diabetes research and clinical practice 107 (2015) 348–354
Contents available at ScienceDirect
Diabetes Research and Clinical Practice jou rnal hom ep ag e: w ww.e l s e v i er . c om/ loca te / d i ab r es
Effectiveness of continuous glucose monitoring in dialysis patients with diabetes: The DIALYDIAB pilot study Michael Joubert a,*, Corinne Fourmy a, Patrick Henri b, Maxence Ficheux b, Thierry Lobbedez b, Yves Reznik a a b
Diabetes Care Unit, University Hospital of Caen, France Dialysis Unit, University Hospital of Caen, France
article info
abstract
Article history:
Aims: The DIALYDIAB trial addresses the contribution of iterative sequences of continuous
Received 20 August 2014
glucose monitoring (CGM) on glucose control in dialysis patients with diabetes.
Received in revised form
Materials and methods: In this before–after monocentric 12-week pilot study, dialysis
13 November 2014
patients with diabetes were monitored with self-monitoring blood glucose (SMBG) 3 times
Accepted 3 January 2015
per day during a 6-week period followed by a 5-day CGM recording at 2-week interval during
Available online 21 January 2015
another 6-week period. SMBG and CGM profiles were remotely analyzed by a single diabetes
Keywords:
Results: Fifteen patients who entered the study had a male/female ratio 8/7, age 60.9 14.8
expert who gave therapeutic counseling to dialysis physicians. Diabetes
years, BMI 29.9 7.8, diabetes duration 19.2 8.5 years and dialysis duration 6.5 6.9 years.
Continuous glucose monitoring
Treatments were diet alone (20%) or diet plus insulin (80%). Mean CGM glucose level was
End-stage renal disease
8.3 2.5 mmol/l at baseline (T0), 8.2 1.6 mmol/l at the end of the SMBG period (T1) (ns) and
Hemodialysis
7.7 1.6 mmol/l at the end of the CGM period (T2) ( p < 0.05 vs T0). Glucose AUC > 10 mmol/l was 0.9 1.4 mmol/l/day at T0 and decreased to 0.4 0.5 at T2 ( p < 0.05)) without change in glucose AUC < 3.3 mmol/l. Treatment adaptation was higher during the CGM period (1.4 1.0 and 2.1 0.9 treatment change at T1 and T2, respectively; p < 0.05). Conclusions: In patients with diabetes on chronic dialysis, iterative CGM monitoring was associated with more frequent treatment changes and finally, better glucose control, without increased risk of hypoglycemia. # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Diabetes has emerged as the leading cause of end stage renal disease in many countries worldwide [1]. Patients with diabetes on chronic hemodialysis are at very high risk of
cardiovascular morbi-mortality [2]. The management of hyperglycemia is particularly challenging in this population due to changes in carbohydrate and insulin metabolism: enhanced insulin resistance, increased hepatic gluconeogenesis, impaired intracellular glucose metabolism, decreased insulin clearance
* Corresponding author at: Diabetes Care Unit, University Hospital of Caen, 14033 Caen Cedex 9, France. Tel.: +33 231064575; fax: +33 231064854. E-mail address: [email protected] (M. Joubert). http://dx.doi.org/10.1016/j.diabres.2015.01.026 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
diabetes research and clinical practice 107 (2015) 348–354
and impaired insulin secretion potentially related to metabolic acidosis [3]. Exogenous insulin and hypoglycemic agents pharmacokinetics are also altered by end stage renal disease and hemodialysis, with different profiles according to the dialysis pattern [4]. The benefit of tight glycemic control is controversial in dialysis patients with diabetes because of an increased hypoglycemic risk that may lead to adverse consequences [5–7]. Management of diabetes is also more complex because classical markers of glycemic control (i.e. HbA1c and fructosamine) may be misleading due to analytical interferences, shortened half-life of red blood cells and abnormal albumin level [8–10]. To address these questions, we conducted a study in a group of patients with diabetes on chronic hemodialysis, in whom continuous glucose monitoring (CGM) was used to evaluate their glucose profile in order to optimize diabetes treatment adjustments.
2.
Subjects, materials and methods
DIALYDIAB was a before–after monocentric 12-week pilot study in a university hospital dialysis unit, to assess the effect of iterative CGM monitoring sequences on glucose control in patients with diabetes mellitus and chronic hemodialysis. The study included two 6-week periods: during the first period, patients were asked to perform 3–6 self monitoring blood glucose (SMBG) per day with their own glucometer device (SMBG period). During the second 6-week period, a 5-day blinded CGM was performed at 2-week interval using the iPro21 CGM (Medtronic; Minneapolis; MN; USA) (CGM period). SMBG data (during the first period) and CGM profiles (during the second period) were downloaded every 2 weeks to a single diabetes expert. Considering these data and the principal clinical information transmitted by secured e-mail by the dialysis center, the diabetes expert remotely proposed antidiabetic drug changes to the nephrologist physician. Primary and secondary outcomes were evaluated at baseline (T0) and at the end of each study period (T1 for SMBG period; T2 for CGM period). Fig. 1 depicts the design of the study.
2.1.
Subjects
All patients aged 18–80 years, with diabetes mellitus (type 1, type 2 or secondary) and on chronic hemodialysis were
349
included from January 2011 to December 2012. Any combination of diabetes treatment was accepted (diet, oral hypoglycemic agents (OHA), GLP-1 analogs and/or insulin). Dialysis program was the same for all patients: a 4 h hemodialysis performed 3 times per week in a university hospital dialysis unit. Dialysis sessions were performed between 8:00 and 12:00 AM for all patients. The dialysate glucose concentration was 5 mmol/l for all the dialysis session. All subjects provided informed written consent to participate in the study, which was approved by the local ethics review board.
2.2.
Self monitoring blood glucose (SMBG)
Patients were asked to use their own glucometer, which was regularly calibrated with a control solution. During the SMBG period, 3–6 SMBG were required (at least before each meal and at the best 2 h after each meal). During the CGM period, 3 SMBG were required in order to calibrate iPro21 CGM as required by the manufacturer.
2.3.
Continuous glucose monitoring (CGM)
The iPro21 CGM (MMT-7741) with sensors MMT-7002 or MMT-7003 (Medtronic, Minneapolis, MN, USA) was used during the CGM period for a 5-day CGM recording every 2 weeks. iPro21 CGM was applied by the dialysis nurse and withdrawn 5 days later during a dialysis session. Therefore CGM recordings covered before, during and after dialysis. Patients controlled SMBG at least 3 times per day during CGM recordings for calibration. CGM data were downloaded with the CareLink Ipro1 software (MMT-7340) (Medtronic, Minneapolis, MN, USA) and then transferred to the diabetes expert for retrospective analysis of glucose profile and treatment adjustment. Additionally, a baseline 5-day blinded CGM was also performed just before the SMBG period as evaluation criterion (cf. statistical analysis). Timing of CGM during the two periods of the trial is summarized in Fig. 1. CGM criteria used for evaluation were: mean glucose (primary outcome measure), area under curve (AUC) over 10 mmol/l, AUC below 3.3 mmol/l and mean average glucose excursion (MAGE). CGM data were also used to represent the glucose profile of a model day with or without dialysis (median and Inter Quartile Range are represented for each model day).
