http://orcid.org/0000-0003-4605-2503
Figures
Abstract
Background
T1D treatment requires informed self-responsible patients, who, however, frequently miss their therapeutic goals, providing considerable potential for improvement.
Methods
This observational report evaluates T1D patients [N = 109], aged ≥18 years (range 22–82), poorly controlled at home, at and 3 weeks after their admission to our diabetes rehabilitation clinic [DRC], where they were offered standardized, but unmonitored life-style modification.
Results
At admission, patients displayed elevated HbA1c values (66 mmol/mol [57; 81]), a high prevalence of co-morbidities (88%), lipodystrophies due to monolocal insulin injections (42%), a low rate of influenza (16%) and pneumococcal (7%) immunization, and underuse of lipid-lowering drugs (-38%). Standardization of life-style improved glucose (p<0.0001) and lipid metabolism (LDL/HDL ratio p<0.01) permitting reduction of insulin dose and reduction of add-on glucose-lowering drugs (GLDs) other than metformin. Outcome was independent of the mode of insulin treatment strategy and more marked at initially high HbA1c, with DRC-costs/d less than 25% of those encountered at standard hospitals.
Citation: Haslacher H, Fallmann H, Waldhäusl C, Hartmann E, Wagner OF, Waldhäusl W (2018) Type 1 diabetes care: Improvement by standardization in a diabetes rehabilitation clinic. An observational report. PLoS ONE 13(3): e0194135. https://doi.org/10.1371/journal.pone.0194135
Editor: Esayas Kebede Gudina, Jimma University, ETHIOPIA
Received: August 22, 2017; Accepted: February 11, 2018; Published: March 12, 2018
Copyright: © 2018 Haslacher et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Raw data are available from Researchgate. DOI: 10.13140/RG.2.2.11836.95364.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
In the absence of a cure, insulin replacement is mandatory for survival in patients with type 1 diabetes mellitus and requires them to be thoroughly informed and trained. They not only have to learn how to replace insulin properly [1], but also to self-administer treatment reliably, as they need to care for themselves alone between visits to their doctor commonly at intervals of 6 to 12 weeks. To shoulder the task of being ‘his own nurse, doctor’s assistant, chemist’ [2] and even physician, and to meet the metabolic goals set [3, 4], patients have to continuously maintain motivation for multiple daily interventions. These include insulin injections and blood glucose [BG] self-measurements to keep their BG within the desired range, i.e. HbA1c below 53 to 58 mmol/mol, if insulin reactions [4] and diabetes-related complications are to be avoided [5]. To provide comprehensive self-care type 1 diabetes patients are also required to document BG values, food intake, administered insulin doses, and bouts of exercise [6, 7], as well as blood pressure (BP), which taken together, may well overstretch individual motivation to comply.
The clinical outcome of this therapeutic approach is variable at best, as type 1 diabetes patients’ HbA1c values, a surrogate marker of long-term BG concentration, fluctuate widely. Thus, set goals are met by a minority only (18%), while 37% even display values >64 mmol/mol (67±18 mmol/mol) [8], which is similar to the 72±23 mmol/mol seen in type 1 diabetes patients 30 years ago in rural area [9] and potentially worsening over time [10]. The associated annual economic costs range in central Europe from € 5,331 (Switzerland 2014) to € 5,899 (Germany 2010) per diabetic patient [11, 12]. The development of novel, cost-effective treatment strategies, as e.g. the integration of lifestyle interventions, is therefore highly warranted. However, only a few studies are available reporting the impact of life style on required medication other than insulin and metabolic parameters in type 1 diabetes patients. This is associated with the odd situation that most standard recommendations on physical activity for T1D patients are based on data from healthy subjects or from patients with T2D [13].
If patient-centered self-care is to be more than rhetoric, its outcome consequently has to be routinely measured by providers of care [14, 15] to identify its potential for metabolic improvement and reduction of costs. This observational study explores besides associated costs and risk factors (smoking, depression, compliance with influenza and pneumococcal immunization), vital variables, metabolic outcome, and medications in type 1 diabetes patients at time of their admission vs. that at discharge after a 3-week stay at our Diabetes Rehabilitation Clinic (DRC).
