Automatic identification of type 2 diabetes, hypertension, ischaemic heart disease, heart failure and their levels of severity from Italian General Practitioners' electronic medical records: a validation study

Diabetes duration, cholesterol levels, and risk of cardiovascular diseases in individuals with type 2 diabetes

OBJECTIVE

To investigate the association of diabetes duration with cardiovascular disease (CVD) risk and to examine the relationship between lipid levels and CVD risk over the duration.

METHODS

Using the Korean National Health Insurance Service Cohort database, we identified 2,359,243 subjects with type 2 DM aged ≥20 years in 2015-2016. Baseline lipid levels and diabetes duration were evaluated, and followed up until December 2020 (mean follow-up, 3.9 years). Subjects were categorized according to diabetes duration (new-onset, <5 years, 5-9 years, or ≥10 years). We analyzed the new-onset diabetes group with low-density lipoprotein cholesterol (LDL-C), <70 mg/dL, as the reference group. The hazard ratios (HRs) and 95% confidence intervals (CIs) of myocardial infarction (MI), and ischemic stroke (IS) were estimated using a Cox proportional hazards model adjusted for potential confounders.

RESULTS

During follow-up, 45,883 cases of MI and 53,538 cases of IS were identified. The risk of MI or IS began to increase at LDL-C ≥160 mg/dL in the new-onset diabetes group, and at LDL-C ≥130 mg/dL in the diabetes duration <5 years group. Among subjects with a diabetes duration of 5-9 years, LDL-C 100-129 mg/dL, LDL-C 130-159 mg/dL, and ≥160 mg/dL were significantly associated with the risk of MI, with HRs (95% CI) of 1.13 (1.04-1.22), 1.28 (1.17-1.39), and 1.58 (1.42-1.76), respectively. The risk of MI in the diabetes duration ≥10 years group was increased by 16%, even in the LDL-C 70-99 mg/dL population (HR [95% CI] 1.16 [1.08-1.25]).

CONCLUSIONS

This population-based longitudinal study revealed that the LDL-C cutoff level for increasing the risk of cardiovascular disease varied with diabetes duration, and that the target LDL-C level should depend on the duration.

LDL-cholesterol target levels achievement in high-risk patients: An (un)expected gender bias

BACKGROUND AND AIMS

Lowering low-density lipoprotein cholesterol (LDL-C) is the cornerstone of cardiovascular disease prevention. Collection of epidemiological data is crucial for monitoring healthcare appropriateness. This analysis aimed to evaluate the proportion of high-risk patients who achieved guidelines recommended LDL-C goal, and explore the predictors of therapeutic failure, with a focus on the role of gender.

METHODS AND RESULTS

Health administrative and laboratory data from seven Local Health Districts in Tuscany were collected for residents aged ≥45 years with a history of major adverse cardiac or cerebrovascular event (MACCE) and/or type 2 diabetes mellitus (T2DM) from January 1, 2019, to January 1, 2021. The study aimed to assess the number of patients with optimal levels of LDL-C (<55 mg/dl for patients with MACCE and <70 mg/dl for patients with T2DM without MACCE). A cohort of 174 200 individuals (55% males) was analyzed and it was found that 11.6% of them achieved the target LDL-C levels. Female gender was identified as an independent predictor of LDL-C target underattainment in patients with MACCE with or without T2DM, after adjusting for age, cardiovascular risk factors, comorbidities, and district area (adjusted-IRR 0.58 ± 0.01; p < 0.001). This result was consistent in subjects without lipid-lowering therapies (adjusted-IRR 0.56 ± 0.01; p < 0.001).

CONCLUSION

In an unselected cohort of high-risk individuals, females have a significantly lower probability of reaching LDL-C recommended targets. These results emphasize the need for action to implement education for clinicians and patients and to establish clinical care pathways for high-risk patients, with a special focus on women.

Diabetes severity is strongly associated with the risk of active tuberculosis in people with type 2 diabetes: a nationwide cohort study with a 6-year follow-up

BACKGROUND

Many have the rising coincidence of diabetes mellitus (DM) and endemic tuberculosis (TB). We evaluated whether the severity of diabetes is associated with an increased risk of active TB infection.

METHODS

 Using a nationally representative database from the Korean National Health Insurance System, 2, 489, 718 people with type 2 DM who underwent a regular health checkup during 2009-2012 were followed up until the end of 2018. The diabetes severity score parameters included the number of oral hypoglycemic agents (≥ 3), insulin use, diabetes duration (≥ 5 years), and the presence of chronic kidney disease (CKD) or cardiovascular disease. Each of these characteristics was scored as one point, and their sum (0-5) was used as the diabetes severity score.