Fig. 1 – Design of the DIALYDIAB study. Data used for diabetes treatment adaptation or for study evaluation/statistical analysis are detailed for each study period. SMBG: self monitoring blood glucose; CGM: continuous glucose monitoring; QOL: quality of life evaluated with the diabetes health profile questionnaire; Trt modif.: treatment modifications applied after suggestions of the diabetes expert. T0: baseline evaluation; T1: evaluation at the end of the SMBG period. T2: evaluation at the end of the CGM period. * CGM recording used both for treatment adaptation and evaluation at the end of the SMBG period.
350
2.4.
diabetes research and clinical practice 107 (2015) 348–354
Remote therapeutic counseling
SMBG and CGM data together with clinical relevant data were regularly downloaded, during the study period to the diabetes expert via secured The diabetes expert remotely proposed treatment changes to the nephrologist physician such as insulin dose change and/or OHA drug change adapted to the periods with or without dialysis. The nephrologist physician applied these recommendations, according to patients’ clinical features. The treatment changes were analyzed for each period as a secondary outcome.
2.5.
HbA1c
A blood sample was drawn at the beginning of a dialysis session at baseline and at the end of each period. HbA1c was measured by high-performance liquid chromatography (HPLC) with HLC-723 GHb G7 analyzer (Tosoh Bioscience, Tokyo, Japan).
2.6.
Quality of life (QOL)
Quality of life was assessed at the end of each study period using the 32-item questionnaire diabetes health profile 1 or 18 (DHP-1/18) for type 1 or type 2 diabetes, respectively. This questionnaire produces a score between 0 (bad QOL) and 100 (excellent QOL) [11].
2.7.
Statistical analysis
Patients’ characteristics were expressed as mean standard deviation and number (percentage) for quantitative and qualitative variables, respectively. The primary evaluation criterion was mean CGM glucose value at the end of each period, compared to baseline. Secondary evaluation criteria, based on CGM, were AUC over 10 mmol/l, AUC below 3.3 mmol/l and MAGE at T1 and T2 compared to T0. HbA1c was also assessed at these same times. QOL and treatment modifications at the end of each period were compared together. We used Wilcoxon signed-rank test to compare primary and secondary quantitative endpoints. Qualitative data were compared with the Fisher exact test. We used Student t-test for paired values to compare CGM glucose profiles in whole and by segments. A p-value below 0.05 was considered to denote significance.
3.
Results
Eighteen patients were included in the study but the analysis relies on 15 subjects as 3 data-sets were excluded due to incomplete CGM recording. The clinical and demographic baseline characteristics of this population were as followed: male/female ratio 8/7, age 60.9 14.8 years, BMI 29.9 7.8, diabetes duration 19.2 8.5 years, dialysis duration 6.5 6.9 years. Among the 15 patients, 2 had type 1 diabetes (13.3%), 9 had type 2 diabetes (60%) and 4 had secondary diabetes (26.7%). Secondary diabetes was due to chronic pancreatitis (n = 1) or long-term corticosteroid therapy for the initial kidney disease (n = 3). The type of kidney disease was diabetic
nephropathy in 8 (53.3%) patients, diabetic nephropathy and nephroangiosclerosis in 3 (20%) and other kidney disease in 4 (26.7%). All patients had 3 sessions of 4 h dialysis per week. Patients were on diet alone (20%) or diet plus insulin (80%) including basal insulin once per day (n = 4), premixed insulin b.i.d. (n = 1), a basal-bolus regimen (n = 4) or continuous subcutaneous insulin injection (CSII) (n = 3). Mean CGM glucose level was 8.3 2.5 mmol/l at T0, 8.2 1.6 mmol/l at T1 (ns) and 7.7 1.6 mmol/l at T2 ( p < 0.05 compared to T0). Glucose AUC > 10 mmol/l decreased significantly from 0.9 1.4 mmol/l/day at T0 to 0.4 0.5 mmol/l/day at T2 ( p < 0.05). MAGE was not different at T0, T1 and T2 (Fig. 2). Glucose AUC < 3.3 mmol/l was very low, below 0.1 mmol/l/day at T0, T1 and T2, with no statistical difference between T0 and T2. In subgroup analysis, after exclusion of patients on diet alone, only mean CGM glucose level decrease remained significant (T0: 8.8 2.5 mmol/l; T2: 8.1 1.5 mmol/l; p < 0.05; n = 12). After exclusion of patients with type 1 diabetes, only glucose AUC < 3.3 mmol/l decrease appeared significant (T0: 0.03 0.05 mmol/l/day; T2: 0.0 0.0 mmol/l/ day; p < 0.05; n = 13). HbA1c decreased from 6.85 1.48% (51 11 mmol/mol) at baseline to 6.46 1.49% (47 11 mmol/mol) at T2 ( p < 0.05) (Fig. 2). The DHP score did not change during the study periods (75.8 17.9 at T1 and 74.7 17.2 at T2). Treatment changes were more frequent during the CGM period compared to the SMBG period (2.1 0.9 vs 1.4 1.0; p < 0.05). Among treatment changes, insulin dose modifications did not differ during the two periods (1.3 1.0 compared to 1.5 1.2, respectively; ns) although treatment regimen modifications were more frequent during the CGM period compared to the SMBG period (0.6 0.6 vs 0.1 0.3; p < 0.05). For example, according to the CGM profiles, repaglinide or rapid acting insulin was added to the three patients on diet alone and to the four patients with basal insulin only. These treatment regimen changes also resulted in two different treatment patterns for days with and without dialysis in 4/15 patients. Model day CGM profiles (median and interquartile range) for day with dialysis (dialysis session between 8:00 and 12:00 AM) (n = 38 days) and day without dialysis (n = 68 days) are presented Fig. 3. Mean CGM glucose level was lower during days with dialysis compared to day without dialysis (7.6 1.0 vs 7.8 0.7 mmol/l; p < 0.05). Such difference in the mean CGM glucose level was explained by the sole 8:00–12:00 AM dialysis period change (6.8 0.3 vs 7.5 0.8 mmol/l for day with vs without dialysis; p < 0.05). Apart from the 8:00 to 12:00 AM dialysis period, no mean CGM glucose level difference was observed between days with and without dialysis (8.0 1.0 and 8.2 0.3 mmol/l, respectively; ns).
4.
Discussion
The present study suggests that the use of iterative CGM sequences in a population of patients with diabetes on chronic dialysis may result in more treatment adaptation and thus, in an improvement of glucose control without increased hypoglycemic risk. The use of CGM in this challenging population has already been validated in a previous study that compared the
351
*
9 8 7 6
2.0 1.5
*
1.0 0.5 0.0 T0
T2
c.