Methods
Study patients
This single-center, explorative study determines improvement in outcome of diabetes care in consecutively admitted type 1 diabetes patients (N = 109), aged ≥18 years (range 22–82), who did not meet their therapeutic goals at home, by identifying changes in selected vital and metabolic variables in response to a 3-week stay at our DRC vs. those observed at admission. The study was approved by the Ethics Committee of the Medical University of Vienna (#1527/2016), and all patients included provided written, informed consent to participate.
Type 1 diabetic patients were referred to the DRC via the Austrian insurance system because of inadequate metabolic control at home, where they were attended either by general practitioners [GP] or by diabetes outpatient services [DOS]. Inclusion in the study required the referring diagnosis to be supported by at least two of the following indicators: i) age at diagnosis <40 years, ii) body mass index [kg/m2] <30, or a description as being a slender or non-obese adult at that time, iii) insulin treatment from onset of disease or serum C-peptide <0.8 ng/ml with blood glucose >4.4 mmol/l at any time, and/or iv) positive GAD-65 antibodies at any time after onset of hyperglycaemia [8], which reduced the number of type 1 diabetes patients included by six to 109.
Treatment
Patients were offered a structured refresher course on diabetes self-care tailored to their individual treatment mode [1] and medical counseling, while their adherence to treatment recommendations remained unmonitored and was only subject to control by local peer pressure.
Insulin treatment strategies. Conventional insulin therapy (CIT) is defined as predetermined dosing and timing of either premixed insulin or long- and short-acting insulin injections accompanied by agreed timing of a defined food and exercise load. Functional insulin therapy (FIT; basis/bolus; physiologic treatment) is defined as disconnected administration of basal (long-acting) and prandial (short-acting) insulin, with individual dosing of the latter as deemed appropriate by the educated patient to compensate for meals or hyperglycaemia, ensuring maximum patient autonomy [1, 5]. Any mix of components of CIT and FIT is termed Intensified insulin treatment (IIT). Both FIT and IIT were administered either by multiple insulin injections or with an insulin pump (CSII, Continuous subcutaneous insulin infusion).
Intake of non-insulin glucose lowering drugs (GLDs) is documented as number of tablets/doses of metformin, DPP4 inhibitors, metformin combined with DPP4 inhibitors, SGLT-2 inhibitors, or GLP-1 receptor agonists administered per day. Additional medication is documented as number of tablets ingested per day for lipid lowering drugs (statins), antihypertensive (ACE inhibitors, angiotensin II receptor blockers [ARBs], diuretics, calcium antagonists, beta-blockers, or alpha-blockers) and antidepressant drugs, and for any other medication.
Life-style. Throughout their stay at our DRC, type 1 diabetes patients were exposed to a standardized, but unmonitored lifestyle offering mixed food, rich in vegetables and fruits providing three meals totaling 1,200 to 2,000 kcal/d depending on BMI, plus an exercise load, such as hiking, swimming and gymnastics, equivalent to an energy expenditure of 400 to 600 kcal/d.
Examinations
Medical history documented standard information, co-morbidities, frequency of past hypoglycaemic events (symptomatic or incidental BG <2.8 mmol/l; N/week), smoking habits and compliance with recommendations for influenza and pneumococcal immunization [16].
Physical examination documented body weight and height, BMI (body mass index, weight [kg]/height [m2]), waist circumference [cm] and blood pressure (BP [mmHg]) as well as lipodystrophies and/or infiltrations at insulin injection sites. Diabetic neuropathy was rated using a neuropathy symptom (NSS) and deficit (NDS) score ranging from 0 (normal) to 10 (severe) [17].
Laboratory tests
Laboratory analyses were performed at the MVZ für Laboratoriumsmedizin, Raubling GmbH, Germany using ISO 15189 accredited standardized procedures and are presented in SI units. Glomerular filtration rate (GFR, ml/min) was estimated using the CKD-EPI equation [18]. HbA1c measurements ([mmol/mol]; HPLC, Bio-Rad Variant II; Bio-Rad Laboratories Inc., Hercules, USA) were made at intervals of three weeks as its changes, though smaller, can already be detected as early as two weeks after an intervention [19].
