Cohort Of Diabetic Patients Research Articles

Abstract WP218: Increased Systolic Blood Pressure Variability Among Diabetic Patients is Associated With Higher Risk of Incident Stroke: A Secondary Analysis of ACCORDION

Background: Increased blood pressure variability (BPV) has been associated with stroke risk, but never specifically in patients with diabetes. Methods: This is a secondary analysis of the Action to Control Cardiovascular Risk in Diabetes Follow-On Study (ACCORDION), the long term follow-up extension of ACCORD. Visit-to-visit BPV was analyzed using all BP readings during the first 36 months. The primary outcome was incident ischemic or hemorrhagic stroke after 36 months. Differences in mean BPV was tested with Student’s t-test. We fit Cox proportional hazards models to estimate the adjusted risk of stroke across lowest vs. highest quintile of BPV and report hazard ratios along with 95% confidence intervals (CI). Results: Our analysis included 9,241 patients, with a mean (SD) age of 62.7 (6.6) years and 61.7% were male. Mean (SD) follow-up was 5.7 (2.4) years and number of BP readings per patient was 12.0 (4.3). Systolic, but not diastolic, BPV was higher in patients who developed stroke (Table 1). The highest quintile of SBP SD was associated with increased risk of incident stroke, independent of mean blood pressure or other potential confounders. (Table 2, Figure 1). There was no interaction between SBP SD and treatment arm assignment, although the interaction for glucose approached significance (Table 2). Conclusion: Higher systolic BPV was associated with incident stroke in a large cohort of diabetic patients. Future trials of stroke prevention may benefit from interventions targeting BPV reduction.

Frailty predicts a higher risk of incident urolithiasis in 525 368 patients with diabetes mellitus: a population-based study

ObjectivePatients with diabetes have an increased risk for urolithiasis, but the associated risk factors remain an active area of research. We investigated whether frailty influenced the probability of patients with...

The patient presenting with chronic limb-threatening ischaemia. Does diabetes influence presentation, limb outcomes and survival?

Peripheral arterial disease (PAD) confers an elevated risk of major amputation and delayed wound healing in diabetic patients with foot ulcers. The major international vascular societies recently developed evidence-based guidelines for the assessment and management of patients with chronic limb-threatening ischaemia (CLTI). CLTI represents the cohort of diabetic and non-diabetic patients who have PAD which is of sufficient severity to delay wound healing and increase amputation risk. Diabetic patients with CLTI are more likely to present with tissue loss, infection and have less favourable anatomy for revascularization than those without diabetes. Although diabetes is not consistently reported as a strong independent risk factor for limb loss, major morbidity and mortality in CLTI patients, it is impossible in clinical practice to isolate diabetes from comorbidities, such as end-stage renal disease and coronary artery disease which occur more commonly in diabetic patients. Treatment of CLTI in the diabetic patient is complex and should involve a multi-disciplinary team to optimize outcomes. Clinicians should use an integrated approach to management based on patient risk assessment, an assessment of the severity of the foot pathology and a structured anatomical assessment of arterial disease as suggested by the Global Vascular Guidelines for CLTI.

The EMPagliflozin compaRative effectIveness and SafEty (EMPRISE) study programme: Design and exposure accrual for an evaluation of empagliflozin in routine clinical care.

BackgroundThe EMPA‐REG OUTCOME trial showed that empagliflozin reduced the risk of cardiovascular death and hospitalization for heart failure (HHF) in diabetic patients with cardiovascular disease. EMPRISE is a study programme on the effectiveness, safety and healthcare utilization of empagliflozin in routine care, leveraging real‐world data from two commercial and one federal US data sources from 2014 to 2019.ObjectivesTo describe rationale and design of EMPRISE, assess ability to minimize confounding and evaluate the time to reach sufficient statistical power for a key study outcome, HHF, using baseline information from the first year of EMPRISE.MethodsIn 3 claims data sets, we identified a 1:1 propensity score (PS)‐matched cohort of diabetic patients ≥18 years initiating empagliflozin or a dipeptidyl peptidase‐4 inhibitor (DPP4i), resulting in 6643 total pairs. The PS model included >140 baseline covariates. We measured covariate balance via standardized differences (SD) and postmatching c‐statistic. We computed the incidence rate (IR) of HHF, predicted exposure accrual over time and calculated expected power.ResultsAfter PS matching, patient characteristics were balanced with SD <0.1 and c‐statistic between 0.54 and 0.59. The population IR of HHF was 4.4 per 1000 person‐years using a specific HHF definition and 14.8 using a broader HHF definition. In our projection, 80%‐powered analyses would require a minimum of 169 HHF events, expected to accumulate by year 3 (specific definition) or year 2 (broader definition).ConclusionBaseline information from EMPRISE provided evidence of solid confounding control and adequate exposure accrual with expected powered analyses for the primary outcomes.

