usually
titative research. Although there were relatively fewer assessed as secondary outcomes in glycaemic technology
studies assessing generic constructs, the findings sug- studies. While there are many nuances among these
gest no significant differences between groups, while findings, hypoglycaemia- specific and diabetes- specific
hypoglycaemia confidence and diabetes-specific positive PROMs appear to show greater benefits of glycaemic
well-being both showed benefits in RCTs. Taken together, technologies than generic PROMs. These findings show
these findings suggest that generic PROMs may be less re- the importance of understanding and appreciating (in
sponsive to glycaemic technologies than diabetes-specific both research and clinical care) the impact that glycaemic
or hypoglycaemia- specific measures, as previously technologies may have on the experiences of the person
discussed.5 with T1D. Where benefits for PROs exist, health care
professionals and policymakers need to value these as
much as the glycaemic benefits, to realise the full potential
6.3 | Implementing PROMs in clinical of technologies for maintaining or improving both health
diabetes care and QoL.
PROMs are undoubtedly valuable tools to inform decision AUTHOR CONTRIBUTIONS

making, improve symptom monitoring and strengthen RAA, PC, EGW and JS conceived and developed the
communication.61–63 Their clinical use has the potential to plan for the review. JS and CH critically reviewed the
increase the holistic care of people with T1D, e.g., through results section, led the discussion section, refined Table 2,
screening and identifying problems, understanding provided Tables 1 and 4, as well as Figure 1; PC and HF led
perceptions and experiences, and monitoring outcomes the CSII section; EGW and TC led the CGM section; WYC
over time,61 as well as through care co- ordination, provided an overview of PROMs; RM critically reviewed
including transition from primary to speciality care or from the work and helped write the abstract and conclusions;
paediatric to adult services. Several studies demonstrate GTT and AH led the section on PROMs and clinical
that most adults with T1D are willing to complete PROMs practice; RAA led the introduction, abstract and isCGM
at annual reviews.64,65 Routine use of clinic consultation sections; RAA and JS coordinated and critically reviewed
tools (incorporating PROMs) enables agenda setting, the overall manuscript. All authors read and approved the
monitoring of the impact of management strategies in real- final paper.
world settings, and truly person-centred care.66,67
However, PROMs are only a tool for identifying expe- FUNDING INFORMATION
riences and perceived problems, requiring follow-up with No specific funding was provided for this review. JS
appropriate action taken by the health care professionals and CH are supported by core funding to the Australian
to improve either biomedical or psychological outcomes Centre for Behavioural Research in Diabetes provided by
of people.68 It is important that health care professionals the collaboration between Diabetes Victoria and Deakin
receive adequate training and resources to enable effec- University. RAA's research is currently supported by
tive implementation. This includes ensuring people with grants from Diabetes UK, NIHR, BBSRC, Avacta Life
|
Sciences and Abbott Diabetes Care. BM and RAA are 4. Gonder-Frederick LA, Grabman JH, Shepard JA. Human factor
members of Diabetes Research Steering Group 2 and considerations for artificial pancreas research. Diabetes Technol
Ther. 2016;18(12):762-764.
would like to thank Diabetes UK for their support.
5. Speight J, Holmes-Truscott E, Hendrieckx C, Skovlund S, Cooke
D. Assessing the impact of diabetes on quality of life: what have
CONFLICT OF INTEREST the past 25 years taught us? Diabet Med. 2020;37(3):483-492.
JS has served on advisory boards for Insulet, Janssen, 6. Wolpert HA, Anderson BJ. Management of diabetes: are doc-
Medtronic, Roche Diabetes Care, and Sanofi Diabetes; tors framing the benefits from the wrong perspective? BMJ.
received unrestricted educational grants and in- kind 2001;323(7319):994-996.
support from Abbott Diabetes Care, AstraZeneca, 7. Black N. Patient reported outcome measures could help trans-