9
d.
8
HbA1c (%)
MAGE (mmol/l)
9
T1
5
b.
7 6 5
T2
10
2.5
T1
a.
11
Glucose AUC > 10mmol/l (mmol/l/day)
12
T0
Mean CGM glucose (mmol/l)
diabetes research and clinical practice 107 (2015) 348–354
*
8 7 6
4 3
T2
T1
T0
T2
T1
T0
5
Fig. 2 – (a) Mean CGM (continuous glucose monitoring) glucose (mmol/l), (b) glucose AUC (area under curve) >10 mmol/l (mmol/l/day), (c) MAGE (mean average glucose excursion) (mmol/l) and (d) HbA1c (%): at baseline (T0, white columns), at the end of the SMBG period (T1, hatched columns) and at the end of the CGM period (T2, black columns). Columns represent means and thin bars represent SD. *p < 0.05 compared to T0.
performance of CGM, SMBG and routine markers (i.e. HbA1c and fructosamine) for the evaluation of glucose control. The authors found a tight correlation between SMBG and CGM glucose values whereas HbA1c and fructosamine were weakly correlated to CGM glucose concentration [12]. HbA1c is biased in the situation of end stage renal disease because of analytical interferences between carbamylated hemoglobin (due to urea metabolites) and HbA1c during High Performance Liquid Chromatography (HPLC) assessment [8]. Furthermore, the shortening of life of red blood cells in chronic dialysis may also underestimate HbA1c [9]. Other markers of diabetes control are also challenged in this population. Fructosamine and glycated albumin are influenced by albumin levels and by uric acid and bilirubin, which impairs their accuracy for glucose control assessment [10,13]. Unlike classic markers, CGM provides day-to-day profile, which gives an overview of daytime glucose control. Longitudinal glucose data are of paramount importance in patients with diabetes and chronic dialysis since several studies have shown that glucose profiles may be different during days with or without dialysis session [14–17]. Kazempour et al. have showed that 24 h CGM mean glucose and 24 h AUC glucose were significantly lower on dialysis compared to off dialysis. In addition, hypoglycemia occurrence was higher during the 24 h following dialysis, with a glucose nadir occurring in a majority of patients during the 12 h post dialysis [14]. Others have found similar results, with a tighter glycemic control and an increased risk of hypoglycemia during the hours following hemodialysis. A drop in glucose variability may also be observed during the dialysis session [15]. Similar results were found during dialysis performed in non-diabetic patients, with a mean glucose level significantly lower the day after dialysis compared to the
day before [18]. In our study, glucose profiles were consistent, with a mean glucose level decreased during the day with dialysis which was due to the sole dialysis session period. This effect relates to the dialysate containing a 1 g/l glucose solution, which corrects hyperglycemic events during dialysis. In contrast, two reports found that mean glucose level and glucose variability were not different or even higher during days with dialysis. This discrepancy might be explained by the heterogeneity of patients involved in these different trials, the latter studies involving subjects with poorer glycemic control despite intensive diabetes treatment including mostly insulin [16,17]. Previous studies of glucose profile in patients on hemodialysis were descriptive. To our knowledge, this is the first CGM-based intervention study for glucose control in this population of patients. CGM was previously used in special groups of patients where tight glucose control is mandatory, i.e. pregnant women with diabetes who require achieving stringent targets to avoid macrosomia and other maternofetal issues. Murphy et al. demonstrated that 3-day CGM profile performed every 4–6 weeks in pregnant women with type 1 or 2 diabetes was superior to the classic glucose monitoring by SMBG for achieving glucose control and decreasing the risk of fetal macrosomia [19]. In the same way, we have shown in the present study that glucose control was tightened thanks to CGM use in patients with diabetes on chronic hemodialysis and that such benefit has not led to an increase of hypoglycemia. We can speculate that such benefit was obtained by iterative glucose profile assessment together with a centralized reading by an independent diabetologist. Non-structured retrospective utilization of CGM readings do not provide any benefit for reaching the glycaemic targets in
352
diabetes research and clinical practice 107 (2015) 348–354
Fig. 3 – Median and IQR (interquartile range) CGM glucose for a model day with or without dialysis. The solid line and dark gray represent median and IQR, respectively, for the model day with dialysis (which was performed between 8:00 and 12:00 AM). The dotted line and light gray represent median and IQR, respectively, for the model day without dialysis. There is no difference between the profiles, expect during hemodialysis session where mean glucose is significantly lower compared to model day without dialysis. *p < 0.05 between 8:00 and 12:00 AM (period of hemodialysis session).
patients with uncontrolled diabetes [20]. In contrast, retrospective CGM analysis by a standardized interpretation strategy, may be a valuable tool for improving glucose control. Miele et al. have reported a referral-based CGM utilization in 88 patients with diabetes and have found that a diabetes expert advice without face to face visit resulted in treatment change in 69% patients and in better glucose control in those patients with the worth baseline HbA1c, together with a reduction of nocturnal hypoglycemia [21]. Our study suffers several weaknesses. First, our sample size is limited and heterogeneous (type 1, type 2 and secondary diabetes, with various treatment, diet or different insulin regimen) and this heterogeneity weakens the results: although subgroup analysis should be cautiously interpreted in such small sample size, we show only mild benefits when excluding data from patients with type 1 diabetes or with type 2 diabetes on diet alone. This finding suggests that additional study could help to define the best patient profile that should benefit from iterative CGM. Second, the before–after design used in the study lacks statistical power and has the potential risk of a ‘‘carry-over’’ effect of the first period (SMBG) on the second period (CGM). Furthermore, the trial has been conducted in a single center. Finally, we acknowledge the lack of recommendations concerning the desirable glucose targets for CGM analysis in our specific population of patients with diabetes on chronic hemodialysis. Recent International guidelines have indicated that HbA1c should be maintained below 8% in kidney failure patients but no equivalent threshold is defined for CGM tracing [22].
However, the strengths of our study are the prospective and interventional design, the large amount of CGM data (iteratively performed) and the positive result despite the small sample size and rather low baseline HbA1c. Our successful CGM-based strategy for improving glucose control is encouraging and open a new field of practice based on a telemedicine approach in subjects with end-stage renal disease and diabetes. However, such a centralized interpretation strategy is time-consuming for the diabetes expert and therefore is probably not applicable to all centers. To address the question of expert availability, several pre-analytical software applications have been developed with CGM devices: these systems may detect glucose drifts and may alert the expert to focus on pathologic glucose profiles, in a time-saving approach. Such software has already been tested and has demonstrated, for example, the automatic detection of hypoglycemia with a sensitivity and specificity close to 100% [23]. Finally, this pilot intervention study suggests that iterative use of CGM with centralized expert diabetes counseling may improve glucose control in dialysis patients with diabetes. Larger prospective studies are needed to confirm these results.