Hypoglycemic episodes are associated with an increased risk of incident frailty among new onset diabetic patients

AimsPatients with diabetes mellitus (DM) are at risk for developing frailty due to the complex interplay between different cardiometabolic factors. We examined whether hypoglycemia could independently increase frailty risk besides these factors. MethodsFrom the Longitudinal Cohort of Diabetic Patients, 210,254 patients with new onset DM between 2004 and 2011 were identified, among whom 2119 non-frail patients had at least 1 hypoglycemic episode within 3 years of DM diagnosis. They were propensity score-matched to 8432 non-frail ones without hypoglycemia throughout the study period. Both groups were followed up longitudinally for incident physical frailty according to a modified FRAIL scale (Fatigue, Resistance, Ambulation, Illness, and Loss of weight). We analyzed the risk of frailty (primary) and mortality (secondary outcome) introduced by hypoglycemia, adjusted for known risk factors of frailty. ResultsThe mean age of patients (46.2% male) was 65.9 ± 14 years; diabetic patients with hypoglycemia had significantly higher comorbidity burden than those without. After 2.68 years, 172 (1.6%) patients with hypoglycemia developed incident frailty, representing a 60% higher risk (hazard ratio [HR] 1.599, 95% confidence interval [CI] 1.14–2.42). After adjusting for other risk factors, those with hypoglycemia had a significantly higher risk of frailty than those without (HR 1.443, 95% CI 1.01–2.05). Additionally, the mortality of those with hypoglycemia was 2-fold higher than those without, and the risk persisted despite confounder adjustment (HR 1.462, 95% CI 1.3–1.65). ConclusionIn this population-based cohort, hypoglycemic episodes among diabetic patients increased the risk of incident frailty and mortality.

Type2 diabetes:changing patterns of use of antidiabetic drugs 8 years after starting treatment(snds)

Abstract Objective To describe the changing patterns in the use of the various classes of antidiabetic drugs in a cohort of patients newly treated for diabetes from 2008 to 2015 on the basis of comprehensive health insurance data. Methods General scheme beneficiaries 45 years and older were identified in the French National Health Data System (SNDS) as newly treated when antidiabetic drugs were dispensed on at least 3 different dates in 2008 and at least once during the last 4 months of 2008, but were dispensed less than 3 times in 2006-2007. Treatment regimens for the last 4 months of each year were defined for this cohort. Results 158,000 people (53% men, mean age 62 years) initiated antidiabetic drug therapy in 2008: monotherapy (without insulin): 74%, dual therapy: 17%, treatment with 3 or more drugs: 3%, and insulin therapy: 7%. The proportion of patients taking monotherapy decreased by 13 percentage points during the second year and only 32% of the cohort was treated by monotherapy 8 years after starting treatment. The proportions of patients taking dual therapy increased by 4 percentage points (21% in 2015), triple therapy increased by 8 percentage points (11%), a combination of insulin+other antidiabetics increased by 4 percentage points (7%) and insulin alone remained stable (4%). At the last quarter of 2015, 12% of patients did not receive any antidiabetic drugs and 13% had died. The creation of a cohort of diabetic patients newly treated in 2013 demonstrated changing prescribing practices: a lower proportion of monotherapy in the first year of treatment, but a growing use of metformin. Conclusions Treatment was intensified during the first 8 years for one half of the cohort. The maximal dosage of Metformin is not used before introducing a second antidiabetic in 1 case in 2. Insufficient use of Metformin, in general and in terms of dosage, was observed, although this use is improving. Key messages Initiation of antidiabetic drug therapy in people 45 years and older consisted of monotherapy in 3/4 of cases in 2008. Insufficient use of Metformin, in general and in terms of dosage, was observed, although this use is improving.