form healthcare. BMJ. 2013;346:f167.
Medtronic, Roche Diabetes Care, and Sanofi Diabetes;
8. Blood Z, Tran A, Caleo L, et al. Implementation of patient-
received sponsorship to attend educational meetings from
reported outcome measures and patient-reported experience
Medtronic, Roche Diabetes Care, and Sanofi Diabetes, measures in melanoma clinical quality registries: a systematic
and consultancy income or speaker fees from Abbott review. BMJ Open. 2021;11(2):e040751.
Diabetes Care, AstraZeneca, Medtronic, Novo Nordisk, 9. DAFNE Study Group. Training in flexible, intensive insulin
Roche Diabetes Care, and Sanofi Diabetes. In all cases, her management to enable dietary freedom in people with type
research group (The Australian Centre for Behavioural 1 diabetes: dose adjustment for normal eating (DAFNE) ran-
Research in Diabetes [ACBRD]) has been the beneficiary of domised controlled trial. BMJ. 2002;325(7367):746.
10. Speight J, Shaw JA. Does one size really fit all? Only by con-
these funds. PC has received personal fees from Medtronic,
sidering individual preferences and priorities will the true im-
Abbott, Dexcom, Insulet, Glooko, Novo Nordisk, Sanofi pact of insulin pump therapy on quality of life be determined.
and Eli Lilly. EGW has received personal fees from Abbott, Diabet Med. 2007;24(7):693-695.
Dexcom, Eli Lilly, Insulet, Medtronic, Novo Nordisk and 11. Barnard KD, Lloyd CE, Skinner TC. Systematic literature re-
Sanofi Aventis. TC has received speaker fees/honoraria view: quality of life associated with insulin pump use in type 1
and educational grants from Sanofi, Abbott Diabetes Care diabetes. Diabet Med. 2007;24(6):607-617.
and NovoNordisk. RAA Institutional Research Grants; 12. Thabit H, Hovorka R. Continuous subcutaneous insulin in-
Abbott, Bayer, Eli Lilly and NovoNordisk. Honoraria/ fusion therapy and multiple daily insulin injections in type 1
diabetes mellitus: a comparative overview and future horizons.
education support/Consultant; Abbott, AstraZeneca,
Expert Opin Drug Deliv. 2016;13(3):389-400.
Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Eli 13. Little SA, Speight J, Leelarathna L, et al. Sustained reduction in
Lilly, GlaxoSmithKline, Menarini Pharmaceuticals Merck severe hypoglycemia in adults with type 1 diabetes complicated
Sharp & Dohme and NovoNordisk. All other authors by impaired awareness of hypoglycemia: two-year follow-up in
reported to conflict of interest in relation to this review. the HypoCOMPaSS randomized clinical trial. Diabetes Care.
2018;41(8):1600-1607.
ORCID 14. Speight J, Holmes-Truscott E, Little SA, et al. Satisfaction with
the use of different Technologies for Insulin Delivery and
Jane Speight https://orcid.org/0000-0002-1204-6896
Glucose Monitoring among Adults with long-standing type 1
Pratik Choudhary https://orcid.
diabetes and problematic hypoglycemia: 2-year follow-up in
org/0000-0001-7635-4735 the HypoCOMPaSS randomized clinical trial. Diabetes Technol
Christel Hendrieckx https://orcid. Ther. 2019;21(11):619-626.
org/0000-0002-0075-828X 15. Heller S, White D, Lee E, et al. A cluster randomised trial, cost-
Thomas Crabtree https://orcid. effectiveness analysis and psychosocial evaluation of insulin
org/0000-0001-7886-8262 pump therapy compared with multiple injections during flex-
Gemma Traviss-Turner https://orcid. ible intensive insulin therapy for type 1 diabetes: the REPOSE
trial. Health Technol Assess. 2017;21(20):1-278.
org/0000-0002-1770-6216
16. Oldham V, Mumford B, Lee D, Jones J, Das G. Impact of insu-
Ramzi A. Ajjan https://orcid.org/0000-0002-1636-3725
lin pump therapy on key parameters of diabetes management
and diabetes related emotional distress in the first 12 months.
REFERENCES Diabetes Res Clin Pract. 2020;166:108281.
1. DiMeglio LA, Evans-Molina C, Oram RA. Type 1 diabetes. 17. Bosi E, Choudhary P, de Valk HW, et al. Efficacy and
Lancet. 2018;391(10138):2449-2462. safety of suspend- before-
low insulin pump technology in
2. Nathan DM. The diabetes control and complications trial/epi- hypoglycaemia- prone adults with type 1 diabetes (SMILE):
demiology of diabetes interventions and complications study at an open- label randomised controlled trial. Lancet Diabetes
