INTRODUCTIONYou may edit the size of the font, the sections distribution and the colour scheme if wanted. All modifications concerning the background/title display are to be done at a Master Slide level.

CONCLUSIONSHighlight this text and replace with your own text.

RESULTS

METHODTips for making a successful poster:• Re-write your paper into poster format i.e.. simplify everything, avoid data

overkill. • Headings of more than 6 words should be in upper and lower case, not all

capitals. Simplify the titles.• Do not to write whole sentences in capitals or underline to stress your point,

use bold characters instead.• When laying out your poster, leave white space around your text. Don’t

overcrowd your poster.• Spell check and get someone else to proof-read.

ACKNOWLEDGEMENTSHighlight this text and replace with your own text.

AIMHow to use this poster template:

Highlight this text and replace it by typing in your own text, or copy and paste your text from a MS Word document or a PowerPoint slide presentation.

The body text / font size should be no smaller than 20 points. Keep body text left-aligned, do not justify text.

REFERENCESHighlight this text and replace with your own text.

Can be Vancouver style i.e. 1 Meyer J-P et al. The treatment of high grade superficial bladder cancer and carcinoma in situ with BCG – a questionnaire survey of Consultant practice in England and Wales. UrolOncol 2002; 2;: 77-80

The recommended font for captions is Calibri, no smaller than 15 pt. Left aligned if it refers to a figure on its left. Start the captions right at the top edge of the picture (graph or photo).

Importing / inserting files:

Images such as photographs, graphs, diagrams, logos, etc, can be added to the poster. Avoid long numerical tables as they will be difficult to read.

To insert images, go through the menus as follows: Insert / Picture / From File. Locate the file on your computer, select and press OK. Please ensure all image files are JPEG/PNG, other formats may cause images to get corrupted after saving.

Be aware of the image size you are importing.

Notes about graphs…

For simple graphs use MS Excel, or create the graph directly in PowerPoint.

Graphs created in a scientific graphing program (e.g.. Sigma Plot, Prism, SPSS, Statistics) should be saved as JPEG or TIFF.

The recommended font for captions is Calibri, no smaller than 15 pt. Left aligned if it refers to a figure on its left. Start the captions right at the top edge of the picture (graph or photo).

Importing / inserting files…

Images such as photographs, graphs, diagrams, logos, etc, can be added to the poster. Avoid long numerical tables.

To insert images, go through the menus as follows: Insert / Picture / From File. Locate the file on your computer, select and press OK. Please ensure all image files are JPEG.

Be aware of the image size you are importing.

Notes about graphs…

For simple graphs use MS Excel, or create the graph directly in PowerPoint.

Graphs created in a scientific graphing program (e.g.. Sigma Plot, Prism, SPSS, Statistics) should be saved as JPEG or TIFF.

CONTACT INFORMATION Highlight this text and replace with your own text.

P o s t e r t i t l e g o e s h e r e Authors, FIRST NAME INITIAL, SURNAME, presenting author(s) underlined, Affiliations1 numbered in superscripte.g. B. SMITH 1, N. HARRISON 2 2 and P. MATTHEWS2

1University of New South Wales, Sydney, Australia2Royal Brisbane Hospital, Brisbane, Australia

QR CODE

The CGM system was inserted in 90 participants in PRECISE II (Table 1). SW602 resulted in 87% of CGM values within 15/15% of reference values over the total glucose range of 40–400 mg/dL (Figure 1) (15,753 matched pairs). Of note, accuracy ≤54 mg/dL showed 89% of CGM values within 15 mg/dL (Figure 2). The overall MARD improved to 8.5% (95% CI: 8.0, 9.1) (Figure 3).

Results

Application of the updated glucose calculation algorithm (SW602) to the raw sensor data from PRECISE II study using a post-hoc analysis method improved the MARD from 8.8% to 8.5% with a strong 15/15% metric of 87%. Severe hypoglycemic accuracy of ≤54 mg/dL was also robust with 89% of readings within 15 mg/dL.

Discussion

An improved MARD of 8.5% was demonstrated for the Eversense CGM system using an updated glucose calculation algorithm through 90 days of sensor wear time.

ConclusionTable 1: Baseline Characteristic (n=90)

Variable

Age, years (SD)Gender, male, n (%)Diabetes duration, years (SD)Baseline HbA1c, % (SD)Insulin-pump use, n (%)BMI, kg/m2 (SD)Type 1 Diabetes, n (%)

Value

45.1 (16.2)54 (60)

20.1 (13.7)7.6 (1.2)43 (48)

29.1 (6.2)61 (68)

Figure 2: 15/15% – Metric Comparison Measurement Range Segments

50

60

70

80

90

study software SW602

71-180

85% 85%

≤54

83% 89%

>180

87% 85%

>54-<71

86% 87%

15/15% metric study software vs SW602

[%]

[mg/dL]

Figure 1: 15/15% – Metric Comparison Total Measurement Range

7678808284868890

study software SW602

86% 87%[%]

15/15% metric40-400 [mg/dL]

Figure 3: MARD Comparison

7

7.5

8

8.5

9

8.8% 8.5%[%]

study software SW602

40-400 [mg/dL]MARD [%]

The algorithm to calculate glucose was updated (SW602) after the study was conducted. The raw sensor data obtained from the PRECISE II study was post-processed using SW602 and changes in performance were evaluated. The goal for SW602 was accuracy improvement early in sensor life and in the hypoglycemic range overall. The e�cacy measures of MARD between paired Eversense and Yellow Springs Instrument (YSI) reference measurements and percent of system agreement within 15 mg/dL or 15% of reference glucose measurements (15/15% metric) through 90 days were evaluated.

Methods

The safety and accuracy of the new implantable Eversense CGM system through 90 days in participants with type 1 or type 2 diabetes (T1D, T2D) was evaluated using a prospective, multi-center, blinded study design (PRECISE II). The study demonstrated that the CGM was safe with a mean absolute relative di�erence (MARD) of 8.8% using the study glucose algorithm software. The objective of this evaluation was to assess the e�ect of a new glucose calculation algorithm on the performance of the Eversense CGM.

Objective

Evaluation of the accuracy of an implantable CGM system with an updated glucose calculation algorithm

Mark Christiansen, MD1; Timothy S Bailey, MD2; Leslie J Kla�, MD, PhD3; Ron Brazg, MD3; Anna R Chang, MD4; Carol J Levy, MD5; David Lam, MD5; Douglas S Denham, DO6; Bruce W Bode, MD7; Ravi Rastogi, PhD8; and Katherine S Tweden, PhD8

1Diablo Clinical Research Inc., Walnut Creek, CA, USA; 2AMCR Institute Inc., Escondido, CA, USA; 3Rainer Clinical Research Center Inc., Renton, WA, USA; 4John Muir Health Clinical Research Center, Concord, CA, USA; 5Mount Sinai Diabetes Center, New York, NY, USA; 6Clinical Trials of Texas Inc., San Antonio, TX, USA; 7Atlanta Diabetes Associates, Atlanta, GA, USA; 8Senseonics Inc. Germantown, MD, USA

CONTACT INFORMATION Senseonics, Inc. 20451 Seneca Meadows Parkway, Germantown, MD 20876 Phone: 844-SENSE4U (736-7348) Email: Katherine.Tweden@senseonics.com