P2489What are the main determinants of an increase in bnp level in asymptomatic diabetic patients without known cardiac disease?

Abstract Background Diabetes mellitus is associated with a high risk of heart failure. The predictors of futures heart failure events in diabetic patients are not clearly understood. BNP measurement can be used as a surrogate endpoint for the diagnosis of heart failure. We investigated the determinants of an increase in BNP level in a large cohort of asymptomatic diabetic patients without known cardiac disease Methods This prospective study included consecutive stable diabetic (type 1 or 2) patients coming for yearly check-up between March 2015 and July 2018 in the university center for the study of diabetes and its complications. Patients with an history of cardiac disease (coronary artery disease, atrial fibrillation, cardiomyopathy, previous heart failure ...) were excluded. All patients had a complete clinical exam, blood pressure measurement (3 consecutive times – mean of 2 lasts measurements), ECG, and blood sample including HbA1C, risk factors assessment, renal function (CKD-EPI) and BNP measurement. Data are presented as mean±SD or median - Spearman's rank and multivariate regression were used for analysis. Results 3743 patients (mean age 57±14 y.o. – 57% male – 78% / 18% / 4% of type 2, type 1 or other type of diabetes respectively – Mean duration of diabetes 17 [1–63] y. – 44% treated with insulin) were studied. Mean±SD / median [min-max] BNP level was 25±39 / 12 [4–737] ng/L. BNP was &lt;20 / 21–35 / 36–50 / 51–100 / 101–400 / &gt;400 ng/L in 69 / 15 / 6 / 7 / 3 / 0.1% of the population respectively. The parameters most correlated with BNP level in type 1 and type 2 diabetes were age, duration of diabetes, renal function, HbA1C, and pulsed pressure. For multivariate analysis, renal function was removed of the model as it was highly correlated with age (r=−0.68). Multivariate analysis demonstrated that in type 1 diabetes, high BNP level was linked to age (p&lt;0.001), pulsed pressure (p&lt;0.001), duration of diabetes (p=0.003) and HbA1C (p=0.02). In type 2 diabetes, high BNP level was linked to age (p&lt;0.0001), pulsed pressure (p&lt;0.0001), duration of diabetes (p=0.005) but not HbA1C (p=0.09). Interestingly the type of treatment (mainly insulin treatment) was not independently related to an increase in BNP level. Conclusion Age, pulsed pressure and duration of diabetes are the main determinants of an increased level of BNP in asymptomatic diabetic patients without any history of cardiac disease. This result could help to select a population who could benefit to a more extensive follow up concerning heart failure.

Left Ventricular Diastolic Dysfunction in Type 2 Diabetes-Progress and Perspectives.

In-depth understanding of early cardiovascular manifestations in diabetes is high on international research and prevention agendas given that cardiovascular events are the leading cause of death for diabetic patients. Our aim was to review recent developments in the echocardiographic assessment of left ventricular diastolic dysfunction (LVDD) as a telltale pre-clinical disturbance preceding diabetic cardiomyopathy. We analyzed papers in which patients had been comprehensively assessed echocardiographically according to the latest LVDD guidelines (2016), and those affording comparisons with previous, widely used recommendations (2009). We found that the updated algorithm for LVDD is more effective in predicting adverse cardiovascular events in patients with established LVDD, and less specific in grading other patients (labelled “indeterminate”). This may prove instrumental for recruiting “indeterminate” LVDD cases among patients with type 2 diabetes mellitus (T2DM) in future screening programs. As an interesting consideration, the elevated values of the index E/e’ can point to early diastolic impairment, foretelling diabetic cardiomyopathy. Identifying subclinical signs early makes clinical sense, but the complex nature of T2DM calls for further research. Specifically, longitudinal studies on rigorously selected cohorts of diabetic patients are needed to better understand and predict the subtle, slow onset of cardiac manifestations with T2DM as a complicating backdrop.

Costs and Length of Stay of Hospitalizations due to Diabetes-Related Complications.