RESULTS

We identified 21, 231 cases of active TB during a median follow-up of 6.8 years. Each parameter of the diabetes severity score was associated with an increased risk of active TB (all P < 0.001). Insulin use was the most significant factor related to the risk of TB, followed by CKD. The risk of TB increased progressively with increasing diabetes severity score. After adjusting for possible confounding factors, the hazard ratio (95% confidence interval) for TB were 1.23 (1.19-1.27) in participants with one parameter, 1.39 (1.33-1.44) in those with two parameters, 1.65 (1.56-1.73) in those with three parameters, 2.05 (1.88-2.23) in those with four parameters, and 2.62 (2.10-3.27) in those with five parameters compared with participants with no parameters.

CONCLUSION

Diabetes severity was strongly associated in a dose-dependent manner with the occurrence of active TB. People with a higher diabetes severity score may be a targeted group for active TB screening.

A nationwide cohort study on diabetes severity and risk of Parkinson disease

There is growing evidence that patients with type 2 diabetes mellitus (DM) have an increased risk of developing Parkinson's disease (PD) and share similar dysregulated pathways. We aimed to determine whether the risk of PD increases as diabetes progresses among patients with type 2 DM. Using a nationally representative database from the Korean National Health Insurance System, 2,362,072 individuals (≥40 years of age) with type 2 DM who underwent regular health checkups during 2009-2012 were followed up until the end of 2018. The diabetes severity score parameters included the number of oral hypoglycemic agents, diabetes duration, insulin use, or presence of chronic kidney disease, diabetic retinopathy, or cardiovascular disease. Each of these characteristics was scored as one unit of diabetes severity and their sum was defined as a diabetes severity score from 0-6. We identified 17,046 incident PD cases during the follow-up. Each component of the diabetes severity score showed a similar intensity for the risk of PD. Compared with subjects with no parameters, HR values (95% confidence intervals) of PD were 1.09 (1.04-1.15) in subjects with one diabetes severity score parameter, 1.28 (1.22-1.35) in subjects with two parameters, 1.55 (1.46-1.65) in subjects with three parameters, 1.96 (1.82-2.11) in subjects with four parameters, 2.08 (1.83-2.36) in subjects with five parameters, and 2.78 (2.05-3.79) in subjects with six parameters. Diabetes severity was associated with an increased risk of developing PD. Severe diabetes may be a risk factor for the development of PD.

How many and who are patients with heart failure eligible to SGLT2 inhibitors? Responses from the combination of administrative healthcare and primary care databases

BACKGROUND

Recent successful findings (i.e. DAPA-HF trial) in patients with heart failure (HF) with/without diabetes treated with sodium-glucose co-transporter inhibitors (SGLT2-I) have fostered real-world data analyses. Fondazione Ricerca e Salute's (ReSD) administrative and Health Search's (HSD) primary healthcare databases were combined in the ReS-HS DB Consortium, to identify and characterize HF-patients eligible to SGLT2-I, and assess their costs charged to the Italian National Health Service (INHS).

METHODS AND RESULTS

Eligibility to SGLT2-I was HF diagnosis, age ≥ 18 years, reduced (≤40%) ejection fraction (HFrEF) and glomerular filtration rate (GFR) ≥30 ml/min. The HSD, including 13,313 HF-patients (1.5% of the total HSD population) was used to develop and test the algorithms for imputing HFrEF and GFR ≥ 30 ml/min, based on a set of covariates, to the ReSD, including 67,369 (1.5% of the total ReSD population). Subjects eligible to SGLT2-I were 2187 in HSD (61.1% of HFrEF); after the imputation, 15,145 in ReSD (58.8% of HFrEF). Prevalence of eligibility to SGLT2-I was higher in males then in females and increased with age; diabetic patients were 44.3% and 33.4% of HSD and ReSD populations eligible to SGLT2-I, respectively. Estimated from ReSD, the mean annual cost charged to the INHS per patient with HF eligible to SGLT2-I was €7122 (68% due to hospitalizations).

CONCLUSIONS

Approximately 20% of patients with HF was eligible to SGLT2-I. Real-world data can identify, quantify and characterize patients eligible to SGLT2-Is and assess related costs for the health care system, thus providing useful information to Regulatory Decision makers.

