basal-bolus insulin_diabetesjournal.pdf

7/30/2019 basal-bolus insulin_diabetesjournal.pdf

1/7

Assessment of Patient-Led orPhysician-Driven Continuous GlucoseMonitoring in Patients With Poorly

Controlled Type 1 Diabetes UsingBasal-Bolus Insulin RegimensA 1-year multicenter study

JEAN-PIERRE RIVELINE, MD1,2

PAULINE SCHAEPELYNCK, MD3

LUCY CHAILLOUS, MD, PHD4

ERIC RENARD, MD, PHD5

AGNS SOLA-GAZAGNES, MD6

ALFRED PENFORNIS, MD, PHD7

NADIA TUBIANA-RUFI, MD8

VRONIQUE SULMONT, MD9

BOGDAN CATARGI, MD, PHD10

CLINE LUKAS, MD11

RGIS P. RADERMECKER, MD12

CHARLES THIVOLET, MD, PHD13

FRANOIS MOREAU, MD14

PIERRE-YVES BENHAMOU, MD, PHD15

BRUNO GUERCI, MD, PHD16

ANNE-MARIE LEGUERRIER, MD17

LUC MILLOT, MD18

CLAUDE SACHON, MD19

GUILLAUME CHARPENTIER, MD1

HLNE HANAIRE, MD20

FOR THE EVADIAC SENSOR STUDYGROUP

OBJECTIVEdThe benefits of real-time continuous glucose monitoring (CGM) have been dem-onstrated in patientswith type 1 diabetes. Our aim was to compare the effect of two modes of use ofCGM, patient led or physician driven, for 1 year in subjects with poorly controlled type 1 diabetes.

RESEARCH DESIGN AND METHODSdPatients with type 1 diabetes aged 860 yearswith HbA1c$8% were randomly assigned to three groups (1:1:1). Outcomes for glucose controlwere assessed at 1 year for two modes of CGM (group 1: patient led; group 2: physician driven)versus conventional self-monitoring of blood glucose (group 3: control).

RESULTSd

A total of 257 subjects with type 1 diabetes underwent screening. Of these, 197 wererandomized, with 178 patients completing the study (age: 36 6 14 years; HbA1c: 8.9 6 0.9%).HbA1c improved similarly in both CGM groups and was reduced compared with the control group(group 1 vs. group 3:20.52%, P= 0.0006; group 2 vs. group 3:20.47%, P= 0.0008; groups1 + 2vs. group 3:20.50%, P, 0.0001).The incidence of hypoglycemia was similar in the three groups.Patient SF-36 questionnaire physical health score improved in both experimental CGM groups (P=0.004). Sensor consumption was 34% lower in group 2 than in group 1 (median [Q1Q3] con-sumption: group 1: 3.42/month [2.203.91] vs. group 2: 2.25/month [1.272.99], P= 0.001).

CONCLUSIONSdBoth patient-led and physician-driven CGM provide similar long-termimprovement in glucose control in patients with poorly controlled type 1 diabetes, but thephysician-driven CGM mode used fewer sensors.

Diabetes Care 35:965971, 2012

Since the report of the results of theDiabetes Control and Complica-tions Trial (1), so-called intensified

insulin therapy has been considered thegold standard for patients with type 1diabetes, together with frequent self-

monitoring of blood glucose (SMBG),patient education about diabetes manage-ment, and close follow-up and support bythe attending health care team. This inten-sified insulin therapy corresponds to insu-lin regimens involving multiple dailyinjections (MDI) or continuous subcuta-neous insulin infusion (CSII). The avail-ability of insulin analogs has enabledfurther improvements in the safety andefficacy of basal-bolus insulin therapy(2). However, a significant subset of pa-tients with type 1 diabetes still fails toreach optimal HbA1c levels and/or expe-

riences debilitating BG variability andrecurrent hypoglycemia (3,4). The avail-ability of devices for real-time continuousglucose monitoring (CGM) recently hasaroused considerable interest among pa-tients and physicians, who anticipatedthat such devices might provide potentialbenefits in terms of BG control. Indeed, sev-eral randomized controlled studies demon-strate that use of CGM could improveHbA1c levels in patientswith type 1 diabeteswith no increased risk of hypoglycemia

c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

From the 1Department of Diabetes and Endocri-nology, Centre Hospitalier Sud Francilien,Corbeil-Essonnes, France; the 2UniversitPierre et Marie Curie-Paris 6, UMRS 872, Paris,France; the 3Service de Nutrition-Endocrinologie-Maladies Mtaboliques, University HospitalSainte Marguerite, Marseille, France; the 4In-stitut du thorax, Service d endocrinologie-diabtologie, Centre Hospitalier Universitairede Nantes, Nantes, France; the 5Department ofEndocrinology, Diabetes, and Nutrition andCentres dInvestigation Clinique INSERM 1001,Montpellier University Hospital, Montpellier,France; the 6Service de Diabtologie, AssistancePublique-Hpitaux de Paris, Centre Hospitalier