Background Diabetes mellitus (DM) has become a significant worldwide public health problem and economic burden because a great proportion of healthcare costs has been spent on the treatment of DM and its related complications. The aim of this study was to examine the costs and length of stay (LoS) of hospitalizations due to diabetes-related complications in Taiwan. Methods This study is a retrospective claim database analysis using the Longitudinal Cohort of Diabetes Patients, with 2012 used as the base year. The hospitalization costs and LoS per admission were estimated for each complication of interest using data from the LHDB 2004 to 2012 cohorts. The presence of eight DM-related complications were identified using the ICD-9-CM codes and procedure codes. ANOVA was used to examine the relationships of diabetes duration with the LoS and costs of the complications. Results A total of 27,473 DM patients who were hospitalized in 2012 due to one of the examined DM-related complications were identified. The most common complications that caused the hospitalizations were nonfatal stroke (34.7%) and nonfatal ischemic heart disease (IHD) (28.7%). Amputation was the complication with the longest hospital stay, with a mean ± SD of 21.6 ± 14.1 days, followed by nonfatal stroke (13.6 ± 11.3), ulcer (12.7 ± 11.8), and fatal IHD (12.2 ± 13.6). The complications with the greatest hospitalization cost were fatal IHD (mean = TWD 306,209.8; median = TWD 221,417.0; 1TWD = 0.034USD) and fatal myocardial infarction (mean = TWD 272,840.1; median = TWD 174,008). Conclusions This study indicates that DM-related complications are associated with significant hospital LoS and costs. The study results could be useful for economic evaluations of diabetes treatments and the estimation of the overall economic impact of diabetes.

Knowledge Learning Symbiosis for Developing Risk Prediction Models from Regional EHR Repositories.

Secondary use of regional EHR data suffers several problems, including data selection bias and limited data size caused by data incompleteness. Here, we propose knowledge learning symbiosis (KLS) as a framework to incorporate domain knowledge to address the problems and make better secondary use of EHR data. Under the framework, we introduce three main categories of methods: knowledge injection to input features, objective functions, and output labels, where knowledge-enhanced neural network (KENN) was first introduced to inject knowledge into objective functions. A case study was conducted to build a cardiovascular disease risk prediction model on the type 2 diabetes patient cohort using regional EHR repositories. By incorporating a well-established knowledge risk model as domain knowledge under our KLS framework, we increased risk prediction performance both on small and biased data, where KENN showed the best performance among all methods.

Dietary practices in type 2 diabetes patients at Bhopal, Madhya Pradesh: a cross-sectional study

Background: Diabetes mellitus (DM) is a rapidly growing health problem in India. Diet and physical activity are important modifiable risk factors affecting the incidence, severity and management of DM. The aim of the study was to assess diet of diabetes mellitus patients. 87 adults from a cohort of diabetic patients attending the Medical Clinics at the Hamidia Hospital Bhopal were invited for the study.Methods: Information about their nutritional status &amp; food consumption pattern was taken by 24-hour dietary recall and food frequency questionnaire. Data was analysed using Epi Info and MS Excel. Frequency counts and percentages were used to describe the demographic characteristics of the participants while the significance difference in between male and female participants. The significance level was set at P &lt;0.05 and 0.01. Statistical analysis used: MS Excel and Epi Info.Results: The study sample was 87 out of which 56 were females and 31 were males; mean age of DM patients was 48.21±12.98.The mean energy intake was 1386.52 kcal and 1125.79 kcal among male and female. The mean protein intake was 68.71 g and 50.50 among male and female diabetics. There was significant difference between male and female age. There were no significant difference in body mass index and waist hip ratio between male and female. 30 (34.5%) &amp; 16 (18.4%) were overweight and obese respectively.Conclusions: Based on these findings, the dietary practices of diabetic patients are inadequate and require improvement. Education and counselling about diet of a diabetes patient is needed.