Conflict of interest The authors declare the following potential conflicts of interest with respect to the research, authorship, and/or
diabetes research and clinical practice 107 (2015) 348–354
publication of this article: Yves Reznik and Michael Joubert carried out clinical trials as co-investigators for Medtronic and Sanofi. Yves Reznik provided advisory services to Medtronic and attended conferences organized by Sanofi and Medtronic as a contributor. Michael Joubert served as a consultant and attended conferences organized by Medtronic and Sanofi.
Statement of human and animal rights All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5).
Statement of informed consent Informed consent was obtained from all patients for being included in the study.
Acknowledgments M.J. designed the trial, performed research, performed statistical analysis and wrote the manuscript. C.F. performed research and statistical analysis. P.H. performed research and reviewed the manuscript. M.F. reviewed the manuscript. T.L. reviewed the manuscript. Y.R. contributed to discussion, and reviewed the manuscript. M.J. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors thank all investigators, study teams, and patients for participating in this study. The authors acknowledge Elisabeth Andrieu from Medtronic Society. The CGM devices and sensors were provided gratis by Medtronic (Minneapolis, MN, USA). Statistical analysis were partially funded by Sanofi France. All researchers declare that these companies did not play any role in the study design, data collection, and analysis or in the publication submission. Previously, this study was presented in abstract form at the 49th Annual EASD Meeting, Barcelona, Spain, September 23– 27, 2013.
references
[1] Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence of chronic kidney disease in the United States. JAMA 2007;298:2038–47. [2] Heras M, Fernandez-Reyes MJ, Sanchez R, Guerrero MT, Molina A, Rodriguez MA, et al. Elderly patients with chronic kidney disease: outcomes after 5 years of follow-up. Nefrology 2012;32:300–5. [3] Mak RH, DeFronzo RA. Glucose and insulin metabolism in uremia. Nephron 1992;61:377–82. [4] Haviv YS, Sharkia M, Safadi R. Hypoglycemia in patients with renal failure. Ren Fail 2000;22:219–23.
353
[5] McMurray SD, Johnson G, Davis S, McDougall K. Diabetes education and care management significantly improve patient outcomes in the dialysis unit. Am J Kidney Dis 2002;40:566–75. [6] Shurraw S, Hemmelgarn B, Lin M, Majumdar SR, Klarenbach S, Manns B, et al. Association between glycemic control and adverse outcomes in people with diabetes mellitus and chronic kidney disease: a population-based cohort study. Arch Intern Med 2011;171:1920–7. [7] Cano N. Inter-relationships between renal metabolism (both in physiology and renal dysfunction) and the liver. Curr Opin Clin Nutr Metab Care 2001;4:279–85. [8] Lee KF, Szeto YT, Benzie IFF. Glycohaemoglobin measurement: methodological differences in relation to interference by urea. Acta Diabetol 2002;39:35–9. [9] Inaba M, Okuno S, Kumeda Y, Yamada S, Imanishi Y, Tabata T, et al. Glycated albumin is a better glycemic indicator than glycated hemoglobin values in hemodialysis patients with diabetes: effect of anemia and erythropoietin injection. J Am Soc Nephrol 2007;18:896–903. [10] Joy MS, Cefalu WT, Hogan SL, Nachman PH. Long-term glycemic control measurements in diabetic patients receiving hemodialysis. Am J Kidney Dis 2002;39:297–307. [11] Meadows KA, Abrams C, Sandbaek A. Adaptation of the Diabetes Health Profile (DHP-1) for use with patients with Type 2 diabetes mellitus: psychometric evaluation and cross-cultural comparison. Diabet Med 2000;17:572–80. [12] Riveline JP, Teynie J, Belmouaz S, Franc S, Dardari D, Bauwens M, et al. Glycaemic control in type 2 diabetic patients on chronic haemodialysis: use of a continuous glucose monitoring system. Nephrol Dial Transplant 2009;24:2866–71. [13] Morgan L, Marenah C, Jeffcoate W, Morgan A. Glycated proteins as indices of glycaemic control in diabetic patients with chronic renal failure. Diabet Med 1996;13:514–9. [14] Kazempour-Ardebili S, Lecamwasam VL, Dassanyake T, Frankel AH, Tam FW, Dornhorst A, et al. Assessing glycemic control in maintenance hemodialysis patients with type 2 diabetes. Diabetes Care 2009;32:1137–42. [15] Chantrel F, Sissoko H, Kepenekian L, Smagala A, Meyer L, Imhoff O, et al. Influence of dialysis on the glucose profile in patients with diabetes: usefulness of continuous glucose monitoring. Horm Metab Res )2014;(March 13). http:// dx.doi.org/10.1055/s-0034-1370963 [Epub ahead of print]. [16] Mirani M, Berra C, Finazzi S, Calvetta A, Radaelli MG, Favareto F, et al. Inter-day glycemic variability assessed by continuous glucose monitoring in insulin-treated type 2 diabetes patients on hemodialysis. Diabetes Technol Ther 2010;12:749–53. [17] Jung HS, Kim HI, Kim MJ, Yoon JW, Ahn HY, Cho YM, et al. Analysis of hemodialysis-associated hypoglycemia in patients with type 2 diabetes using a continuous glucose monitoring system. Diabetes Technol Ther 2010;12:801–7. [18] Sobngwi E, Ashuntantang G, Ndounia E, Dehayem M, Azabji-Kenfack M, Kaze F, et al. Continuous interstitial glucose monitoring in non-diabetic subjects with end-stage renal disease undergoing maintenance haemodialysis. Diabetes Res Clin Pract 2010;90:22–5. [19] Murphy HR, Rayman G, Lewis K, Kelly S, Johal B, Duffield K, et al. Effectiveness of continuous glucose monitoring in pregnant women with diabetes: randomised clinical trial. BMJ 2008;337:a1680. [20] Pepper GM, Steinsapir J, Reynolds K. Effect of short-term iPRO continuous glucose monitoring on hemoglobin A1c levels in clinical practice. Diabetes Technol Ther 2012;14:654–7.
354
diabetes research and clinical practice 107 (2015) 348–354
[21] Miele A, Weiland K, Dungan KM. Clinical outcomes associated with referral-based continuous glucose monitoring using a central standardized interpretation strategy. Diabetes Technol Ther 2012;14:765–71. [22] Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, Ferrannini E, Nauck M, et al. Management of hyperglycemia in type 2 diabetes: a patient-centered approach: position statement of the American Diabetes
Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2012;35:1364–79. [23] Jensen MH, Christensen TF, Tarnow L, Mahmoudi Z, Johansen MD, Hejlesen OK. Professional continuous glucose monitoring in subjects with type 1 diabetes: retrospective hypoglycemia detection. J Diabetes Sci Technol 2013;7:135–43.