Statistical analyses
Continuous data are given as means ± SD or median (interquartile range, IQR), categorical data as counts and percentages. Comparisons of paired continuous data were made using the non-parametric Wilcoxon test, or, where appropriate, Student's t-test. Correlations between continuous data were calculated according to Pearson (linear) and Spearman (non-linear). Independent data were compared by ANOVA and post-hoc t-tests and categorical data by Pearson’s χ2 tests, whereas deviations of dichotomous variables from uniform distributions were assessed by binomial tests.
Outcome was evaluated by interpreting main effects and interactions derived from general linear models with repeated measurement design (SPSS [IBM, Armonk, NY, USA]).
Analytically relevant differences between HbA1c concentrations at baseline and discharge are defined as changes exceeding the reference change value [20]:
Given a Z (Z-score) of 1.96 for significance at 95% probability level, a CVA (analytical coefficient of variance) of 1.05% and a CVI (individual biological variation) of 1.9% [21], a relative difference >6% in HbA1c concentration between baseline and discharge was considered diagnostically relevant.
Possible decreases in HbA1c were predicted by binary logistic regression models, whereas goodness of fit was evaluated by interpreting areas under the curve (AUC) of receiver operator characteristic (ROC) plots. p-Values, recalculated according to Benjamini and Hochberg, were considered significant if <0.05.
Improvements of HbA1c, LDL, and mean arterial blood pressure (MAP, diastolic BP plus BP-amplitude/3) were determined after three weeks as reduction at discharge vs. values at admission, and expressed as percentage of their respective initial deviation from ADA benchmarks (HbA1c, 53 mmol/mol; LDL, 1.81 mmol/mol and MAP, <107 mmHg) [3].
Figures were drawn using GraphPad Prism 6 (GraphPad Software Inc., La Jolla, Ca, USA), MedCalc version 15.8 (MedCalc Software bvba, Ostend, Belgium), or SPSS 23 (IBM), which was also used for calculations.
Results
Study population
Baseline characteristics of type 1 diabetes patients are presented in Tables 1 and 2. In brief, we registered a preponderance of males (54%), with a median duration of disease of 18 years and a high frequency of lipodystrophies and infiltrations at monolocal insulin injection sites (42%). Insulin demand was 15 percent greater at admission (0.53 [0.44; 0.65] U/kg b.w. [body weight]) than at the time of discharge (0.46 [0.39; 0.53] U/kg b.w.; p<0.0001), when infiltrated areas were avoided. Hypoglycaemia was encountered once/week and its rate not different between sexes, while nephropathy (GFR < 60 ml/min) was seen in 9 per cent of patients.
The combined share of active (N = 40) and former smokers (N = 15) among type 1 diabetes patients was 52%, smoking a median of 20 (10; 20) cigarettes per day over 25 (17; 31) years, and did not differ between sexes. The decision of patients to stop smoking was commonly coincident with a major health hazard (stroke, myocardial infarction etc.), but never a free personal decision. Non-smokers had a higher BMI (+2.8±0.9 kg/m2, p<0.05), developing, similar to their waist circumference (-2.3±0.3 vs. -0.8±0.4 cm, p<0.05), more favorably at the DRC (-0.4±0.1 kg/m2) than in current smokers (0.0±0.1 kg/m2, p<0.05), who also had higher HbA1c values (+8±3 mmol/mol, p<0.05).
Recommendations of influenza (16%) and pneumococcal (7%) immunization were only rarely implemented by type 1 diabetes patients.
Co-morbidities (Fig 1A) were diagnosed in 88% of type 1 diabetes patients and dominated by hyperlipidaemia (52%) followed by arterial hypertension, depression, cardiovascular disease, hypothyroidism/Schmidt syndrome (8%), chronic obstructive pulmonary disease (COPD), and other disorders with a prevalence of <6% including alcoholism, carcinoma, psychosis, sleep apnea, and others.