30 years: overview. Diabetes Care. 2014;37(1):9-16. Endocrinol. 2019;7(6):462-472.
3. Battelino T, Danne T, Bergenstal RM, et al. Clinical targets for 18. McAuley SA, Lee MH, Paldus B, et al. Six months of hybrid
continuous glucose monitoring data interpretation: recommen- closed-loop versus manual insulin delivery with Fingerprick
dations from the international Consensus on time in range. Blood glucose monitoring in adults with type 1 diabetes: a ran-
Diabetes Care. 2019;42(8):1593-1603. domized. Controlled Trial Diabetes Care. 2020;43(12):3024-3033.
|
19. Benhamou PY, Franc S, Reznik Y, et al. Closed-loop insulin 31. Pratley RE, Kanapka LG, Rickels MR, et al. Effect of contin-
delivery in adults with type 1 diabetes in real-life conditions: a uous glucose monitoring on hypoglycemia in older adults
12-week multicentre, open-label randomised controlled cross- with type 1 diabetes: a randomized clinical trial. JAMA.
over trial. Lancet Digit Health. 2019;1(1):e17-e25. 2020;323(23):2397-2406.
20. Kropff J, DeJong J, Del Favero S, et al. Psychological outcomes 32. Feig DS, Donovan LE, Corcoy R, et al. Continuous glucose mon-
of evening and night closed-loop insulin delivery under free itoring in pregnant women with type 1 diabetes (CONCEPTT):
living conditions in people with type 1 diabetes: a 2-month ran- a multicentre international randomised controlled trial. Lancet.
domized crossover trial. Diabet Med. 2017;34(2):262-271. 2017;390(10110):2347-2359.
21. Bisio A, Gonder-Frederick L, McFadden R, et al. The impact 33. Nefs G, Bazelmans E, Marsman D, Snellen N, Tack CJ, de Galan
of a recently approved automated insulin delivery system on BE. RT-CGM in adults with type 1 diabetes improves both gly-
glycemic, sleep, and psychosocial outcomes in older adults caemic and patient- reported outcomes, but independent of
with type 1 diabetes: a pilot study. J Diabetes Sci Technol. each other. Diabetes Res Clin Pract. 2019;158:107910.
2021;16(3):663-669. 34. Jimenez- Sahagun R, Gomez Hoyos E, Diaz Soto G, et al.
22. Pinsker JE, Muller L, Constantin A, et al. Real-world patient- Impact of flash glucose monitoring on quality of life and gly-
reported outcomes and glycemic results with initiation of caemic control parameters in adults with type 1 diabetes
Control-IQ technology. Diabetes Technol Ther. 2021;23(2): mellitus small star, filled. Endocrinol Diabetes Nutr (Engl ed).
120-127. 2022;69(5):345-353.
23. Beato-Vibora PI, Gallego-Gamero F, Lazaro-Martin L, Romero- 35. Reddy M, Jugnee N, El Laboudi A, Spanudakis E, Anantharaja
Perez MDM, Arroyo-Diez FJ. Prospective analysis of the im- S, Oliver N. A randomized controlled pilot study of continu-
pact of commercialized hybrid closed-loop system on glycemic ous glucose monitoring and flash glucose monitoring in people
Control, glycemic variability, and patient- related outcomes with type 1 diabetes and impaired awareness of hypoglycaemia.
in children and adults: a focus on superiority over predic- Diabet Med. 2018;35(4):483-490.
tive low-glucose suspend technology. Diabetes Technol Ther. 36. Haskova A, Radovnicka L, Petruzelkova L, et al. Real-time CGM
2020;22(12):912-919. is superior to flash glucose monitoring for glucose Control in
24. Polonsky WH, Hood KK, Levy CJ, et al. How introduction of type 1 diabetes: the CORRIDA randomized controlled trial.
automated insulin delivery systems may influence psychosocial Diabetes Care. 2020;43(11):2744-2750.
outcomes in adults with type 1 diabetes: findings from the first 37. Bolinder J, Antuna R, Geelhoed- Duijvestijn P, Kroger J,
investigation with the Omnipod(R) 5 system. Diabetes Res Clin Weitgasser R. Novel glucose-sensing technology and hypogly-
Pract. 2022;190:109998-110007. caemia in type 1 diabetes: a multicentre, non-masked, ran-
25. Polonsky WH, Hessler D, Ruedy KJ, Beck RW, Group DS. domised controlled trial. Lancet. 2016;388(10057):2254-2263.
The impact of continuous glucose monitoring on markers of 38. Fokkert M, van Dijk P, Edens M, et al. Improved well-being
quality of life in adults with type 1 diabetes: further findings and decreased disease burden after 1-year use of flash glu-
from the DIAMOND randomized clinical trial. Diabetes Care. cose monitoring (FLARE-NL4). BMJ Open Diabetes Res Care.
2017;40(6):736-741. 2019;7(1):e000809.