Universitaire Htel-Dieu, Paris, France; the7Department of Endocrinology-Metabolismand Diabetology-Nutrition, University Hospital ofBesanon and EA 3920, University of Franche-Comte, Besanon, France; the 8Service Endo-crinologie et Centres dInvestigation Clinique,Hpital Robert Debr, APHP, Paris, France; the9Service de pdiatrie A, American MemorialHospital, Centre Hospitalier Universitaire deReims, Reims, France; the 10Department ofEndocrinology,Centre Hospitalier UniversitaireBordeaux, Pessac, France; the 11Service den-docrinologie, Centre Hospitalier Universitairede Reims, Reims, France; the 12Division of Di-abetes, Nutrition, and Metabolic Disorders,

Department of Medicine, Centre HospitalierUniversitaire Sart Tilman, University of Lige,Lige, Belgium; the 13Department of Endocri-nology, Diabetes, and Nutrition, Hopital Lyon-Sud, HCL, Universit Lyon 1, INSERM 1060,Lyon, France; the 14Department of Endocri-nology and Diabetology, University Hospital,Strasbourg, France; the 15Department of En-docrinology, Ple DigiDune, Grenoble Uni-versity Hospital, Joseph-Fourier University,Grenoble, France; the 16Universit de Nancy Iet service de diabtologie, maladies mtabo-liques et maladies de la nutrition, Hpitauxde Brabois, hpital d adultes, Centre Hospital-ier Universitaire de Nancy, Nancy, France; the

care.diabetesjournals.org DIABETES CARE, VOLUME 35, MAY 2012 965

C l i n i c a l C a r e / E d u c a t i o n / N u t r i t i o n / P s y c h o s o c i a l R e s e a r c h

O R I G I N A L A R T I C L E


2/7

(57). However, most of the reportedstudies were short-term (3 to 6 months)and included close monitoring by theinvestigators, thus precluding any conclu-sions about expected outcomes in commonclinical practice and in the long-term.

While compliance with CGM therapy wasidentified as a strong predictor of HbA1c

reduction, the role of specific patient edu-cation in the use of CGM data to adjustinsulin therapy and the contribution ofpatient empowerment havenot beenwidelyexamined. Hence, it is still not knownwhether patient-led or physician-drivenCGM use and management have differentoutcomes. Answering such questionswould be of great value, considering theadditional cost of CGM use. We thereforedesigned this randomized multicenter con-trolled study to assess in poorly controlledpatients with type 1 diabetes the outcomeson glucose control and quality of life (QoL)of two CGM approaches following a simi-lar initial specific education procedure:patient self-management with CGM orphysician-prescribed use of CGM.

RESEARCH DESIGN ANDMETHODSdPatients were recruited at19 diabetes care centers. Inclusion criteriaincluded agebetween 8 and 60 years, type1 diabetes diagnosed.12 months earlier,treatment by insulin analogs using eitherCSII or MDI, HbA1c level $8.0%, andSMBG performed at least twice daily. Be-

fore inclusion, the patients were instruc-ted in the technical use of the study CGMdevice and were required to wear itduring a 10-day test period to confirmtheir ability and willingness to useCGM. The trial was approved by the eth-ics committeeof theParis VI Medical Fac-ulty. All patients (or the parents ofminors) had read the patient informationsheet and signed informed consent formsbefore enrolment (Clinical trial reg. no.NCT00726440).

Study protocolRandomization. Patients meeting the in-clusion criteria were randomly assigned1:1:1 to one of the three groups: patient-led

use of CGM (group 1), physician-pre-scribed CGM (group 2), or conventionalSMBG (group 3 [control group]) for 12months.

In group 1, use of the CGM devicewas managed entirely by the patientsthemselves. Patients were advised to useCGM continuously throughout the study,

as they would with a glucose meter, es-pecially if glucose targets were not ach-ieved. In group 2, use of the CGM devicewas prescribed by the patient s physician,who asked the patient to use the sensorsintermittently according to guidelinesbased on glucose outcomes: 2 weeks sen-sor use per month during the first 3months, thereafter continuing either inthe same manner or with more intensiveuse during the following 3 months if atany visit the patient presented one of thefollowing criteria: HbA1c $7.5%, greaterthan four mild hypoglycemic episodesper week, or at least one severe hypogly-cemic episode. Thus, use of the sensorscould be gradually increased every 3months to 20, 25,or even 30 days/month.In the control group (group 3), patientswere asked to carry out standard homeSMBG. Abbott Diabetes Care providedthe Navigator CGM systems and theirconsumables, as well as the home glucosemeters and test strips used for the study.Training. At the time of inclusion, allpatients received intensive educationabout target glucose values, insulin dose

management, and insulin-to-carbohydrateratios and correction factors, and theywere asked to perform SMBG at leastthree times daily. Patients randomized togroups 1 and 2 were provided with theFreeStyle Navigator glucose needle-typesensor system for 1 year (Abbott DiabetesCare, Alameda, CA) (8). These patientsreceived specific training in how to ana-lyze and make use of the CGM data and toconfirm glucose values using the meter in-cluded in the Navigator device beforemaking therapeutic decisions (see Sup-plementary Appendices 13). The pa-tients skills were assessed quarterlyduring the study by means of a shortquestionnaire comprising six questions:

Is the patient able to 1) identify fastingand premeal glucose levels, 2) interpretthese glucose results and adapt the insulindoses, 3) identify postmeal glucose levels,4) interpret these glucose results andadaptthe insulin doses, 5) identify low glucoselevels during the night, and 6) interpretthe nocturnal glucose levels and adapt

the basal insulin doses?Visits. Visits were scheduled 20 daysafter randomization and at 3, 6, 9, and12 months. Patients were asked to per-form an 8-point BG profile in the weekpreceding each visit. At each of thesevisits, HbA1c was measured and adverseevents were reported. Adverse events col-lected included mild hypoglycemia (de-fined as an SMBG value ,70 mg/dL orsymptoms of low BG) during the preced-ing week, severe hypoglycemia (definedas an event requiring assistance from an-other person), ketoacidosis, unexpectedstudy- or device-related events, and anyserious adverse event, regardless of cause,since the previous visit. In both CGMgroups, the glucose data were down-loaded from the device memory and theamount of actually used sensors was re-corded. Insulin dose adjustment was dis-cussed with the patient on the basis ofCGM graphs and according to individualglucose targets. Further attention waspaid to specific circumstances, such asphysical activity or unusual meals withhigh carbohydrate content. Physicians

evaluated the patients skill in managinginsulin doses using the six-item question-naire. Training was considered optimal ifthe six items were answered positively ateach visit throughout the entire study andnonoptimal if at least one item was not an-swered positively at any visit. Alarm set-tings, sensor insertion, and calibrationtechniques were checked. QoL was as-sessedat baseline and at the end of the studyusing the Diabetes Quality of Life (DQoL)(9) and SF-36 questionnaires (911).

Evaluation criteriaThe main aim of this study was to assessthe impact on BG control of 1 year ofCGM using two approaches to sensor

c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

17Unit dEndocrinologie, Dpartement deMdecine, Hpital Sud, Centre HospitalierUniversitaire Rennes, Rennes, France; the18Department of Endocrinology, Diabetes, andMetabolism, University Hospital, Saint-Etienne,France; the 19Service de Diabtologie-MtabolismeGroupe Hospitalier Piti-Salptrire, Paris,France; and the 20Ple Cardio-Vasculaireet Mtabolique, Service de Diabetologie,

Centre Hospitalier Universitaire de Tou-louse, Universit de Toulouse, Toulouse,France.

Corresponding author: Jean-Pierre Riveline, [email protected].

Received 16 October 2011 and accepted 31 January2012.

DOI: 10.2337/dc11-2021. Clinical trial reg. no.NCT00726440, clinicaltrials.gov.

This article contains Supplementary Data online athttp://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1.

A slide set summarizing this article is available online. 2012 by the American Diabetes Association. Readers

may use this article as long as the work is properlycited, the use is educational and not for profit, andthe work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/for details.

966 DIABETES CARE, VOLUME 35, MAY 2012 care.diabetesjournals.org

One-year CGM in type 1 diabetic patients
http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1mailto:[email protected]:[email protected]://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://creativecommons.org/licenses/by-nc-nd/3.0/http://creativecommons.org/licenses/by-nc-nd/3.0/http://creativecommons.org/licenses/by-nc-nd/3.0/http://creativecommons.org/licenses/by-nc-nd/3.0/http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1mailto:[email protected]:[email protected]://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc11-2021/-/DC1


3/7

use (patient led or physician driven) com-paratively with current standard SMBG inpatients with poorly controlled type 1diabetes. The primary end point was thereduction in HbA1c at 12 months versusbaseline. A 0.5% change in HbA1c valuewas considered clinically meaningful. Onthe basis of a single-sided two-sample

t test with a 95% CI, a sample size of 50patients per group was required to ensurestatistical power of 80% to demonstrate a0.5% difference between the groups re-garding absolute change in HbA1c betweenbaseline and 12 months. A total of 180 pa-tients had to be randomized to achieve anevaluable dataset of 150 patients while al-lowing for a 15% drop-out rate per groupand equal distribution of patients betweencenters. Because of the nature of the treat-ment, the study was not blinded.

The secondary objectives were evalu-ation of changes in glucose variability,frequency of mild and/or severe hypogly-cemic events, changes in QoL, efficacy ofCGM in patients treated by CSII and MDI,and efficacy of CGM in optimally edu-cated patients versus nonoptimally edu-cated patients. Secondary end pointsincluded SD of BG calculated from the8-point BG profiles, number of mild andsevere hypoglycemic events, number ofketoacidosis episodes, change in QoL scoresfrom baseline to 12 months, and dailyinsulinrequirements. Finally, the relationship be-tween compliance (i.e., mean number of

sensors used per month) and DHbA1c wasevaluated in each study group.