(a) Co-morbidities (N,%) in type 1 diabetes patients (N = 109). CHD, coronary heart disease; PAD, peripheral artery disease; HT, hypothyreoidism; COPD, chronic obstructive pulmonary disease, and (b) correlation of NDS (neuropathic deficit score) with duration of disease (ρ = 0.368, p<.01).
Arterial hypertension was strongly associated with older age (52 (46; 57) vs. 46 (36; 54) years; p<0.01), increased waist circumference (+8.3±2.4 cm, p<0.01), and by trend with higher BMI (+2.0±0.9 kg/m2) and triglycerides (+0.29±0.13 mmol/l; both p = 0.057).
Interestingly, presence of depression neither depended on age, duration of illness, insulin dose, or lipodystrophies.
Sixteen percent of type 1 diabetes patients suffered, dependent on duration of disease (Fig 1B; ρ = 0.368, p<0.01), from diabetic neuropathy with the neuropathic deficit score (NDS, 0 [0; 3]) being somewhat smaller than the corresponding symptom score (NSS, 2 [0; 4]), demonstrating the subjective burden of complaints.
Outcome
Vital and metabolic variables (Table 2A).
Analyzing clinical outcome after three weeks, significant improvement was seen in both vital (BMI, waist circumference, BP) and metabolic variables.
At admission, 86% of type 1 diabetes patients had HbA1c values above the ADA benchmark (>53 mmol/mol) and only 14% at or below the therapeutic target, with this share increasing to 17% at discharge. During the DRC stay, HbA1c did not change in eight patients (8%) and even deteriorated in 17%. Of the 75% with decreasing HbA1c values, 41% showed an analytically relevant change >6% from baseline, displaying an inverse linear relationship with HbA1c values recorded at admission (r = -0.621, p<0.0001; Fig 2A).
(a) Inverse correlation between HbA1c values at admission and their relative changes in response to a 3-week stay at the DRC in type 1 diabetes patients (N = 109). Insert: ROC-curve from a binary logistic regression model predicting from baseline possible HbA1c improvement (>6.0 relative %) in response to proper treatment. (b) Strategies of insulin treatment used by type 1 diabetes patients (N = 109) at admission and at discharge. Note the shift towards more elaborate treatment modes. BOT, Basal supported oral therapy; CIT, Conventional insulin therapy; IIT (± CSII), Intensified insulin therapy ± continuous subcutaneous insulin infusion, FIT (± CSII), Functional insulin (basis/bolus) therapy ± continuous subcutaneous insulin infusion.
Notably, the fall in serum LDL/HDL ratio did not depend on medication alone, but also on changes in lifestyle as its decline was observed in patients with pre-established (-17±31%, p<0.01) and newly established (-34±21%, p<0.0001) hypolipidaemic treatment (interaction for all three groups p<0.01). A similar pattern applied to MAP, which decreased within three weeks to the same extent in patients pretreated, newly treated and those not treated with antihypertensives (interaction p>0.05).
Of note, changes in both metabolic and vital parameters were independent from the patients’ choice of insulin treatment strategy, emphasizing the need to reliably implement pertinent treatment recommendations (see Table A and Fig A in S1 File).
Significantly, patients with initially elevated HbA1c and LDL levels reduced this deviation within 3 weeks by more than thirty percent towards their respective ADA benchmark, while MAP fell even more markedly (Fig 2C).
When HbA1c concentration at admission was categorized by residential postal code, some regions displayed better HbA1c values (p<0.05) than others (H, 58 mmol/mol [53; 65]; <B, Δ+15±6 mmol/mol; <C, Δ+17±6 mmol/mol; <<F, Δ+21±7 mmol/mol; and <<<G, Δ+22±9 mmol/mol). In particular, G showed considerably poorer care than H, reflecting regional differences in diabetes education and care.
Type 1 diabetes patients were grouped according to visits to their general practitioner (GP, 42%) or use of a specialized diabetes outpatient service (DOS, 58%) at home. As depicted in Fig 3, those seen by the former benefitted most from their stay at the DRC (HbA1c: GP -7±4%, p<0.0001; DOS -3±6%, p<0.01; interaction: p<0.05. LDL/HDL ratio: GP -25±23%, p<0.001; DOS -4±37%, n.s.; interaction: p<0.05).