26. Lind M, Polonsky W, Hirsch IB, et al. Continuous glucose 39. Al Hayek AA, Al Dawish MA. Assessing diabetes distress
monitoring vs conventional therapy for glycemic Control and sleep quality in Young adults with type 1 diabetes using
in adults with type 1 diabetes treated with multiple daily in- FreeStyle libre: a prospective cohort study. Diabetes Ther.
sulin injections: the GOLD randomized clinical trial. JAMA. 2020;11(7):1551-1562.
2017;317(4):379-387. 40. Deshmukh H, Wilmot EG, Gregory R, et al. Effect of flash glu-
27. Olafsdottir AF, Polonsky W, Bolinder J, et al. A randomized cose monitoring on glycemic control, hypoglycemia, diabetes-
clinical trial of the effect of continuous glucose monitoring on related distress, and resource utilization in the Association
nocturnal hypoglycemia, daytime hypoglycemia, glycemic vari- of British Clinical Diabetologists (ABCD) Nationwide audit.
ability, and hypoglycemia confidence in persons with type 1 di- Diabetes Care. 2020;43(9):2153-2160.
abetes treated with multiple daily insulin injections (GOLD-3). 41. Little SA, Leelarathna L, Walkinshaw E, et al. Recovery of hy-
Diabetes Technol Ther. 2018;20(4):274-284. poglycemia awareness in long-standing type 1 diabetes: a mul-
28. Heinemann L, Freckmann G, Ehrmann D, et al. Real-time ticenter 2 x 2 factorial randomized controlled trial comparing
continuous glucose monitoring in adults with type 1 dia- insulin pump with multiple daily injections and continuous
betes and impaired hypoglycaemia awareness or severe hy- with conventional glucose self-monitoring (HypoCOMPaSS).
poglycaemia treated with multiple daily insulin injections Diabetes Care. 2014;37(8):2114-2122.
(HypoDE): a multicentre, randomised controlled trial. Lancet. 42. Haynes E, Ley M, Talbot P, Dunbar M, Cummings E. Insulin
2018;391(10128):1367-1377. pump therapy improves quality of life of Young patients with
29. van Beers CA, DeVries JH, Kleijer SJ, et al. Continuous glu- type 1 diabetes enrolled in a government- funded insulin
cose monitoring for patients with type 1 diabetes and impaired pump program: a qualitative study. Can. J Diabetes. 2020;45:
awareness of hypoglycaemia (IN CONTROL): a randomised, 395-402.
open- label, crossover trial. Lancet Diabetes Endocrinol. 43. Anyanwagu U, Olaoye H, Jennings P, et al. Continuous sub-
2016;4(11):893-902. cutaneous insulin infusion (CSII) therapy at Derby teaching
30. Laffel LM, Kanapka LG, Beck RW, et al. Effect of continuous hospitals: sustained benefits in glucose control. Diabet Med.
glucose monitoring on glycemic control in adolescents and 2017;34(8):1154-1157.
Young adults with type 1 diabetes: a randomized clinical trial. 44. Beck RW, Riddlesworth T, Ruedy K, et al. Effect of continuous
JAMA. 2020;323(23):2388-2396. glucose monitoring on glycemic Control in adults with type 1
|
diabetes using insulin injections: the DIAMOND randomized diabetes: design and validation protocol of the hypo-METRICS
clinical trial. Jama. 2017;317(4):371-378. application. BMJ Open. 2022;12(2):e051651.
45. Pease A, Lo C, Earnest A, Kiriakova V, Liew D, Zoungas S. 59. Schipp J, Skinner T, Holloway E, et al. How adults with type 1
The efficacy of technology in Type 1 diabetes: a systematic re- diabetes are navigating the challenges of open-source artificial
view, network meta-analysis, and narrative synthesis. Diabetes pancreas systems: a qualitative study. Diabetes Technol Ther.
Technol Ther. 2020;22(5):411-421. 2021;23(8):546-554.
46. Barnard KD, Wysocki T, Ully V, et al. Closing the loop in adults, 60. Hendrieckx C, Poole LA, Sharifi A, et al. "It is definitely a game
children and adolescents with suboptimally controlled type 1 changer": a qualitative study of experiences with in-home over-
diabetes under free living conditions: a psychosocial substudy. night closed-loop technology among adults with type 1 diabe-
J Diabetes Sci Technol. 2017;11(6):1080-1088. tes. Diabetes Technol Ther. 2017;19(7):410-416.
47. Berget C, Messer LH, Vigers T, et al. Six months of hybrid closed 61. Greenhalgh J, Meadows K. The effectiveness of the use of
loop in the real-world: an evaluation of children and young adults patient-based measures of health in routine practice in im-