Statistical analysisThe full analysis set (FAS) was defined asall randomized patients having at leastone postbaseline HbA1c available. MissingHbA1c values during follow-up were re-placed by the mean of the previous andfollowing measurements where available(2.5% of HbA1c measurements). MissingHbA1c values at the end of follow-up werereplaced using the last observation carriedforward method (9.7% of HbA1c mea-surements). The level of significance wasa = 0.05. For comparison of CGM-usegroups versus the control group (groups1+2vs.3,group1vs.3,andgroup2vs.3),a superiority hypothesis was tested (one-sided situation). For comparison ofgroup 1 versus group 2, a difference hy-pothesis was tested (two-sided situation).Repeated assessments over time were an-alyzed by means of a mixed model withrepeated measures (ANCOVA), using anautoregressive correlation structure tak-ing treatment groups, visits, and their

interaction asfixed factorsand baseline val-ues and age as covariates. Changes frombaseline were analyzed using ANCOVA,taking baseline values and age as covari-ates. Binary criteria assessing HbA1c at12months (HbA1c level,7.5%), changes inHbA1c at the end of follow-up (a relativedecrease.10% between baseline and 12

months), and severe hypoglycemia epi-sodes during follow-up were comparedbetween groups using logistic regressionwith adjustment of two covariates (base-line values and age). Sensor consumptionwas well compared using a Wilcoxon ranksum test between groups 1 and 2.

RESULTS

PatientsFrom May 2008 to June 2009, 257 pa-tients with type 1 diabetes were screenedfor inclusion in the study. After the 10-day test period using CGM, 197 patientsfulfilled the inclusion criteria and wererandomly assigned. An additional 19patients were excluded from the studyanalysis owing to lack of any HbA1c re-sults between baseline and the end ofthe study. The FAS population thus com-prised 178 patients. Compared with theFAS population, patients failing thescreening test (n = 60) were younger,had diabetes of shorter duration, had pre-sented more ketoacidosis events, and hadlower levels of education (Table 1). Using

the study randomization procedure, theFAS population included 62 patients ingroup1,55ingroup2,and61ingroup3.The patients baseline characteristics werecomparable between the three groups ex-cept for a higher number of patients withone or more episodes of severe hypoglyce-mia during the previous year in group 1(Table 1). The per-protocol populationexcluded 3 FAS patients on account ofmajor protocol deviations: switch fromMDI to CSII in 2 participants and preg-nancy in 1 participant.

HbA1c levels and insulin dosesThe reduction in HbA1c from baseline to12 months was significantly greater ingroup 1 (20.50% [95% CI 20.70 to20.29]) and group 2 (20.45% [20.66 to20.24]) than in group 3 (0.02% [20.18to 0.23]; group 1 vs. 3:P= 0.0006; group 2vs. 3: P= 0.0018) (Fig. 1). This result wasalso observed when the results of bothCGM groups were combined: DHbA1cgroups 1 + 2 (20.48% [20.63 to 20.33])versus group 3: P, 0.0001. The decreasein HbA1c was similar in groups 1 and 2

(group 1 vs. group 2: 0.05% [20.34 to0.25]; P= 0.7644). These results were re-ported from 3 months and continuedthroughout the study. More patients ingroups 1 and 2 had a relative HbA1c re-duction .10% versus baseline (29.0,25.5, and 11.5% in groups 1, 2, and 3,respectively; group 1 vs. group 3: P =

0.044; group 2 vs. group 3: P = 0.036),and more patients achieved the HbA1c tar-get of,7.5% (9.7, 14.6, and 1.6% ingroups 1, 2, and 3, respectively; group 1vs. group 3: P= 0.025; group 2 vs. group3: P = 0.026). In both CGM groups, pa-tients on CSII had a more pronounced re-duction in HbA1c by the end of the studythan patients on MDI compared with thecontrol group. In patients on CSII, the in-tergroup HbA1c was 20.67% (21.01 to20.33; P= 0.0001) for groups 1 + 2 (n =60) versus group 3 (n = 24); in patients onMDI, the intergroup HbA1c was only20.28% (20.67 to 0.10; NS) for groups1 + 2 (n = 56) versus group 3 (n = 37).Changes in basal or prandial insulin dosesranged from22 to 2 units/day during thestudy period, not reaching significance.

Glucose stability, hypoglycemia,and ketoacidosis

At 1 year, glycemic stability significantlyimproved in group 2 (Table 2). The fre-quency of severe hypoglycemia episodesdid not differ among groups after adjust-ment for both age and the presence of at

least one severe hypoglycemia episodedur-ing the year before inclusion. However, anonsignificant increase was observed ingroup 1, mainly with regard to one patientwho experienced seven episodes of severehypoglycemia. Ketoacidosis events wereinfrequent in all groups. In patients onCSII, the frequency of severe and mild hy-poglycemia was similar in the CGM groupscompared with the control group.