Means ± SD.
To identify type 1 diabetes patients who might benefit most from a standardized environment we investigated potential predictors of metabolic improvement such as age, sex, BMI, waist circumference, HbA1c, fasting BG, total daily insulin dose, LDL/HDL ratio, and duration of disease. Of those, only baseline HbA1c proved to be an independent variable predicting potential relative HbA1c reductions ≥6% from baseline, exhibiting an odds ratio of 3.119 (2.011–4.838). The resulting model presented with considerable goodness of fit (Model: χ2 = 42.361, df = 1, p<0.0001, Nagelkerkes R2 = 0.434) and a good predictive capacity upon ROC analysis (AUC 0.841 (0.758–0.904)) with a maximal Youden’s index of 53% predicted probability (Fig 2A insert).
Medications (Table 2B).
Insulin. Median total insulin dose at admission was 40 U/d (32; 52), of which 22 (17; 30) U/d were long-acting and 18 (14; 23) U/d short-acting insulin. Although total daily insulin dose was to some extent (ρ = 0.202, p>0.05) associated with duration of disease, this correlation was nonlinear, with its first quartile (≤7 years) receiving less insulin at admission than the other quartiles (36 (32; 39) p<0.05).
To avoid hypoglycaemia due to better compliance with treatment rules at the DRC, total daily insulin dose had to be reduced by 13% from 0.53 (0.44; 0.65) to 0.46 (0.39; 0.53; p Type 1 diabetes patients with cutaneous infiltrations due to monolocal insulin injections (49%) required somewhat higher insulin doses at admission (+19%) than those without lipodystrophy (0.61 (0.47; 0.70) vs. 0.51 (0.42; 0.56) U/kg b.w., p<0.05).
At admission, most type 1 diabetes patients used some kind of intensified insulin therapy (N = 84, 77%), while only 10% employed Functional (basis/bolus) insulin therapy (FIT), 12% used Conventional insulin therapy (CIT), and only one chose basal insulin alone. Following counseling, the number of patients opting for CIT fell to three, while the majority preferred more elaborate strategies (basal/bolus 55%, IIT 42%), while the use of Continuous subcutaneous insulin infusion (CSII) rose from 17 to 26% (p = 0.002, Fig 2B).
Other medications. At admission, 85 (78%) patients received at least one additional medica-tion besides insulin. This proportion was unchanged at the time of discharge (N = 86 [79%]).
Although the number of patients receiving GLDs remained constant, add-on antidiabetic treatment shifted towards metformin, in part to avoid overinsulinzation and weight gain, replacing sulfonylureas, glitazones and SGLT-2 inhibitors (Table 2B). Such superfluous medication of GLDs other than metformin was more prevalent in type 1 diabetes patients previously attended at home by GPs than in those taken care of by specialized DOS, which also achieved better metabolic control.
The use of statins had to be increased by 38%, while that of antihypertensives remained constant for both ACE inhibitors (18%) and ARBs (22%). Their supplement with diuretics (11%), however, was replaced at discharge in part (40%) by calcium antagonists, whereas the rate of beta-blocker medication remained constant, and the use of antidepressants and other medications did not change.
Costs
Costs per day at the DRC including board, lodging, physical rehabilitation services plus expenses for supplementary and/or modification of medication as used are modest (€131.—per day) vs. those in standard (acute) hospitals ranging from 594 to 2042 €/d (see Table B in S1 File). Of note, due to the reduction in daily insulin dose, savings, calculated as pharmacy prices, were offset by an increased need for lipid-lowering drugs, while costs for any other medication remained constant.
Discussion
This exploratory study shows remarkable improvement of metabolic outcome if diabetes self-care is supported by standardized life-style modification, peer pressure and mutual interaction between patients at a DRC (Table 2, Fig 2A). This suggests that the outcome gap in diabetes care, i.e. the gap between diabetes control at admission and benchmark values, could possibly be narrowed or even closed at home by simple means, if type 1 diabetes patients internalized that concerted action as to insulin dosing, food intake, and exercise is requisite to therapeutic success and not merely optional.
