. 2023 Oct 5;3:132. doi: 10.1038/s43856-023-00358-x

Table 6.

AID Articles with Precision-Directed Analyses or Subpopulations of Interest.

Author year (trial/cohort title) location	Analysis type/study design/number of randomized	Primary endpoint (SIG vs. NS)—% dropout multiple test correction performed?	Subgroup analyses—technology superior to standard of care	Subgroup analyses—technology comparable to standard of care
Thabit⁹² (APCam8; AP@home04) Europe	Primary Randomized crossover 58	TIR/time in a tight range overnight at 3 months (SIG)—0% No	Age AID use in adults (aged ≥ 18 years) and AID use at night in children (aged 6–17 years) improved time in the target range (p < 0.001 for both).	Age Secondary nighttime endpoints were similar between age groups except for TBR, time <50 mg/dL, and number of nights with glucose <63 mg/dL, which were not improved in ages 6–17 years (n = 25).
Tauschmann⁹¹ (APCam11) USA, UK	Primary RCT 86	TIR 3 months (SIG)—0% No	Age < 13 vs. 13–21 vs. ≥21, sex, HbA1c < 8.5 vs. ≥8.5% AID use in the entire cohort (aged ≥ 6 years) increased TIR by 10.8% (95% CI: 8.2–13.5%; p < 0.0001)
Collyns⁸³ New Zealand	Primary Randomized crossover 60	TIR 3 months (SIG)—2% No	Age AID use increased TIR in ages 7–13 years (n = 19; +11.8 ± 7.4%), ages 14–21 (n = 14; +14.4 ± 8.4%), and ages 22–80 (n = 26; +11.9 ± 9.5%; p < 0.001 for all ages).
Ware⁹⁰ (DAN05) USA, UK	Primary RCT 133	Δ HbA1c 6 months (SIG)—8% Yes^a	Age, device, device use AID use in children, aged 6–18 years, improved HbA1c (–0.32% [95% CI: −0.59 to −0.04]; p = 0.023). CamAPS FX (n = 46) showed an HbA1c change of −1.05% {95% CI: −1.43 to −0.67; p < 0.0001) with median 93% device use.	Device, device use FlorenceM (n = 75) showed an HbA1c change of +0.21% (95% CI: −0.14 to 0.57; p = 0.23) with median 40% device use.
Benhamou⁷⁷ (Diabeloop WP7) France	Primary Randomized crossover 68	TIR 3 months (SIG)—7% No	HbA1c Increased TIR for ages ≥ 18 years (+9.2% [95% CI: 6.4–11.9]; p < 0.0001) and across HbA1c groups: <7% (n = 16), 7–7.4 (n = 10), 7.5–7.9 (n = 18), 8–8.4 (n = 8), and ≥8.5 (n = 11; p < 0.05 for all)
Breton⁸⁰ (iDCL) USA	Primary RCT 101	TIR 4 months (SIG)—1% Unsure	Age, sex, BMI, household income, parental education, HbA1c AID increased TIR in ages 6–13 years by +11% (95% CI: 7–14; p < 0.001).
Kovatchev⁸⁴ (iDCL) USA	Primary RCT 127	TBR & TAR 3 months (SIG)—2% Yes	Age, sex, BMI, HbA1c, C-peptide, TBR, TAR, previous CGM use, household income, education AID reduced TBR (−1.7% [95% CI: −2.4 to −1.0]; p < 0.0001 superiority) & TAR (−3.0% [95% CI: −6.2 to 0.1]; p < 0.0001 noninferiority) in ages ≥ 14 years.	Sex, TBR In male participants, there was no change in TAR (p_interaction = 0.018) compared to −11% in females (n = 59). Those with baseline TBR ≤ 4.0% (n = 63) had less improvement in TBR (−1%) (p_interaction = 0.0406) TBR > 4.0% (−3%).
Ekhlaspour⁹⁴ (iDCL) USA	Secondary RCT 168	TIR 6 months (SIG)—0% No	Age, device use AID increased TIR in ages 6–13 years by +2.6 hours with no significant interaction with baseline device use for TIR or TBR.
McAuley⁸⁶ (Australian JDRF Closed-Loop)	Primary RCT 120	TIR 6 months (SIG)—8% Unsure	Sex, insulin modality prior to trial AID increased TIR in ages 25–70 years by +15% (95% CI: 11–19; p < 0.0001) similarly for males/females, MDI/pump users.

Author year (trial/cohort title) location

Analysis type/study design/number of randomized

Primary endpoint (SIG vs. NS)—% dropout multiple test correction performed?