Sensor prescription and real useIn group 1, actual sensor use was 3.42 permonth (median [Q1Q3] [2.203.91]),that is, 57 6 20% of the prescribedtime. No correlation between sensor useand HbA1c level was noted in this group(P= 0.117, R2 = 0.0449). In group 2, thenumber of sensors prescribed per monthwas as follows: at baseline, three sensorsfor all patients; at 3 months, three sensorsfor 13% and four sensors for 87% of pa-tients; at 6 months, three sensors for 7%,four sensors for 56%, and five sensors for37% of patients; and at 9 months, threesensors for 2%, four sensors for 45%, fivesensors for 21%, and six sensors for 32%


Riveline and Associates


4/7

of patients. In this group, sensor use was

2.25 per month (1.272.99),that is,65629% of the prescribed time. Compliance interms of sensor use was negatively corre-lated with HbA1c level (P = 0.026, R

2 =0.1050). Total sensor consumption during1 year was significantly lower (34%) ingroup 2 than in group 1 (P= 0.001).

From baseline to the end of thestudy, SMBG per week significantly de-creased in both CGM groups versus thecontrol group (296 12 vs. 16 12, P,0.0001).

TrainingIn groups 1 and 2, 47.6% of patients hadreceived optimal training, as assessed bythe six-item questionnaire. These opti-mally trained patients exhibited greaterimprovement in HbA1c than the others.This difference remained significant afteradjustment for compliance with CGM use(DHbA1c: 20.71 6 0.81 vs. 20.30 60.81, P= 0.033).

QoLAt 1 year, a significant improvement inthe physical component summary of the

SF-36 questionnaire was reported in

groups 1 and 2 in comparison with group 3(groups 1 + 2: 1.47 6 6.52; group 3:22.48 6 6.52; P = 0.0042). The mentalcomponent summary did not differ signifi-cantly between the groups (groups 1 + 2:0.65 6 10.55; group 3: 21.03 6 10.62;NS), and neither did theglobal DQoL score.However, patient satisfaction, assessed byone scale of the DQoL, improved signifi-cantly in groups 1 and 2 (groups 1 + 2:2.83 6 12.61 vs. group 3: 22.12 612.61; P= 0.0447).

CONCLUSIONSdThe aim of thisrandomized multicenter study was tocompare the efficacy of 1 year of use ofCGM with that of standard SMBG. More-over, it assessed the performance of twoapproaches in CGM use, patient led andphysician driven. After 1 year, there was asignificant difference in HbA1c of 0.5%between the control group and the othertwo groups using CGM. Furthermore, after1 year, the decrease in HbA1c was similarin both the patient-led and physician-driven groups. The 0.5% decrease inHbA1c observed with CGM compared with

traditional SMBG was similar to that reported

in previous studies assessing the efficacy ofCGM in patients with suboptimally con-trolled type 1 diabetes (57,12). Maximumreduction in HbA1c was achieved at 3months (20.75 and 20.53% in groups 1and 2, respectively), probably as a resultof a study effect consisting offirst accessto an innovative and appealing technologyfollowed by only slight impairment in glu-cose control thereafter. However, signifi-cant improvement in glucose control wasmaintained for 1 year through CGM, a re-sult already observedin another study (13).The conditions of the only study with thesame durationreported thus far(theSensor-

Augmented Pump Therapyfor A1C Reduc-tion Study 3 [STAR-3]) were differentsince it compared the effectiveness of aninsulin pump combined with CGM versusthat of insulin injections combined withSMBG (12). Of note, the authors of thatstudy found the same degree of efficacy.It is interesting that metabolic efficacywith CGM was identical in both thepatient- and physician-driven groups,although sensor use by patients in thephysician-driven group was 34% lower.

Table 1dCharacteristics of patients with type 1 diabetes screened for inclusion in the study

FAS population (n = 178)

Screen failure

population (n = 60)Group 1 (n = 62)

Group 2

(n = 55)

Group 3

(n = 61)

All patients

(n = 178)

Male patients, n (%) 31 (50.0) 25 (45.5) 39 (63.9) 95 (53.4) 33 (55.0)

Age (years) 37.5 6 13.4 33.5 6 13.3 37.8 6 13.9 36.4 6 13.6 31.26 12.3

Age ,19, n (%) 7 (11.3) 8 (14.6) 9 (14.8) 24 (13.5) 9 (15.0)BMI (kg/m2) 24.1 6 3.9 24.7 6 3.2 25.3 6 3.6 24.76 3.6 23.8 6 4.2

Duration of diabetes (years) 16.4 6 9.1 15.4 6 8.9 18.8 6 10.6 16.96 9.6 13.2 6 8.2

Insulin regimen, n (%)

CSII 30 (48.4) 27 (49.1) 37 (60.7) 94 (52.8) 26 (43)

MDI 32 (51.6) 28 (50.9) 24 (39.3) 84 (47.2) 34 (57)

HbA1c (%) 9.0 6 0.8 8.9 6 0.9 8.8 6 0.9 9.0 6 0.9 9.3 6 1.3

Patients with $1 episode of severe

hypoglycemia, n (%)x 15 (24.2) 5 (9.1) 5 (8.2) 25 (14.0) 6 (10.0)

Patients with $1 diabetic ketoacidosis

during previous year, n (%) 1 (1.6) 1 (1.8) 0 (0.0) 2 (1.1) 4 (6.7)