The need to comply with respective recommendations for treatment and lifestyle is also demonstrated by the identical rate of improvement seen after three weeks at the DRC in vital (BMI) and metabolic variables (HbA1c, LDL/HDL ratio) independent of the applied strategy of insulin treatment.
Poor metabolic control may, however, also be due to mood swings associated with nicotine abuse [22], or to the patients’ desire to avoid hypoglycaemia as seen in males, simply by increasing carbohydrate load or reducing insulin dose [23]. Moreover, recommendations for T1D-patients as to exercise and nutrition are often drawn from type 2 diabetes-patients [24], but not evaluated as such. In this context, it is of no surprise if metaanalyses of the effect of physical exercise in addition to insulin treatment on long-term glycaemic control provide inconclusive results [25], or report only some small metabolic improvement in Type 1 diabetic children and adolescents [26–28].
Our report shows that inadequate compliance with standards of type 1 diabetes care is also commonplace with regard to i) smoking, ii) attention to co-morbidities, which ought to be treated properly as previously shown in type 2 diabetes patients [29], and iii) influenza and pneumococcal immunization [30], whose importance is frequently dismissed by both type 1 diabetes patients and by their attending physicians [31].
Any such disregard of treatment recommendations is particularly detrimental in type 1 diabetes patients, as they suffer from multiple co-morbidities. In our cohort, those were dominated by hyperlipidaemia, hypertension and obesity, the latter possibly even precipitating double diabetes, all being strong risk factors for the development of diabetic micro- and macrovascular disease [32]. The prevalence of comorbidities increased with duration of type 1 diabetes, and patients displayed diabetic nephropathy and neuropathy at a rate similar to that reported by previous studies [33].
This is to be regretted, as occurrence of diabetic complications can be minimized by aiming as closely as possible at benchmark values of diabetes care, as convincingly shown by the DCCT/EDIC study [34].
Significantly, in addition to its beneficial effects for the patients themselves, type 1 diabetes care at a DRC also comes with an economic advantage, as the costs per day amount to only 15 to 25% of that in standard hospitals (Table B in S1 File) frequently accepting the same patients with HbA1c far above therapeutic target.
In addition, lowering HbA1c somewhat below therapeutic target also provides some economic benefit per se, since patients sustaining HbA1c levels <53 mmol/mol for >3 years require less financial support during this period (−$5,214.—) than those with HbA1c ≥53 mmol/mol [35].
The limitations of this observational study, include its (i) single center observational nature, (ii) reliance on patient compliance with standardization at the DRC, and (iii) absence of control of patient adherence to treatment recommendations in their external environment after discharge, where they tend to relapse at a considerable rate [36]. Thus, the long-term benefit of intervention at a DRC needs to be answered in randomized multi-center clinical trials.
Conclusion
We conclude that i) to meet goals of type 1 diabetes care insulin treatment, food intake and exercise have to be handled in concert and the respective recommendations implemented, ii) this need is helped by simple standardization of lifestyle, which easily also could be implemented at home, iii) compliance with rules of insulin treatment in T1D care cannot be replaced by arbitrary medication with add-on glucose-lowering drugs, and iv) costs of necessary patient education can be considerably reduced if offered at a DRC instead of acute hospitals.
Supporting information
S1 File.
Table A of S1 File gives detailed results from general linear models assessing changes in vital and metabolic variables in response to 3 weeks at the DRC in patients with and without changes in insulin treatment modes. Table B summarizes costs of hospitalization depending on Austrian hospital types. Fig A depicts improvement of vital and metabolic variables in patients with and without changes in insulin treatment strategies.
https://doi.org/10.1371/journal.pone.0194135.s001
(DOCX)
Acknowledgments
The authors acknowledge the patients, who chose to participate in this observational study and the physical therapy staff at the DRC for their hard work and dedication. The authors also thank Prof. Michael Roden, Düsseldorf, for critically reviewing the manuscript and Dr. John Heath for English language editing (service paid by the Department of Laboratory Medicine, Medical University of Vienna).