Long-term Outcomes of an Endovascular-First Approach for Diabetic Patients With Predominantly Tibial Disease Treated in a Multidisciplinary Setting

Randomized studies suggest that open lower extremity revascularization procedures are associated with improved outcomes compared with endovascular peripheral vascular interventions (PVIs). However, advances in endovascular technologies and treatment by multidisciplinary limb preservation teams have shown improved outcomes. The aim of our study was to compare perioperative and long-term outcomes after open versus PVI procedures in diabetic patients with chronic limb-threatening ischemia (CLTI) treated in a multidisciplinary setting. All patients presenting to our multidisciplinary diabetic limb-preservation service from 6/2012 to 07/2018 were enrolled in a prospective database. Patients who underwent either an open lower extremity bypass (LEB) or a PVI for CLTI were included in the analysis. Perioperative (30-day) complications and 4-year patency and limb salvage rates were compared between PVI and LEB using chi-squared tests, Kaplan-Meier curve analyses, and stepwise multivariable Cox proportional hazards models. A total of 195 lower extremity revascularization procedures were performed in 120 patients (mean age: 65.0±1.0years, 61.7% male, 63.3% black), including 53 (27.2%) open procedures and 142 (72.8%) PVIs. Nearly two-thirds of procedures (65.6%) treated multilevel diseases, while 27.2% treated isolated tibial disease and 7.2% treated isolated femoropopliteal disease. More than half of the procedures (53.3%) were performed for Wound, Ischemia, and foot Infection (WIfI) classification stage 4 limbs, 25.1% for stage 3, and 21.6% for stage 1/2. In the LEB group, 67.9% of targets were infrapopliteal. In the PVI group, 63.4% of procedures were isolated tibial interventions or were multilevel interventions including the tibial segment. Perioperative complications occurred in 52.8% of LEB versus 12.0% of PVI (P<0.001). At 4years postoperatively, there was no significant difference in crude (unadjusted) primary patency for PVI versus LEB (34.5±6.6% vs. 49.6±8.1, P=0.89). Secondary patency was better for the LEB group (50.3±7.4% vs. 55.4±7.5%; P=0.04), but amputation-free survival was similar (65.1±6.7% vs. 60.9±9.7%; P=0.79). After adjusting for baseline differences between groups, primary patency (hazard ratio [HR]: 0.61; 95% confidence interval [CI]: 0.34 to 1.10) and amputation-free survival (HR: 1.51; 95% CI: 0.71 to 2.34) remained similar for PVI versus LEB, but secondary patency was persistently lower for PVI (HR: 0.35; 95% CI: 0.14 to 0.90). In this cohort of diabetic patients with CLTI undergoing predominantly tibial interventions, open revascularization was associated with a higher risk of perioperative complications than PVIs. While secondary patency rates were better after LEBs, our data suggest that an endovascular-first approach results in equivalent long-term amputation-free survival for diabetic patients treated in a multidisciplinary setting.

Vitamin D levels and platelet reactivity in diabetic patients receiving dual antiplatelet therapy

BackgroundHypovitaminosis D represents an emerging cardiovascular risk factor, and especially among higher-risk subsets of patients, such as in those with diabetes mellitus. The anti-inflammatory and anti-thrombotic properties of vitamin D, in fact, could be even more beneficial among diabetics, where platelet hyperreactivity and suboptimal response to antiplatelet drugs has been associated with poorer outcomes. However, no study has so far evaluated the impact of vitamin D levels on platelet reactivity and high-on treatment platelet reactivity (HRPR) among diabetic patients receiving dial antiplatelet therapy (DAPT). MethodsOur population is represented by a consecutive cohort ofdiabetic patients treated with DAPT (ASA + clopidogrel or ticagrelor or dose-adjusted prasugrel) for an acute coronary syndrome or elective PCI, undergoing platelet reactivity assessment at 30–90 days post-discharge. Aggregation was assessed by multiple-electrode aggregometry. HRPR was defined for values above the lower limit of normality (in non-treated patients). ResultsWe included 440 patients, that were divided according to quartiles values of vitamin D (< 9.4; 9.4–15.59; 15.6–21.64; ≥ 21.65 ng/ml). Among them, 31 were excluded as chronically treated with vitamin D supplementation. Lower vitamin D quartiles were associated with more advanced age (p = 0.01), female gender (p = 0.04), renal failure (p = 0.005), history of previous MI (p = 0.01), CABG and use of diuretics (p = 0.003), severe coronary disease (p = 0.002), but lower ejection fraction (p = 0.001), treatment with statins (p = 0.04) and new ADP-antagonists (p = 0.002). Vitamin D levels related with higher HbA1c (p = 0.001), cholesterol (p = 0.02) and creatinine (p = 0.004) and lower hemoglobin (p = 0.004).The prevalence of HRPR with ASA was low and not related to vitamin D quartiles (3.4% vs 2.7% vs 1.8% vs 2.1%, p = 0.44; adjusted OR[95%CI] = 1.16[0.60–2.26], p = 0.67).The prevalence of HRPR for ADP antagonists was associated to hypovitaminosis D (40.2% vs 29.1% vs 29.4% vs 25.5%, p = 0.03; (adjusted OR[95%CI] = 1.76[1.04–2.98], p = 0.036for I vs II-IV quartile). The impact of vitamin D quartiles, was significant only in patients on new ADP antagonists (n = 225, of whom 81 on prasugrel 5 mg; p = 0.03; adjusted OR[95%CI] = 3.12[1.34–7.49], p = 0.009) but not with clopidogrel (p = 0.85, adjusted OR[95%CI] = 1.05[0.49–2.24], p = 0.89). ConclusionsAmong diabetic patients receiving dual antiplatelet therapy for an acute coronary syndrome or elective percutaneous coronary intervention, severe vitamin D deficiency is associated with a higher ADP-mediated platelet reactivity and rate of HRPR, and especially for new ADP-antagonists over clopidogrel.