Daily home glucose meter readings

(n/week) 29.2 6 15.4 25.1 6 13.7 30.0 6 14.8 28.2 6 14.7 23.86 13.0

Daily insulin doses (units/day)Total 54.7 6 26.3 53.6 6 17.2 57.0 6 23.2 55.1 6 22.6

Basal 24.4 6 11.9 25.2 6 8.9 27.0 6 11.1 25.5 6 10.7

Prandial 30.3 6 18.1 28.3 6 13.4 30.1 6 17.2 29.6 6 16.4

Educational level (%)

No diploma 27.1 37.9

College graduate 20.0 27.6

Higher education 52.9 34.5

Data are mean 6 SD unless otherwise indicated. P, 0.05 for comparison between FAS groups. P, 0.05 for comparison between FAS population and screen-failure population. xPatients with $1 episodes of severe hypoglycemia during previous year. A severe episode of hypoglycemia was defined as an event requiringassistance from another person.




5/7

In addition, compliance with CGM usewas greater in the physician-driven group,with patients adhering more closely to theprescribed time of CGM use (57 and 65%of the prescribed times in the patient- and

physician-driven groups, respectively).The correlation between CGM use andHbA1c level was significant only in thephysician-driven group. Glucose stabil-ity was also better in these patients than

in the patients from the control group, afinding that tallies with the results ob-served in several other studies (7,1315).The step-by-step approach to prescrip-tion of sensor use seems to be more eco-nomical than patient-led prescription.In the physician-driven group, the quan-tity of sensors prescribed depended on

individual metabolic outcomes at eachvisit. Finally, at the end of the study, halfof the patients in this group required CGMfor 15 to 20 days/month, with the re-mainder needing to use CGM for 25 to30 days/month. CGM requirements thusvary widely from one patient to another.

A number of previous studies suggestthat the devices must be worn as muchas possible to produce any significantbenefit (6,12). However, sensors arerelatively cumbersome, invasive, and ex-pensive. Our observations suggest thatthe initial prescription of CGM 15 days/month, followed by an incremental useif needed, may be more agreeable and,finally, cost effective.

In the patient-driven group, a non-significant trend toward higher incidenceof severe hypoglycemia events was notedin comparison with the control group.

Figure 1dHbA1c change from baseline (M0) at 3, 6, 9, and 12 months in groups 1, 2, and 3.Values aremeans (95% CIs). The data correspond well to theestimated mean and CI of the model.*P , 0.05 for comparisons between the two experimental groups and the control group at each

time point. M, month.

Table 2dGlycemic stability, hypoglycemia, and diabetic ketoacidosis

Group 1 (patient)

(n = 62)

Group 2 (physician)

(n = 55)

Group 3 (control)

(n = 61)

Pvalue

Groups

1 + 2 vs. 3

Group:*

1 vs. 3

2 vs. 3

1 vs. 2

DSD in BG (mg/dL) between

month 12 and month 0x 29.3 [219.0 to 0.4] 215.7 [226.8 to 24.6] 20.6 [28.9 to 7.7] 0.018

0.183

0.049

0.393

Patients with severe hypoglycemic

event, n (%){0 47 (75.8) 50 (90.9) 55 (90.2)

1 8 (12.9) 4 (7.3) 3 (4.9)

2 4 (6.5) 0 (0.0) 0 (0.0)

3 1 (1.6) 1 (1.8) 2 (3.3)

4 1 (1.6) 0 (0.0) 0 (0.0)

6 0 (0.0) 0 (0.0) 1 (1.6)

7 1 (1.6) 0 (0.0) 0 (0.0)

Patients with $1 event, n (%) 15 (24.2) 5 (9.19) 6 (9.8) 0.1682 0.9962

0.2153

0.9962

Patients with diabetic ketoacidosis, n (%)

0 60 (96.8) 53 (96.4) 59 (96.7)

1 1 (1.6) 2 (3.6) 1 (1.6)

2 0 (0.0) 0 (0.0) 1 (1.6)

8 1 (1.6) 0 (0.0) 0 (0.0)

*Alladjusted using the Hochberg procedure. xSDof 8-point dailybloodglucoseprofile (mg/dL), values are means (95% CIs).{A severe episode of hypoglycemia wasdefined as an event requiring assistance from another person to administer carbohydrate/glucagon.




6/7

Other randomized studies assessing theefficacy of CGM found no increase in thefrequency of severe hypoglycemic eventswith CGM, except for the STAR 1 study(16). Some studies even note an improve-ment in glycemic stability (7,13,14). Anexplanation of this difference is that onesubject in the patient-driven group pre-

sented seven events during the study.Furthermore, at baseline, patients in thisgroup had already experienced more fre-quent episodes of severe hypoglycemia.These two observations could accountfor the trend noted. Comparison of thefrequency of severe hypoglycemia be-tween groups was made after adjustmentfor baseline values, including severe hy-poglycemic events, and, consequently,there is no significant difference in se-vere hypoglycemia rates between thegroups.