Subgroup analyses—technology superior to standard of care

Subgroup analyses—technology comparable to standard of care

Thabit⁹² (APCam8; AP@home04)

Europe

Primary

Randomized crossover

TIR/time in a tight range overnight at 3 months (SIG)—0%

Age

AID use in adults (aged ≥ 18 years) and AID use at night in children (aged 6–17 years) improved time in the target range (p < 0.001 for both).

Age

Secondary nighttime endpoints were similar between age groups except for TBR, time <50 mg/dL, and number of nights with glucose <63 mg/dL, which were not improved in ages 6–17 years (n = 25).

Tauschmann⁹¹ (APCam11)

USA, UK

Primary

RCT

TIR 3 months (SIG)—0%

Age < 13 vs. 13–21 vs. ≥21, sex, HbA1c < 8.5 vs. ≥8.5%

AID use in the entire cohort (aged ≥ 6 years) increased TIR by 10.8% (95% CI: 8.2–13.5%; p < 0.0001)

Collyns⁸³

New Zealand

Primary

Randomized crossover

TIR 3 months (SIG)—2%

Age

AID use increased TIR in ages 7–13 years (n = 19; +11.8 ± 7.4%), ages 14–21 (n = 14; +14.4 ± 8.4%), and ages 22–80 (n = 26; +11.9 ± 9.5%; p < 0.001 for all ages).

Ware⁹⁰ (DAN05)

USA, UK

Primary

RCT

133

Δ HbA1c 6 months (SIG)—8%

Yes^a

Age, device, device use

AID use in children, aged 6–18 years, improved HbA1c (–0.32% [95% CI: −0.59 to −0.04]; p = 0.023). CamAPS FX (n = 46) showed an HbA1c change of −1.05% {95% CI: −1.43 to −0.67; p < 0.0001) with median 93% device use.

Device, device use

FlorenceM (n = 75) showed an HbA1c change of +0.21% (95% CI: −0.14 to 0.57; p = 0.23) with median 40% device use.

Benhamou⁷⁷ (Diabeloop WP7)

France

Primary

Randomized crossover

TIR 3 months (SIG)—7%

HbA1c

Increased TIR for ages ≥ 18 years (+9.2% [95% CI: 6.4–11.9]; p < 0.0001) and across HbA1c groups: <7% (n = 16), 7–7.4 (n = 10), 7.5–7.9 (n = 18), 8–8.4 (n = 8), and ≥8.5 (n = 11; p < 0.05 for all)

Breton⁸⁰ (iDCL)

USA

Primary

RCT

101

TIR 4 months (SIG)—1%

Unsure

Age, sex, BMI, household income, parental education, HbA1c

AID increased TIR in ages 6–13 years by +11% (95% CI: 7–14; p < 0.001).

Kovatchev⁸⁴ (iDCL)

USA

Primary

RCT

127

TBR & TAR 3 months (SIG)—2%

Yes

Age, sex, BMI, HbA1c, C-peptide, TBR, TAR, previous CGM use, household income, education

AID reduced TBR (−1.7% [95% CI: −2.4 to −1.0]; p < 0.0001 superiority) & TAR (−3.0% [95% CI: −6.2 to 0.1]; p < 0.0001 noninferiority) in ages ≥ 14 years.

Sex, TBR

In male participants, there was no change in TAR (p_interaction = 0.018) compared to −11% in females (n = 59). Those with baseline TBR ≤ 4.0% (n = 63) had less improvement in TBR (−1%) (p_interaction = 0.0406) TBR > 4.0% (−3%).

Ekhlaspour⁹⁴ (iDCL)

USA

Secondary

RCT

168

TIR 6 months (SIG)—0%

Age, device use

AID increased TIR in ages 6–13 years by +2.6 hours with no significant interaction with baseline device use for TIR or TBR.

McAuley⁸⁶ (Australian JDRF Closed-Loop)

Primary

RCT

120

TIR 6 months (SIG)—8%

Unsure

Sex, insulin modality prior to trial

AID increased TIR in ages 25–70 years by +15% (95% CI: 11–19; p < 0.0001) similarly for males/females, MDI/pump users.

^aFDR Adjustment using Benjamini Hochberg procedure, IAH impaired awareness of hypoglycemia.

SH severe hypoglycemia, IRR incidence rate ratio.

NH nocturnal hypoglycemia, RAS renin–angiotensin system.

MAGE mean amplitude of glucose excursion (mean of blood glucose values exceeding 1 SD from the 24‐h mean blood glucose, used as an index of glycemic variability).