967-P: Association between Visit-to-Visit Blood Pressure Variability and Blood Glucose Control, a Retrospective Analysis of Three Population-Based Cohorts of Diabetic Patients

Background: Relationship between blood pressure (BP) and blood glucose control is still unsure. This study aimed to find whether visit-to-visit variability (VVV) of BP could be an independent risk factor for blood glucose control. Methods: A retrospective analysis was conducted in 95254 Chinese primary care diabetic patients from Oct. 2014 to Nov. 2018 with repeated measurements of BP and fasting plasma glucose (FPG) every 3 months, including 3 cohorts: Putuo District diabetes cohort (PTD, N=43590), Xuhui District diabetes cohort (XHD, N=45451) and Changning District diabetes cohort (CND, N=6213). The VVV of systolic blood pressure (SBP) was evaluated by average successive difference (ASD), and glucose control was evaluated by “FPG well-controlled proportion” (FPG≤7mmol/L as “well-controlled”), over 16 follow-ups. Logistic regression was used to explore the association of SBP ASD and FPG well-controlled proportion. The full model adjusted the baseline characteristics considering age, gender, BMI, exercise, medication and diet. Model was built in PTD and validated in the other 2 cohorts. Results: Patients in PTD can be divided into 3 groups by VVV of SBP, which were “stable,” “fluctuating” and “highly fluctuating” groups. With “stable” group as reference, higher VVV of SBP was associated with lower FPG well-controlled proportion. In PTD, “fluctuating” group, OR: 1.18, 95% CI: 1.13-1.24; “highly fluctuating” group, OR: 1.42, 95% CI: 1.35-1.49. Results were consistent in the 2 validated cohorts and ORs for the “highly fluctuating” group were 2.37, 95% CI: 2.23-2.52 and 1.34, 95% CI: 1.15-1.56. In the whole population (N =95254), “highly fluctuating” group was associated with the worst blood glucose control with OR: 1.67, 95% CI: 1.62-1.73. Conclusion: High VVV of SBP could be an independent risk factor for blood glucose control, which suggests extra surveillance for those with highly fluctuating SBP. Disclosure S. Zhang: None. L. Xie: None. W. Zhang: None. B. Qian: None.

Gut microbiome-derived phenyl sulfate contributes to albuminuria in diabetic kidney disease

Diabetic kidney disease is a major cause of renal failure that urgently necessitates a breakthrough in disease management. Here we show using untargeted metabolomics that levels of phenyl sulfate, a gut microbiota-derived metabolite, increase with the progression of diabetes in rats overexpressing human uremic toxin transporter SLCO4C1 in the kidney, and are decreased in rats with limited proteinuria. In experimental models of diabetes, phenyl sulfate administration induces albuminuria and podocyte damage. In a diabetic patient cohort, phenyl sulfate levels significantly correlate with basal and predicted 2-year progression of albuminuria in patients with microalbuminuria. Inhibition of tyrosine phenol-lyase, a bacterial enzyme responsible for the synthesis of phenol from dietary tyrosine before it is metabolized into phenyl sulfate in the liver, reduces albuminuria in diabetic mice. Together, our results suggest that phenyl sulfate contributes to albuminuria and could be used as a disease marker and future therapeutic target in diabetic kidney disease.