Our study underscores the value of atest period during which patients areinvited to use the sensors before startinglong-term CGM to evaluate their techni-cal skills and their motivation to useCGM, as has been previously suggested(17). Patients unwilling to use the CGMdevice on a long-term basis after a 10-daytest period were younger, less compliantwith SMBG, and probably less likely touse CGM than those finally included inthe study. Thus, such a test period couldbe useful to ensure selection of patientsmotivated to use CGM.

Our study also shows that CGM ismore effective in patients treated withCSII. This result is consistent with the factthat when using an insulin pump, a pa-tient can more easily adjust insulin de-livery in real time according to both CGMvalues and trends. However, since extraboluses were not recorded during thestudy, it is difficult to determine clearlywhether patients were acting in responseto CGM data or to alarms.

CGM is also of greater benefit topatients trained to adjust their insulindoses according to CGM data,whence theimportance of structured training to helppatients translate CGM technology intoan efficient self-management tool, as sug-gested by others (18,19).

The pediatric population was com-posed of 24 subjects aged ,19 years. Be-cause of the small size of this subgroup, itwas not possible to draw any specific con-clusions about the pediatric population;however, major differences didnotshow up.

Although long-term CGM might pos-sibly increase the overallburdenof diabetesmanagement, patient QoL in this study

was not impaired after 1 year, and animprovement was in fact seen in boththe physical component summary of theSF-36 questionnaire and the satisfactioncomponent of the DQoL questionnaire. Assuggested by others, use of CGM may thuspotentially produce positive psychologicaland behavioral effects (18,20).

One limitation of our study concernsthe inability to compare glucose profilesbetween the control group and the studygroups, since masked CGM was not usedin the control group.

In summary, this study demonstratesthat use of CGM significantly improvesboth HbA1c level and glucose stability inthe long-term compared with conven-tional SMBG in patients with poorly con-trolled type 1 diabetes. This benefit isparticularly evident in patients on insulinpump therapy. In addition, patient satis-faction was improved in a study that las-ted 1 year. The uniqueness of this study isin assessing two approaches in the use ofCGM (patient led or physician driven)with step-by-step physician-drivensensor use. We report similar metabolicefficacy with both approaches. Sincethe physician-driven group used 34%fewer sensors, this strategy appearscost effective.

AcknowledgmentsdThis study was sup-ported by the Association Franaise des Dia-

btiques and the Leon Fredericq Foundationof the University of Lige (for the Belgian partof this trial).

FreeStyle Navigators, home glucose meters,and test strips used for the study were pro-vided by Abbott Diabetes Care. J.-P.R. partic-ipates in advisory boards and as a consultantfor Abbott Diabetes Care, LifeScan, sanofi-aventis, and Eli Lilly and has received hono-raria and payment for presentations, travel,and accommodation expenses from AbbottDiabetes Care and Novo Nordisk. E.R. is aconsultant for Abbott Diabetes Care and hasreceived honoraria and payment for the de-velopment of educational presentations, travel,

and accommodation expenses from AbbottDiabetes Care.A.P. isa memberof the boardsofBoehringer Ingelheim, Merck-Serono, Novartis,Novo Nordisk, and sanofi-aventis; has receivedpayment for the development of educationalpresentations including speakers office servicesfrom Abbott, Merck Sharp & Dohme, Novartis,Novo Nordisk, and sanofi-aventis; and has hadtravel and accommodation expenses covered orreimbursed by Abbott, Boehringer Ingelheim,Merck Sharp & Dohme, and sanofi-aventis.N.T.-R. is a member of a scientific advisorypanel. P.-Y.B. participates in advisory boardsand as a consultant for LifeScan, sanofi-aventis,andEli Lillyand hasreceivedfeesfor speakingin

symposia organized by Abbott Diabetes Care,Medtronic, Eli Lilly, Roche Diagnostics, andNovo Nordisk. H.H. participates in advisoryboards and as a consultant for Abbott DiabetesCare, Medtronic, sanofi-aventis, Eli Lilly, andNovo Nordisk and has received fees for speak-ing in symposia organized by Abbott DiabetesCare, Medtronic, Eli Lilly, and LifeScan. Noother potential conflicts of interest relevant to

this article were reported.Abbott Diabetes Care was not involved in

study design, data collection, or analysis.J.-P.R., G.C., and H.H. contributed to fund-

ing and to the conception and design of thestudy; acquired, analyzed, and interpreted data;and wrote, reviewed, and edited the manuscript.P.S. contributed patients with diabetes andclinical data and wrote the manuscript. L.C.,E.R., A.S.-G., A.P., N.T.-R., V.S., B.C., C.L.,R.P.R., C.T., F.M., P.-Y.B., B.G., A.-M.L., L.M.,and C.S. contributed patients with diabetesand clinical data and reviewed the manuscript.J.-P.R. is the guarantor of this work and, assuch, hadfull accessto all the data inthe study

and takes responsibility for the integrity of thedata and the accuracy of the data analysis.

Partsof this studywerepresented in abstractform at the 71st Scientific Sessions of theAmerican Diabetes Association, San Diego,California, 2428 June 2011, and at the 47thAnnual Meeting of the European Associationfor the Study of Diabetes, Portugal, Lisbon,1216 September 2011.