References
- 1. Waldhäusl WK. The physiological basis of insulin treatment—clinical aspects. Diabetologia. 1986;29(12):837–49. Epub 1986/12/01. pmid:3552811
- 2.
Barnett D, Elliott P, Joslin M. A Centennial Portrait. Boston: Joslin Diabetes Center. 1998.
- 3. American Diabetes Association. Standards of Medical Care in Diabetes—2015. Diabetes care. 2015;38:S33–40.
- 4. Böhm BO, Dreyer M, Fritsche A, Füchtenbusch M, Gölz S, Martin S, et al. Therapie des Typ-1-Diabetes. Diabetologie und Stoffwechsel. 2016;11(S 02):S106–S16. 10.1055/s-0042-110814.
- 5. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. The New England journal of medicine. 1993;329(14):977–86. Epub 1993/09/30. pmid:8366922
- 6. Miller KM, Beck RW, Bergenstal RM, Goland RS, Haller MJ, McGill JB, et al. Evidence of a strong association between frequency of self-monitoring of blood glucose and hemoglobin A1c levels in T1D exchange clinic registry participants. Diabetes care. 2013;36(7):2009–14. Epub 2013/02/05. pmid:23378621
- 7. Tamborlane WV, Beck RW, Bode BW, Buckingham B, Chase HP, Clemons R, et al. Continuous glucose monitoring and intensive treatment of type 1 diabetes. The New England journal of medicine. 2008;359(14):1464–76. Epub 2008/09/10. pmid:18779236
- 8. Beck RW, Tamborlane WV, Bergenstal RM, Miller KM, DuBose SN, Hall CA. The T1D Exchange clinic registry. The Journal of clinical endocrinology and metabolism. 2012;97(12):4383–9. Epub 2012/09/22. pmid:22996145
- 9. Derfler K, Waldhäusl W, Zyman HJ, Howorka K, Holler C, Freyler H. Diabetes care in rural area: clinical and metabolic evaluation. Diabetes care. 1986;9(5):509–17. Epub 1986/09/01. pmid:3533476
- 10. Miller KM, Foster NC, Beck RW, Bergenstal RM, DuBose SN, DiMeglio LA, et al. Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D Exchange clinic registry. Diabetes care. 2015;38(6):971–8. Epub 2015/05/23. pmid:25998289
- 11. Huber CA, Schwenkglenks M, Rapold R, Reich O. Epidemiology and costs of diabetes mellitus in Switzerland: an analysis of health care claims data, 2006 and 2011. BMC endocrine disorders. 2014;14:44. Epub 2014/06/05. pmid:24894889
- 12. Kanavos P, van den Aardweg S, Schurer W. Diabetes expenditure, burden of disease and management in 5 EU countries. LSE Health and Social Care. 2012.
- 13. Pilacinski S, Zozulinska-Ziolkiewicz DA. Influence of lifestyle on the course of type 1 diabetes mellitus. Archives of medical science AMS. 2014;10(1):124–34. Epub 2014/04/05. pmid:24701225
- 14. Wennberg J, Gittelsohn. Small area variations in health care delivery. Science (New York, NY). 1973;182(4117):1102–8. Epub 1973/12/14.
- 15. Steinberg EP. Improving the quality of care—can we practice what we preach? The New England journal of medicine. 2003;348(26):2681–3. Epub 2003/06/27. pmid:12826644
- 16. Smith SA, Poland GA. Influenza and pneumococcal immunization in diabetes. Diabetes care. 2004;27 Suppl 1:S111–3. Epub 2003/12/25.
- 17. Ziegler D. Update 2014 on diabetic neuropathy. Der Diabetologe. 2014;10(5):376–83. 10.1007/s11428-014-1195-7.
- 18. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Annals of internal medicine. 2009;150(9):604–12. Epub 2009/05/06. pmid:19414839
- 19. Hirst JA, Stevens RJ, Farmer AJ. Changes in HbA1c Level over a 12-Week Follow-up in Patients with Type 2 Diabetes following a Medication Change. PloS one. 2014;9(3):e92458. pmid:24667212
- 20. Fraser CG. Reference change values. Clinical chemistry and laboratory medicine. 2011;50(5):807–12. Epub 2011/10/01. pmid:21958344
- 21.