SUN-139 Pneumococcal Vaccination Rates among Pediatric Diabetics

Since 1997 it has been recommended by the CDC that pediatric patients over the age of two with certain chronic illnesses, including diabetes mellitus, receive 23 valent pneumococcal polysaccharide vaccine (PPSV23) in addition to the PCV7/PCV13 primary series given to all children (1). This recommendation is attributable to the need to prevent invasive pneumococcal infection in these susceptible populations. School aged children are frequently colonized with Streptococcus pneumonia bacteria (2), and those with diabetes are at a greater risk of developing serious pneumococcal infections (3). There is some evidence in the international literature suggesting a lack of compliance with this guideline (4). Our aim was to assess the pneumococcal polysaccharide vaccination rate among a cohort of diabetic patients in our pediatric endocrinology clinic. We conducted a retrospective chart review of all diabetic patients seen at the pediatric endocrinology clinic in Staten Island, New York between the ages of 2-20. We excluded those who had no follow up for 1 year. We collected data including age of diagnosis, type of diabetes, and vaccination records. Vaccine records were obtained from the New York Citywide Immunization Registry (CIR). Out of all the diabetics in our clinic, 126 charts met eligibility criteria. Of the 126 patients, 105 had type 1 diabetes mellitus, 14 were type 2, and 7 were other (including MODY and secondary to other underlying conditions). The type 1 diabetics had the lowest mean age of diagnosis of 7 years, compared with mean age of 12 years for the other two groups. Of the 126 patients, only 3.2% (95% CI: 0.9% , 7.9%) were found to have received the PPSV23, all type 1 diabetics. Our results show a poor rate of PPSV23 vaccination among pediatric patients with diabetes. Additionally, as the age of diagnosis is lowest in type 1 diabetics, most type 1 diabetics have a longer period of exposure to potentially serious pneumococcal infections. Thus, it is imperative that all diabetics, especially type 1 diabetics, receive the vaccine as early as possible following diagnosis for optimal protection. Our findings, in combination with the limited additional pediatric literature published to date, suggest that these low vaccination rates are likely generalizable to other geographic regions as well as other chronic illnesses in which children have increased susceptibility to pneumococcal disease (5, 6). Given these results, we recommend endocrinologists and pediatricians alike take steps to ensure adherence to the PPSV23 recommendations.

Determinants of inertia with lipid-lowering treatment in patients with type 2 diabetes mellitus

ObjectiveTo assess the control of cLDL in diabetic patients, to measure the impact on such control of inertia with lipid-lowering agents and to explore factors that allow for predicting this inertia. MethodsStudy of historical cohorts of diabetic patients. The proportion of patients who achieved the target cLDL levels was estimated. Therapeutic inertia was considered when the dose of the lipid-lowering agents was not adjusted, or a lipid-lowering agent was not changed or added in patients with initial cLDL outside the target. Change in cLDL from the first to the last visit and inertia with lipid-lowering drugs were analyzed according to comorbidities, cardiovascular risk factors and treatments used. ResultsThe study simple consisted of 639 patients (mean follow-up time 11.1±11.2 months), of whom 27.5% achieved target cLDL levels. Inertia occurred in 43.6% of patients with initial cLDL outside the target. Independent predictors of inertia were the initial cLDL (p<0.001), polyneuropathy (p=0.014), adjustment of antihypertensive agents (p=0.002), adequacy of lipid-lowering agents (p<0.001), use of ezetimibe (p=0.001) and adherence to lipid-lowering drugs (p=0.015). ConclusionsInertia with lipid-lowering agents in a diabetic patient is less frequent in the presence of higher cLDL values, in cases of polyneuropathy, when antihypertensive agents are adjusted or changed, and when non-adherence is detected. The adequate initial prescription of statins and the association with ezetimibe decrease the likelihood of committing inertia.

Clinical factors associated with persistently poor diabetes control in the Veterans Health Administration: A nationwide cohort study.