Study coordination, monitoring visits, datamanagement, and data analysis were performedindependently by CERITD, a nonprofit clinicaltranslational research center located in CorbeilHospital (Corbeil-Essonnes, France). The au-thors would like to thank Frederic Mistretta,MSC, and Didier Not of Randomized ClinicalTrials, Lyon, France, for their assistance withstatistical analysis, as well as Lydie Canipel andCaroline Peschard of CERITD for their involve-ment in the study management. The authorsare very grateful to the patients for their par-ticipation in the study.

References1. The Diabetes Control and Complications

Trial Research Group. The effect of in-tensive treatment of diabetes on the de-velopment and progression of long-term

complications in insulin-dependent di-abetes mellitus. N Engl J Med 1993;329:977986

2. American Diabetes Association. Standardsof medical care in diabetesd2011. Di-abetes Care 2011;34(Suppl. 1):S11S61

3. Bode BW, Schwartz S, Stubbs HA, BlockJE. Glycemic characteristics in continu-ously monitored patients with type 1 andtype 2 diabetes: normative values. Di-abetes Care 2005;28:23612366

4. Thompson CJ, Cummings JF, Chalmers J,Gould C, Newton RW. How have patientsreacted to the implications of the DCCT?Diabetes Care 1996;19:876879




7/7

5. Deiss D, Bolinder J, Riveline JP, et al. Im-proved glycemic control in poorly con-trolled patients with type 1 diabetes usingreal-time continuous glucose monitoring.Diabetes Care 2006;29:27302732

6. Tamborlane WV, Beck RW, Bode BW,et al.; Juvenile Diabetes Research Foun-dation Continuous Glucose MonitoringStudy Group. Continuous glucose moni-

toring and intensive treatment of type 1diabetes. N Engl J Med 2008;359:14641476

7. Raccah D, Sulmont V, Reznik Y, et al. In-cremental value of continuous glucosemonitoring when starting pump therapyin patients with poorly controlled type 1diabetes: the RealTrend study. DiabetesCare 2009;32:22452250

8. McGarraugh GV,Clarke WL,KovatchevBP.Comparison of the clinical informationprovided by the FreeStyle Navigatorcontinuous interstitial glucose monitorversus traditional blood glucose read-ings. Diabetes Technol Ther 2010;12:

3653719. Renard E, Apostol D, Lauton D, Boulet F,

Bringer J. Quality of life in diabetic pa-tients treated by insulin pumps. Quality ofLife Newsletter 2002;28:1113

10. Jacobson AM, de Groot M, Samson JA.The evaluation of two measures of qualityof life in patients with type I and type IIdiabetes. Diabetes Care 1994;17:267274

11. Leplge A, Ecosse E, Verdier A, PernegerTV. The French SF-36 Health Survey:translation, cultural adaptation and pre-liminary psychometric evaluation. J ClinEpidemiol 1998;51:10131023

12. Bergenstal RM, Tamborlane WV, AhmannA, et al.; STAR 3 Study Group. Effectivenessof sensor-augmented insulin-pump ther-apy in type 1 diabetes. N Engl J Med 2010;363:311320

13. Bode B, Beck RW, Xing D, et al.; JuvenileDiabetes Research Foundation ContinuousGlucose Monitoring Study Group. Sustainedbenefit of continuous glucose monitoringon A1C, glucose profiles, and hypoglyce-mia in adults with type 1 diabetes. DiabetesCare 2009;32:20472049

14. Battelino T, Phillip M, Bratina N, Nimri R,Oskarsson P, Bolinder J. Effect of continu-ous glucose monitoring on hypoglycemia

in type 1 diabetes. Diabetes Care 2011;34:795800

15. Beck RW, Hirsch IB, Laffel L, et al.; JuvenileDiabetes Research Foundation ContinuousGlucose Monitoring Study Group. The

effect of continuous glucose monitoring inwell-controlled type 1 diabetes. DiabetesCare 2009;32:13781383

16. Hirsch IB, Abelseth J, Bode BW, et al.Sensor-augmented insulin pump therapy:results of the first randomized treat-to-target study. Diabetes Technol Ther 2008;10:377383

17. Hermanides J, DeVries JH. Sense and

nonsense in sensors. Diabetologia 2010;53:593596

18. Jenkins AJ, Krishnamurthy B, Best JD,et al. An algorithmguidingpatientresponsesto real-time-continuous glucose monitoringimproves quality of life. Diabetes TechnolTher 2011;13:105109

19. Jenkins AJ, Krishnamurthy B, Best JD,et al. Evaluation of an algorithm to guidepatients with type 1 diabetes treated withcontinuoussubcutaneous insulin infusionon how to respond to real-time continuousglucose levels: a randomized controlled trial[erratum in: Diabetes Care 2010;33:1911].Diabetes Care 2010;33:12421248

20. Diabetes Research in Children Network(DirecNet) Study Group. Psychologicalaspects of continuous glucose monitoringin pediatric type 1 diabetes. Pediatr Di-abetes 2006;7:3238



basal-bolus insulin_diabetesjournal.pdf

Documents