Contents available at ScienceDirect
Diabetes Research and Clinical Practice jou rnal hom ep ag e: w ww.e l s e v i er . c om/ loca te / d i ab r es
Effectiveness of continuous glucose monitoring in dialysis patients with diabetes: The DIALYDIAB pilot study Michael Joubert a,*, Corinne Fourmy a, Patrick Henri b, Maxence Ficheux b, Thierry Lobbedez b, Yves Reznik a a b
Diabetes Care Unit, University Hospital of Caen, France Dialysis Unit, University Hospital of Caen, France
article info
abstract
Article history:
Aims: The DIALYDIAB trial addresses the contribution of iterative sequences of continuous
Received 20 August 2014
glucose monitoring (CGM) on glucose control in dialysis patients with diabetes.
Received in revised form
Materials and methods: In this before–after monocentric 12-week pilot study, dialysis
13 November 2014
patients with diabetes were monitored with self-monitoring blood glucose (SMBG) 3 times
Accepted 3 January 2015
per day during a 6-week period followed by a 5-day CGM recording at 2-week interval during
Available online 21 January 2015
another 6-week period. SMBG and CGM profiles were remotely analyzed by a single diabetes
Keywords:
Results: Fifteen patients who entered the study had a male/female ratio 8/7, age 60.9 14.8
expert who gave therapeutic counseling to dialysis physicians. Diabetes
years, BMI 29.9 7.8, diabetes duration 19.2 8.5 years and dialysis duration 6.5 6.9 years.
Continuous glucose monitoring
Treatments were diet alone (20%) or diet plus insulin (80%). Mean CGM glucose level was
End-stage renal disease
8.3 2.5 mmol/l at baseline (T0), 8.2 1.6 mmol/l at the end of the SMBG period (T1) (ns) and
Hemodialysis
7.7 1.6 mmol/l at the end of the CGM period (T2) ( p < 0.05 vs T0). Glucose AUC > 10 mmol/l was 0.9 1.4 mmol/l/day at T0 and decreased to 0.4 0.5 at T2 ( p < 0.05)) without change in glucose AUC < 3.3 mmol/l. Treatment adaptation was higher during the CGM period (1.4 1.0 and 2.1 0.9 treatment change at T1 and T2, respectively; p < 0.05). Conclusions: In patients with diabetes on chronic dialysis, iterative CGM monitoring was associated with more frequent treatment changes and finally, better glucose control, without increased risk of hypoglycemia. # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Diabetes has emerged as the leading cause of end stage renal disease in many countries worldwide [1]. Patients with diabetes on chronic hemodialysis are at very high risk of
cardiovascular morbi-mortality [2]. The management of hyperglycemia is particularly challenging in this population due to changes in carbohydrate and insulin metabolism: enhanced insulin resistance, increased hepatic gluconeogenesis, impaired intracellular glucose metabolism, decreased insulin clearance
* Corresponding author at: Diabetes Care Unit, University Hospital of Caen, 14033 Caen Cedex 9, France. Tel.: +33 231064575; fax: +33 231064854. E-mail address: [email protected] (M. Joubert). http://dx.doi.org/10.1016/j.diabres.2015.01.026 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
diabetes research and clinical practice 107 (2015) 348–354
and impaired insulin secretion potentially related to metabolic acidosis [3]. Exogenous insulin and hypoglycemic agents pharmacokinetics are also altered by end stage renal disease and hemodialysis, with different profiles according to the dialysis pattern [4]. The benefit of tight glycemic control is controversial in dialysis patients with diabetes because of an increased hypoglycemic risk that may lead to adverse consequences [5–7]. Management of diabetes is also more complex because classical markers of glycemic control (i.e. HbA1c and fructosamine) may be misleading due to analytical interferences, shortened half-life of red blood cells and abnormal albumin level [8–10]. To address these questions, we conducted a study in a group of patients with diabetes on chronic hemodialysis, in whom continuous glucose monitoring (CGM) was used to evaluate their glucose profile in order to optimize diabetes treatment adjustments.
2.
Subjects, materials and methods
DIALYDIAB was a before–after monocentric 12-week pilot study in a university hospital dialysis unit, to assess the effect of iterative CGM monitoring sequences on glucose control in patients with diabetes mellitus and chronic hemodialysis. The study included two 6-week periods: during the first period, patients were asked to perform 3–6 self monitoring blood glucose (SMBG) per day with their own glucometer device (SMBG period). During the second 6-week period, a 5-day blinded CGM was performed at 2-week interval using the iPro21 CGM (Medtronic; Minneapolis; MN; USA) (CGM period). SMBG data (during the first period) and CGM profiles (during the second period) were downloaded every 2 weeks to a single diabetes expert. Considering these data and the principal clinical information transmitted by secured e-mail by the dialysis center, the diabetes expert remotely proposed antidiabetic drug changes to the nephrologist physician. Primary and secondary outcomes were evaluated at baseline (T0) and at the end of each study period (T1 for SMBG period; T2 for CGM period). Fig. 1 depicts the design of the study.
2.1.
Subjects
All patients aged 18–80 years, with diabetes mellitus (type 1, type 2 or secondary) and on chronic hemodialysis were
349
included from January 2011 to December 2012. Any combination of diabetes treatment was accepted (diet, oral hypoglycemic agents (OHA), GLP-1 analogs and/or insulin). Dialysis program was the same for all patients: a 4 h hemodialysis performed 3 times per week in a university hospital dialysis unit. Dialysis sessions were performed between 8:00 and 12:00 AM for all patients. The dialysate glucose concentration was 5 mmol/l for all the dialysis session. All subjects provided informed written consent to participate in the study, which was approved by the local ethics review board.
2.2.
Self monitoring blood glucose (SMBG)
Patients were asked to use their own glucometer, which was regularly calibrated with a control solution. During the SMBG period, 3–6 SMBG were required (at least before each meal and at the best 2 h after each meal). During the CGM period, 3 SMBG were required in order to calibrate iPro21 CGM as required by the manufacturer.
2.3.
Continuous glucose monitoring (CGM)
The iPro21 CGM (MMT-7741) with sensors MMT-7002 or MMT-7003 (Medtronic, Minneapolis, MN, USA) was used during the CGM period for a 5-day CGM recording every 2 weeks. iPro21 CGM was applied by the dialysis nurse and withdrawn 5 days later during a dialysis session. Therefore CGM recordings covered before, during and after dialysis. Patients controlled SMBG at least 3 times per day during CGM recordings for calibration. CGM data were downloaded with the CareLink Ipro1 software (MMT-7340) (Medtronic, Minneapolis, MN, USA) and then transferred to the diabetes expert for retrospective analysis of glucose profile and treatment adjustment. Additionally, a baseline 5-day blinded CGM was also performed just before the SMBG period as evaluation criterion (cf. statistical analysis). Timing of CGM during the two periods of the trial is summarized in Fig. 1. CGM criteria used for evaluation were: mean glucose (primary outcome measure), area under curve (AUC) over 10 mmol/l, AUC below 3.3 mmol/l and mean average glucose excursion (MAGE). CGM data were also used to represent the glucose profile of a model day with or without dialysis (median and Inter Quartile Range are represented for each model day).