using the 670G system. Pediatr Diabetes. 2020;21(2):310-318. proving the process and outcomes of patient care: a literature
48. Messer LH, Campbell K, Pyle L, Forlenza GP. Basal-IQ tech- review. J Eval Clin Pract. 1999;5(4):401-416.
nology in the real world: satisfaction and reduction of diabe- 62. Haywood K, Marshall S, Fitzpatrick R. Patient participation in
tes burden in individuals with type 1 diabetes. Diabet Med. the consultation process: a structured review of intervention
2020;38:e14381. strategies. Patient Educ Couns. 2006;63(1–2):12-23.
49. Lewis D. History and perspective on DIY closed looping. 63. Velikova G, Booth L, Smith AB, et al. Measuring quality of life
J Diabetes Sci Technol. 2019;13(4):790-793. in routine oncology practice improves communication and pa-
50. Braune K, Gajewska KA, Thieffry A, et al. Why tient well-being: a randomized controlled trial. J Clin Oncol.
#WeAreNotWaiting- motivations and self- reported outcomes 2004;22(4):714-724.
among users of open-source automated insulin delivery systems: 64. Hernar I, Graue M, Richards D, et al. Electronic capturing of
multinational survey. J Med Internet Res. 2021;23(6):e25409. patient-reported outcome measures on a touchscreen computer
51. Litchman ML, Lewis D, Kelly LA, Gee PM. Twitter analy- in clinical diabetes practice (the DiaPROM trial): a feasibility
sis of #OpenAPS DIY artificial pancreas technology use sug- study. Pilot Feasibility Stud. 2019;5:29.
gests improved A1c and quality of life. J Diabetes Sci Technol. 65. Hendrieckx C, Halliday JA, Russell-Green S, et al. Adults with
2019;13(2):164-170. diabetes distress often want to talk with their health profession-
52. Adaptive Platform Trials Coalition. Adaptive platform trials: als about it: findings from an audit of 4 Australian specialist
definition, design, conduct and reporting considerations. Nat diabetes clinics. Can J Diabetes. 2020;44(6):473-480.
Rev Drug Discov. 2019;18(10):797-807. 66. Todd PJ, Edwards F, Spratling L, et al. Evaluating the relation-
53. Lawton J, Blackburn M, Rankin D, et al. Participants' expe- ships of hypoglycaemia and HbA1c with screening-detected di-
riences of, and views about, daytime use of a day-and-night abetes distress in type 1 diabetes. Endocrinol Diabetes Metab.
hybrid closed- loop system in real life settings: longitudinal 2018;1(1):e00003.
qualitative study. Diabetes Technol Ther. 2019;21(3):119-127. 67. Barnard-Kelly K, Kelly RC, Chernavvsky D, Lal R, Cohen L,
54. Cleal B, Nexo MA, Schmidt S, Rytter K, Norgaard K, Willaing Ali A. Feasibility of spotlight consultations tool in routine care:
I. Values and control in type 1 diabetes beyond glycemic out- real-world evidence. J Diabetes Sci Technol. 2022;16(4):939-944.
comes: a qualitative interview study of everyday life with an 68. Greenhalgh J, Gooding K, Gibbons E, et al. How do patient re-
insulin pump. Chronic Illn. 2021;18(3):620-633. ported outcome measures (PROMs) support clinician-patient
55. Speight J, Holmes-Truscott E, Harvey DM, et al. Structured type communication and patient care? A realist synthesis. J Patient
1 diabetes education delivered in routine care in Australia re- Rep Outcomes. 2018;2:42.
duces diabetes-related emergencies and severe diabetes-related 69. Hendrieckx C, Halliday JA, Beeney LJ, Speight J. Diabetes and
distress: the OzDAFNE program. Diabetes Res Clin Pract. Emotional Health: A Practical Guide for Healthcare Professionals
2016;112:65-72. Supporting Adults with Type 1 and Type 2 Diabetes. Diabetes
56. Snyder CF, Aaronson NK, Choucair AK, et al. Implementing UK; 2019.
patient-reported outcomes assessment in clinical practice:
a review of the options and considerations. Qual Life Res.
2012;21(8):1305-1314. How to cite this article: Speight J, Choudhary P,
57. Movsas B, Hunt D, Watkins-Bruner D, et al. Can electronic Wilmot EG, et al. Impact of glycaemic technologies
web-based technology improve quality of life data collection? on quality of life and related outcomes in adults
Analysis of radiation therapy oncology group 0828. Pract
with type 1 diabetes: A narrative review. Diabet
Radiat Oncol. 2014;4(3):187-191.
58. Soholm U, Broadley M, Zaremba N, et al. Investigating the day-
Med. 2023;40:e14944. doi: 10.1111/dme.14944
to-day impact of hypoglycaemia in adults with type 1 or type 2