Ricos C. Desirable Biological Variation Database specifications 2014 [2016-07-31]. Available from: http://www.westgard.com/biodatabase1.htm.
- 22. Clyde M, Smith KJ, Gariepy G, Schmitz N. The association between smoking and depression in a Canadian community-based sample with type 2 diabetes. Canadian journal of diabetes. 2013;37(3):150–5. Epub 2013/09/28. pmid:24070837
- 23. Choleau C, Albisser AM, Bar-Hen A, Bihan H, Campinos C, Gherbi Z, et al. A novel method for assessing insulin dose adjustments by patients with diabetes. Journal of diabetes science and technology. 2007;1(1):3–7. Epub 2007/01/01. pmid:19888373
- 24. Chimen M, Kennedy A, Nirantharakumar K, Pang TT, Andrews R, Narendran P. What are the health benefits of physical activity in type 1 diabetes mellitus? A literature review. Diabetologia. 2012;55(3):542–51. Epub 2011/12/23. pmid:22189486
- 25. Kennedy A, Nirantharakumar K, Chimen M, Pang TT, Hemming K, Andrews RC, et al. Does exercise improve glycaemic control in type 1 diabetes? A systematic review and meta-analysis. PloS one. 2013;8(3):e58861. Epub 2013/04/05. pmid:23554942
- 26. Tonoli C, Heyman E, Roelands B, Buyse L, Cheung SS, Berthoin S, et al. Effects of Different Types of Acute and Chronic (Training) Exercise on Glycaemic Control in Type 1 Diabetes Mellitus. Sports Medicine. 2012;42(12):1059–80. pmid:23134339
- 27. Quirk H, Blake H, Tennyson R, Randell TL, Glazebrook C. Physical activity interventions in children and young people with Type 1 diabetes mellitus: a systematic review with meta-analysis. Diabetic medicine a journal of the British Diabetic Association. 2014;31(10):1163–73. Epub 2014/06/27. pmid:24965376
- 28. Yardley JE, Hay J, Abou-Setta AM, Marks SD, McGavock J. A systematic review and meta-analysis of exercise interventions in adults with type 1 diabetes. Diabetes research and clinical practice. 2014;106(3):393–400. Epub 2014/12/03. pmid:25451913
- 29. Gaede P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. The New England journal of medicine. 2008;358(6):580–91. Epub 2008/02/08. pmid:18256393
- 30.
Clark N. Goals of Treatment. In: Leahy J, Cefalu W, editors. Insulin Therapy. New York: Marcel Dekker; 2002. p. 13–20.
- 31. Smith SA, Poland GA. Use of influenza and pneumococcal vaccines in people with diabetes. Diabetes care. 2000;23(1):95–108. Epub 2000/06/17. pmid:10857977
- 32. Merger SR, Kerner W, Stadler M, Zeyfang A, Jehle P, Muller-Korbsch M, et al. Prevalence and comorbidities of double diabetes. Diabetes research and clinical practice. 2016;119:48–56. Epub 2016/07/28. pmid:27449710
- 33. Harris M, Eastman R, Cowie C. Symptoms of sensory neuropathy in adults with NIDDM in the U.S. population. Diabetes care. 1993;16(11):1446–52. Epub 1993/11/01. pmid:8299433
- 34. Nathan DM. The diabetes control and complications trial/epidemiology of diabetes interventions and complications study at 30 years: overview. Diabetes care. 2014;37(1):9–16. Epub 2013/12/21. pmid:24356592
- 35. Juarez D, Goo R, Tokumaru S, Sentell T, Davis J, Mau M. Association Between Sustained Glycated Hemoglobin Control and Healthcare Costs. The American journal of pharmacy benefits. 2013;5(2):59–64. Epub 2014/01/01. pmid:24379909
- 36. Graber AL, Davidson P, Brown AW, McRae JR, Woolridge K. Dropout and relapse during diabetes care. Diabetes care. 1992;15(11):1477–83. Epub 1992/11/01. pmid:1468274