Fig. 1 – Design of the DIALYDIAB study. Data used for diabetes treatment adaptation or for study evaluation/statistical analysis are detailed for each study period. SMBG: self monitoring blood glucose; CGM: continuous glucose monitoring; QOL: quality of life evaluated with the diabetes health profile questionnaire; Trt modif.: treatment modifications applied after suggestions of the diabetes expert. T0: baseline evaluation; T1: evaluation at the end of the SMBG period. T2: evaluation at the end of the CGM period. * CGM recording used both for treatment adaptation and evaluation at the end of the SMBG period.
350
2.4.
diabetes research and clinical practice 107 (2015) 348–354
Remote therapeutic counseling
SMBG and CGM data together with clinical relevant data were regularly downloaded, during the study period to the diabetes expert via secured The diabetes expert remotely proposed treatment changes to the nephrologist physician such as insulin dose change and/or OHA drug change adapted to the periods with or without dialysis. The nephrologist physician applied these recommendations, according to patients’ clinical features. The treatment changes were analyzed for each period as a secondary outcome.
2.5.
HbA1c
A blood sample was drawn at the beginning of a dialysis session at baseline and at the end of each period. HbA1c was measured by high-performance liquid chromatography (HPLC) with HLC-723 GHb G7 analyzer (Tosoh Bioscience, Tokyo, Japan).
2.6.
Quality of life (QOL)
Quality of life was assessed at the end of each study period using the 32-item questionnaire diabetes health profile 1 or 18 (DHP-1/18) for type 1 or type 2 diabetes, respectively. This questionnaire produces a score between 0 (bad QOL) and 100 (excellent QOL) [11].
2.7.
Statistical analysis
Patients’ characteristics were expressed as mean standard deviation and number (percentage) for quantitative and qualitative variables, respectively. The primary evaluation criterion was mean CGM glucose value at the end of each period, compared to baseline. Secondary evaluation criteria, based on CGM, were AUC over 10 mmol/l, AUC below 3.3 mmol/l and MAGE at T1 and T2 compared to T0. HbA1c was also assessed at these same times. QOL and treatment modifications at the end of each period were compared together. We used Wilcoxon signed-rank test to compare primary and secondary quantitative endpoints. Qualitative data were compared with the Fisher exact test. We used Student t-test for paired values to compare CGM glucose profiles in whole and by segments. A p-value below 0.05 was considered to denote significance.
3.
Results
Eighteen patients were included in the study but the analysis relies on 15 subjects as 3 data-sets were excluded due to incomplete CGM recording. The clinical and demographic baseline characteristics of this population were as followed: male/female ratio 8/7, age 60.9 14.8 years, BMI 29.9 7.8, diabetes duration 19.2 8.5 years, dialysis duration 6.5 6.9 years. Among the 15 patients, 2 had type 1 diabetes (13.3%), 9 had type 2 diabetes (60%) and 4 had secondary diabetes (26.7%). Secondary diabetes was due to chronic pancreatitis (n = 1) or long-term corticosteroid therapy for the initial kidney disease (n = 3). The type of kidney disease was diabetic
nephropathy in 8 (53.3%) patients, diabetic nephropathy and nephroangiosclerosis in 3 (20%) and other kidney disease in 4 (26.7%). All patients had 3 sessions of 4 h dialysis per week. Patients were on diet alone (20%) or diet plus insulin (80%) including basal insulin once per day (n = 4), premixed insulin b.i.d. (n = 1), a basal-bolus regimen (n = 4) or continuous subcutaneous insulin injection (CSII) (n = 3). Mean CGM glucose level was 8.3 2.5 mmol/l at T0, 8.2 1.6 mmol/l at T1 (ns) and 7.7 1.6 mmol/l at T2 ( p < 0.05 compared to T0). Glucose AUC > 10 mmol/l decreased significantly from 0.9 1.4 mmol/l/day at T0 to 0.4 0.5 mmol/l/day at T2 ( p < 0.05). MAGE was not different at T0, T1 and T2 (Fig. 2). Glucose AUC < 3.3 mmol/l was very low, below 0.1 mmol/l/day at T0, T1 and T2, with no statistical difference between T0 and T2. In subgroup analysis, after exclusion of patients on diet alone, only mean CGM glucose level decrease remained significant (T0: 8.8 2.5 mmol/l; T2: 8.1 1.5 mmol/l; p < 0.05; n = 12). After exclusion of patients with type 1 diabetes, only glucose AUC < 3.3 mmol/l decrease appeared significant (T0: 0.03 0.05 mmol/l/day; T2: 0.0 0.0 mmol/l/ day; p < 0.05; n = 13). HbA1c decreased from 6.85 1.48% (51 11 mmol/mol) at baseline to 6.46 1.49% (47 11 mmol/mol) at T2 ( p < 0.05) (Fig. 2). The DHP score did not change during the study periods (75.8 17.9 at T1 and 74.7 17.2 at T2). Treatment changes were more frequent during the CGM period compared to the SMBG period (2.1 0.9 vs 1.4 1.0; p < 0.05). Among treatment changes, insulin dose modifications did not differ during the two periods (1.3 1.0 compared to 1.5 1.2, respectively; ns) although treatment regimen modifications were more frequent during the CGM period compared to the SMBG period (0.6 0.6 vs 0.1 0.3; p < 0.05). For example, according to the CGM profiles, repaglinide or rapid acting insulin was added to the three patients on diet alone and to the four patients with basal insulin only. These treatment regimen changes also resulted in two different treatment patterns for days with and without dialysis in 4/15 patients. Model day CGM profiles (median and interquartile range) for day with dialysis (dialysis session between 8:00 and 12:00 AM) (n = 38 days) and day without dialysis (n = 68 days) are presented Fig. 3. Mean CGM glucose level was lower during days with dialysis compared to day without dialysis (7.6 1.0 vs 7.8 0.7 mmol/l; p < 0.05). Such difference in the mean CGM glucose level was explained by the sole 8:00–12:00 AM dialysis period change (6.8 0.3 vs 7.5 0.8 mmol/l for day with vs without dialysis; p < 0.05). Apart from the 8:00 to 12:00 AM dialysis period, no mean CGM glucose level difference was observed between days with and without dialysis (8.0 1.0 and 8.2 0.3 mmol/l, respectively; ns).
4.
Discussion
The present study suggests that the use of iterative CGM sequences in a population of patients with diabetes on chronic dialysis may result in more treatment adaptation and thus, in an improvement of glucose control without increased hypoglycemic risk. The use of CGM in this challenging population has already been validated in a previous study that compared the
351
*
9 8 7 6
2.0 1.5
*
1.0 0.5 0.0 T0
T2
c.