Diabetic Medicine - 2022 - Speight - Impact of Glycaemic Technologies On Quality of Life and Related Outcomes in Adults

Uploaded by

Copyright:

Available Formats

Diabetic Medicine - 2022 - Speight - Impact of Glycaemic Technologies On Quality of Life and Related Outcomes in Adults

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Diabetic Medicine - 2022 - Speight - Impact of Glycaemic Technologies On Quality of Life and Related Outcomes in Adults

Uploaded by

Copyright:

Available Formats

Received: 7 February 2022

| Accepted: 18 August 2022

Impact of glycaemic technologies on quality of life and

Jane Speight1,2 | Pratik Choudhary3 | Emma G. Wilmot4,5 |

Diabetic Medicine. 2023;40:e14944. ﻿ wileyonlinelibrary.com/journal/dme | 1 of 23

study design, population and length of follow-­up. Differences in PRO findings

T A B L E 1 Definition of key constructs used when validating

Study name Study duration &

Technologya (Citation) designb (N) Inclusion criteriac PROMsd Findings

Study name Study duration &

Study name Study duration &

Study name Study duration &

≤8.0% (p = 0.019), while those with HbA1c >8.0% demonstrating no

Study name Study duration &

Confidence in Diabetes Self-­care 6-­month Cohort (Nefs et al, 2020)33 +a

(HFS / HFS-­II), Worry 2017)32

REPOSE: 6/12/24-­month RCT (Heller et al, 2017) − −

Impaired awareness of hypoglycaemia

IN CONTROL: 16-­week RCT (van Beers et al, 2016)29 − −

Questionnaire (GME-­Q) HypoCOMPaSS: 24-­month (preference) (Speight et al, − −

5.2 | Diabetes-­specific PROMs 6.1 | Implication of study design in

Implications of study design

Randomised controlled trials Cohort studies

Advantages Disadvantages Advantages Disadvantages

Partially-randomised preference trials

PROMs are undoubtedly valuable tools to inform decision AUTHOR CONTRIBUTIONS

You might also like

Diabetic Medicine. 2023;40:e14944. wileyonlinelibrary.com/journal/dme | 1 of 23

study design, population and length of follow-up. Differences in PRO findings

Confidence in Diabetes Self-care 6-month Cohort (Nefs et al, 2020)33 +a

(HFS / HFS-II), Worry 2017)32

REPOSE: 6/12/24-month RCT (Heller et al, 2017) − −

IN CONTROL: 16-week RCT (van Beers et al, 2016)29 − −

Questionnaire (GME-Q) HypoCOMPaSS: 24-month (preference) (Speight et al, − −

5.2 | Diabetes-specific PROMs 6.1 | Implication of study design in