ObjectivePatients with persistent poorly-controlled diabetes mellitus (PPDM) despite engagement in clinic-based care are at particularly high risk for diabetes complications and costs. Understanding this population’s demographics, comorbidities and care utilization could guide strategies to address PPDM. We characterized factors associated with PPDM in a large sample of Veterans with type 2 diabetes.MethodsWe identified a cohort of Veterans with medically treated type 2 diabetes, who received Veterans Health Administration primary care during fiscal years 2012 and 2013. PPDM was defined by hemoglobin A1c levels uniformly >8.5% during fiscal year (FY) 2012, despite engagement with care during this period. We used FY 2012 demographic, comorbidity and medication data to describe PPDM in relation to better-controlled diabetes patients and created multivariable models to examine associations between clinical factors and PPDM. We also constructed multivariable models to explore the association between PPDM and FY 2013 care utilization.ResultsIn our cohort of diabetes patients (n = 435,820), 12% met criteria for PPDM. Patients with PPDM were younger than better-controlled patients, less often married, and more often Black/African-American and Hispanic or Latino/Latina. Of included comorbidities, only retinopathy (OR 1.68, 95% confidence interval (CI): 1.63,1.73) and nephropathy (OR 1.26, 95% CI: 1.19,1.34) demonstrated clinically significant associations with PPDM. Complex insulin regimens such as premixed (OR 10.80, 95% CI: 10.11,11.54) and prandial-containing regimens (OR 18.74, 95% CI: 17.73,19.81) were strongly associated with PPDM. Patients with PPDM had higher care utilization, particularly endocrinology care (RR 3.56, 95% CI: 3.47,3.66); although only 26.4% of patients saw endocrinology overall.ConclusionPPDM is strongly associated with complex diabetes regimens, although heterogeneity in care utilization exists. While there is evidence of underutilization, inadequacy of available care may also contribute to PPDM. Our findings should inform tailored approaches to meet the needs of PPDM, who are among the highest-risk, highest-cost patients with diabetes.

ConvSCCS: convolutional self-controlled case series model for lagged adverse event detection

With the increased availability of large electronic health records databases comes the chance of enhancing health risks screening. Most post-marketing detection of adverse drug reaction (ADR) relies on physicians' spontaneous reports, leading to under-reporting. To take up this challenge, we develop a scalable model to estimate the effect of multiple longitudinal features (drug exposures) on a rare longitudinal outcome. Our procedure is based on a conditional Poisson regression model also known as self-controlled case series (SCCS). To overcome the need of precise risk periods specification, we model the intensity of outcomes using a convolution between exposures and step functions, which are penalized using a combination of group-Lasso and total-variation. Up to our knowledge, this is the first SCCS model with flexible intensity able to handle multiple longitudinal features in a single model. We show that this approach improves the state-of-the-art in terms of mean absolute error and computation time for the estimation of relative risks on simulated data. We apply this method on an ADR detection problem, using a cohort of diabetic patients extracted from the large French national health insurance database (SNIIRAM), a claims database containing medical reimbursements of more than 53 million people. This work has been done in the context of a research partnership between Ecole Polytechnique and CNAMTS (in charge of SNIIRAM).

Maintenance of good glycaemic control is challenging – A cohort study of type 2 diabetes patient in North Karelia, Finland

This study assessed type 2 diabetes treatment outcomes and process indicators using a comprehensive type 2 diabetes patient cohort in North Karelia, Finland, from 2011 to 2016. Data from all diagnosed type 2 diabetes patients (n=8429) living in North Karelia were collated retrospectively from regional electronic patient records. We assessed whether HbA1c and low-density lipoprotein (LDL) were measured and managed as recommended. The HbA1c measurement rate improved (78% vs 89%) during 2011-2012 and 2015-2016, but a gradual deterioration in glycaemic control (HbA1c<7.0% or 53mmol/mol) was observed among both females (75% vs 67%) and males (72% vs 64%). The LDL measurement rate initially improved from the baseline. LDL control (<2.5mmol/L) improved among both females (52% vs 59%) and males (58% vs 66%). A gender difference was observed in the achievement of the treatment target for LDL, with females showing worse control. Low-density lipoprotein (LDL) control in type 2 diabetes patients has improved, but the existence of gender disparities needs further attention. Maintaining appropriate HbA1c control among type 2 diabetes patients over time appears to be difficult. Active follow-up and tailored treatment have the potential to improve the quality of care. Electronic patient records could be more efficiently used to improve the quality of care and to support decision-making.