9
d.
8
HbA1c (%)
MAGE (mmol/l)
9
T1
5
b.
7 6 5
T2
10
2.5
T1
a.
11
Glucose AUC > 10mmol/l (mmol/l/day)
12
T0
Mean CGM glucose (mmol/l)
diabetes research and clinical practice 107 (2015) 348–354
*
8 7 6
4 3
T2
T1
T0
T2
T1
T0
5
Fig. 2 – (a) Mean CGM (continuous glucose monitoring) glucose (mmol/l), (b) glucose AUC (area under curve) >10 mmol/l (mmol/l/day), (c) MAGE (mean average glucose excursion) (mmol/l) and (d) HbA1c (%): at baseline (T0, white columns), at the end of the SMBG period (T1, hatched columns) and at the end of the CGM period (T2, black columns). Columns represent means and thin bars represent SD. *p < 0.05 compared to T0.
performance of CGM, SMBG and routine markers (i.e. HbA1c and fructosamine) for the evaluation of glucose control. The authors found a tight correlation between SMBG and CGM glucose values whereas HbA1c and fructosamine were weakly correlated to CGM glucose concentration [12]. HbA1c is biased in the situation of end stage renal disease because of analytical interferences between carbamylated hemoglobin (due to urea metabolites) and HbA1c during High Performance Liquid Chromatography (HPLC) assessment [8]. Furthermore, the shortening of life of red blood cells in chronic dialysis may also underestimate HbA1c [9]. Other markers of diabetes control are also challenged in this population. Fructosamine and glycated albumin are influenced by albumin levels and by uric acid and bilirubin, which impairs their accuracy for glucose control assessment [10,13]. Unlike classic markers, CGM provides day-to-day profile, which gives an overview of daytime glucose control. Longitudinal glucose data are of paramount importance in patients with diabetes and chronic dialysis since several studies have shown that glucose profiles may be different during days with or without dialysis session [14–17]. Kazempour et al. have showed that 24 h CGM mean glucose and 24 h AUC glucose were significantly lower on dialysis compared to off dialysis. In addition, hypoglycemia occurrence was higher during the 24 h following dialysis, with a glucose nadir occurring in a majority of patients during the 12 h post dialysis [14]. Others have found similar results, with a tighter glycemic control and an increased risk of hypoglycemia during the hours following hemodialysis. A drop in glucose variability may also be observed during the dialysis session [15]. Similar results were found during dialysis performed in non-diabetic patients, with a mean glucose level significantly lower the day after dialysis compared to the
day before [18]. In our study, glucose profiles were consistent, with a mean glucose level decreased during the day with dialysis which was due to the sole dialysis session period. This effect relates to the dialysate containing a 1 g/l glucose solution, which corrects hyperglycemic events during dialysis. In contrast, two reports found that mean glucose level and glucose variability were not different or even higher during days with dialysis. This discrepancy might be explained by the heterogeneity of patients involved in these different trials, the latter studies involving subjects with poorer glycemic control despite intensive diabetes treatment including mostly insulin [16,17]. Previous studies of glucose profile in patients on hemodialysis were descriptive. To our knowledge, this is the first CGM-based intervention study for glucose control in this population of patients. CGM was previously used in special groups of patients where tight glucose control is mandatory, i.e. pregnant women with diabetes who require achieving stringent targets to avoid macrosomia and other maternofetal issues. Murphy et al. demonstrated that 3-day CGM profile performed every 4–6 weeks in pregnant women with type 1 or 2 diabetes was superior to the classic glucose monitoring by SMBG for achieving glucose control and decreasing the risk of fetal macrosomia [19]. In the same way, we have shown in the present study that glucose control was tightened thanks to CGM use in patients with diabetes on chronic hemodialysis and that such benefit has not led to an increase of hypoglycemia. We can speculate that such benefit was obtained by iterative glucose profile assessment together with a centralized reading by an independent diabetologist. Non-structured retrospective utilization of CGM readings do not provide any benefit for reaching the glycaemic targets in
352
diabetes research and clinical practice 107 (2015) 348–354
Fig. 3 – Median and IQR (interquartile range) CGM glucose for a model day with or without dialysis. The solid line and dark gray represent median and IQR, respectively, for the model day with dialysis (which was performed between 8:00 and 12:00 AM). The dotted line and light gray represent median and IQR, respectively, for the model day without dialysis. There is no difference between the profiles, expect during hemodialysis session where mean glucose is significantly lower compared to model day without dialysis. *p < 0.05 between 8:00 and 12:00 AM (period of hemodialysis session).
patients with uncontrolled diabetes [20]. In contrast, retrospective CGM analysis by a standardized interpretation strategy, may be a valuable tool for improving glucose control. Miele et al. have reported a referral-based CGM utilization in 88 patients with diabetes and have found that a diabetes expert advice without face to face visit resulted in treatment change in 69% patients and in better glucose control in those patients with the worth baseline HbA1c, together with a reduction of nocturnal hypoglycemia [21]. Our study suffers several weaknesses. First, our sample size is limited and heterogeneous (type 1, type 2 and secondary diabetes, with various treatment, diet or different insulin regimen) and this heterogeneity weakens the results: although subgroup analysis should be cautiously interpreted in such small sample size, we show only mild benefits when excluding data from patients with type 1 diabetes or with type 2 diabetes on diet alone. This finding suggests that additional study could help to define the best patient profile that should benefit from iterative CGM. Second, the before–after design used in the study lacks statistical power and has the potential risk of a ‘‘carry-over’’ effect of the first period (SMBG) on the second period (CGM). Furthermore, the trial has been conducted in a single center. Finally, we acknowledge the lack of recommendations concerning the desirable glucose targets for CGM analysis in our specific population of patients with diabetes on chronic hemodialysis. Recent International guidelines have indicated that HbA1c should be maintained below 8% in kidney failure patients but no equivalent threshold is defined for CGM tracing [22].
However, the strengths of our study are the prospective and interventional design, the large amount of CGM data (iteratively performed) and the positive result despite the small sample size and rather low baseline HbA1c. Our successful CGM-based strategy for improving glucose control is encouraging and open a new field of practice based on a telemedicine approach in subjects with end-stage renal disease and diabetes. However, such a centralized interpretation strategy is time-consuming for the diabetes expert and therefore is probably not applicable to all centers. To address the question of expert availability, several pre-analytical software applications have been developed with CGM devices: these systems may detect glucose drifts and may alert the expert to focus on pathologic glucose profiles, in a time-saving approach. Such software has already been tested and has demonstrated, for example, the automatic detection of hypoglycemia with a sensitivity and specificity close to 100% [23]. Finally, this pilot intervention study suggests that iterative use of CGM with centralized expert diabetes counseling may improve glucose control in dialysis patients with diabetes. Larger prospective studies are needed to confirm these results.
Conflict of interest The authors declare the following potential conflicts of interest with respect to the research, authorship, and/or
diabetes research and clinical practice 107 (2015) 348–354
publication of this article: Yves Reznik and Michael Joubert carried out clinical trials as co-investigators for Medtronic and Sanofi. Yves Reznik provided advisory services to Medtronic and attended conferences organized by Sanofi and Medtronic as a contributor. Michael Joubert served as a consultant and attended conferences organized by Medtronic and Sanofi.
Statement of human and animal rights All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5).
Statement of informed consent Informed consent was obtained from all patients for being included in the study.
Acknowledgments M.J. designed the trial, performed research, performed statistical analysis and wrote the manuscript. C.F. performed research and statistical analysis. P.H. performed research and reviewed the manuscript. M.F. reviewed the manuscript. T.L. reviewed the manuscript. Y.R. contributed to discussion, and reviewed the manuscript. M.J. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The authors thank all investigators, study teams, and patients for participating in this study. The authors acknowledge Elisabeth Andrieu from Medtronic Society. The CGM devices and sensors were provided gratis by Medtronic (Minneapolis, MN, USA). Statistical analysis were partially funded by Sanofi France. All researchers declare that these companies did not play any role in the study design, data collection, and analysis or in the publication submission. Previously, this study was presented in abstract form at the 49th Annual EASD Meeting, Barcelona, Spain, September 23– 27, 2013.
references
[1] Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, et al. Prevalence of chronic kidney disease in the United States. JAMA 2007;298:2038–47. [2] Heras M, Fernandez-Reyes MJ, Sanchez R, Guerrero MT, Molina A, Rodriguez MA, et al. Elderly patients with chronic kidney disease: outcomes after 5 years of follow-up. Nefrology 2012;32:300–5. [3] Mak RH, DeFronzo RA. Glucose and insulin metabolism in uremia. Nephron 1992;61:377–82. [4] Haviv YS, Sharkia M, Safadi R. Hypoglycemia in patients with renal failure. Ren Fail 2000;22:219–23.
353
[5] McMurray SD, Johnson G, Davis S, McDougall K. Diabetes education and care management significantly improve patient outcomes in the dialysis unit. Am J Kidney Dis 2002;40:566–75. [6] Shurraw S, Hemmelgarn B, Lin M, Majumdar SR, Klarenbach S, Manns B, et al. Association between glycemic control and adverse outcomes in people with diabetes mellitus and chronic kidney disease: a population-based cohort study. Arch Intern Med 2011;171:1920–7. [7] Cano N. Inter-relationships between renal metabolism (both in physiology and renal dysfunction) and the liver. Curr Opin Clin Nutr Metab Care 2001;4:279–85. [8] Lee KF, Szeto YT, Benzie IFF. Glycohaemoglobin measurement: methodological differences in relation to interference by urea. Acta Diabetol 2002;39:35–9. [9] Inaba M, Okuno S, Kumeda Y, Yamada S, Imanishi Y, Tabata T, et al. Glycated albumin is a better glycemic indicator than glycated hemoglobin values in hemodialysis patients with diabetes: effect of anemia and erythropoietin injection. J Am Soc Nephrol 2007;18:896–903. [10] Joy MS, Cefalu WT, Hogan SL, Nachman PH. Long-term glycemic control measurements in diabetic patients receiving hemodialysis. Am J Kidney Dis 2002;39:297–307. [11] Meadows KA, Abrams C, Sandbaek A. Adaptation of the Diabetes Health Profile (DHP-1) for use with patients with Type 2 diabetes mellitus: psychometric evaluation and cross-cultural comparison. Diabet Med 2000;17:572–80. [12] Riveline JP, Teynie J, Belmouaz S, Franc S, Dardari D, Bauwens M, et al. Glycaemic control in type 2 diabetic patients on chronic haemodialysis: use of a continuous glucose monitoring system. Nephrol Dial Transplant 2009;24:2866–71. [13] Morgan L, Marenah C, Jeffcoate W, Morgan A. Glycated proteins as indices of glycaemic control in diabetic patients with chronic renal failure. Diabet Med 1996;13:514–9. [14] Kazempour-Ardebili S, Lecamwasam VL, Dassanyake T, Frankel AH, Tam FW, Dornhorst A, et al. Assessing glycemic control in maintenance hemodialysis patients with type 2 diabetes. Diabetes Care 2009;32:1137–42. [15] Chantrel F, Sissoko H, Kepenekian L, Smagala A, Meyer L, Imhoff O, et al. Influence of dialysis on the glucose profile in patients with diabetes: usefulness of continuous glucose monitoring. Horm Metab Res )2014;(March 13). http:// dx.doi.org/10.1055/s-0034-1370963 [Epub ahead of print]. [16] Mirani M, Berra C, Finazzi S, Calvetta A, Radaelli MG, Favareto F, et al. Inter-day glycemic variability assessed by continuous glucose monitoring in insulin-treated type 2 diabetes patients on hemodialysis. Diabetes Technol Ther 2010;12:749–53. [17] Jung HS, Kim HI, Kim MJ, Yoon JW, Ahn HY, Cho YM, et al. Analysis of hemodialysis-associated hypoglycemia in patients with type 2 diabetes using a continuous glucose monitoring system. Diabetes Technol Ther 2010;12:801–7. [18] Sobngwi E, Ashuntantang G, Ndounia E, Dehayem M, Azabji-Kenfack M, Kaze F, et al. Continuous interstitial glucose monitoring in non-diabetic subjects with end-stage renal disease undergoing maintenance haemodialysis. Diabetes Res Clin Pract 2010;90:22–5. [19] Murphy HR, Rayman G, Lewis K, Kelly S, Johal B, Duffield K, et al. Effectiveness of continuous glucose monitoring in pregnant women with diabetes: randomised clinical trial. BMJ 2008;337:a1680. [20] Pepper GM, Steinsapir J, Reynolds K. Effect of short-term iPRO continuous glucose monitoring on hemoglobin A1c levels in clinical practice. Diabetes Technol Ther 2012;14:654–7.
354
diabetes research and clinical practice 107 (2015) 348–354
[21] Miele A, Weiland K, Dungan KM. Clinical outcomes associated with referral-based continuous glucose monitoring using a central standardized interpretation strategy. Diabetes Technol Ther 2012;14:765–71. [22] Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, Ferrannini E, Nauck M, et al. Management of hyperglycemia in type 2 diabetes: a patient-centered approach: position statement of the American Diabetes
Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2012;35:1364–79. [23] Jensen MH, Christensen TF, Tarnow L, Mahmoudi Z, Johansen MD, Hejlesen OK. Professional continuous glucose monitoring in subjects with type 1 diabetes: retrospective hypoglycemia detection. J Diabetes Sci Technol 2013;7:135–43.