Large-scale evidence generation and evaluation across a network of databases for type 2 diabetes mellitus (LEGEND-T2DM): a protocol for a series of multinational, real-world comparative cardiovascular effectiveness and safety studies

INTRODUCTION

Therapeutic options for type 2 diabetes mellitus (T2DM) have expanded over the last decade with the emergence of cardioprotective novel agents, but without such data for older drugs, leaving a critical gap in our understanding of the relative effects of T2DM agents on cardiovascular risk.

METHODS AND ANALYSIS

The large-scale evidence generations across a network of databases for T2DM (LEGEND-T2DM) initiative is a series of systematic, large-scale, multinational, real-world comparative cardiovascular effectiveness and safety studies of all four major second-line anti-hyperglycaemic agents, including sodium-glucose co-transporter-2 inhibitor, glucagon-like peptide-1 receptor agonist, dipeptidyl peptidase-4 inhibitor and sulfonylureas. LEGEND-T2DM will leverage the Observational Health Data Sciences and Informatics (OHDSI) community that provides access to a global network of administrative claims and electronic health record data sources, representing 190 million patients in the USA and about 50 million internationally. LEGEND-T2DM will identify all adult, patients with T2DM who newly initiate a traditionally second-line T2DM agent. Using an active comparator, new-user cohort design, LEGEND-T2DM will execute all pairwise class-versus-class and drug-versus-drug comparisons in each data source, producing extensive study diagnostics that assess reliability and generalisability through cohort balance and equipoise to examine the relative risk of cardiovascular and safety outcomes. The primary cardiovascular outcomes include a composite of major adverse cardiovascular events and a series of safety outcomes. The study will pursue data-driven, large-scale propensity adjustment for measured confounding, a large set of negative control outcome experiments to address unmeasured and systematic bias.

ETHICS AND DISSEMINATION

The study ensures data safety through a federated analytic approach and follows research best practices, including prespecification and full disclosure of results. LEGEND-T2DM is dedicated to open science and transparency and will publicly share all analytic code from reproducible cohort definitions through turn-key software, enabling other research groups to leverage our methods, data and results to verify and extend our findings.

Machine learning optimization of an electronic health record audit for heart failure in primary care

Aims

The diagnosis of heart failure (HF) is an important problem in primary care. We previously demonstrated a 74% increase in registered HF diagnoses in primary care electronic health records (EHRs) following an extended audit procedure. What remains unclear is the accuracy of registered HF pre‐audit and which EHR variables are most important in the extended audit strategy. This study aims to describe the diagnostic HF classification sequence at different stages, assess general practitioner (GP) HF misclassification, and test the predictive performance of an optimized audit.

Methods and results

This is a secondary analysis of the OSCAR‐HF study, a prospective observational trial including 51 participating GPs. OSCAR used an extended audit based on typical HF risk factors, signs, symptoms, and medications in GPs' EHR. This resulted in a list of possible HF patients, which participating GPs had to classify as HF or non‐HF. We compared registered HF diagnoses before and after GPs' assessment. For our analysis of audit performance, we used GPs' assessment of HF as primary outcome and audit queries as dichotomous predictor variables for a gradient boosted machine (GBM) decision tree algorithm and logistic regression model. Of the 18 011 patients eligible for the audit intervention, 4678 (26.0%) were identified as possible HF patients and submitted for GPs' assessment in the audit stage. There were 310 patients with registered HF before GP assessment, of whom 146 (47.1%) were judged not to have HF by their GP (over‐registration). There were 538 patients with registered HF after GP assessment, of whom 374 (69.5%) did not have registered HF before GP assessment (under‐registration). The GBM and logistic regression model had a comparable predictive performance (area under the curve of 0.70 [95% confidence interval 0.65–0.77] and 0.69 [95% confidence interval 0.64–0.75], respectively). This was not significantly impacted by reducing the set of predictor variables to the 10 most important variables identified in the GBM model (free‐text and coded cardiomyopathy, ischaemic heart disease and atrial fibrillation, digoxin, mineralocorticoid receptor antagonists, and combinations of renin‐angiotensin system inhibitors and beta‐blockers with diuretics). This optimized query set was enough to identify 86% (n = 461/538) of GPs' self‐assessed HF population with a 33% reduction (n = 1537/4678) in screening caseload.

Conclusions

Diagnostic coding of HF in primary care health records is inaccurate with a high degree of under‐registration and over‐registration. An optimized query set enabled identification of more than 80% of GPs' self‐assessed HF population.

Assessing the severity of cardiovascular disease in 213 088 patients with coronary heart disease: a retrospective cohort study

Objective

Most current cardiovascular disease (CVD) risk stratification tools are for people without CVD, but very few are for prevalent CVD. In this study, we developed and validated a CVD severity score in people with coronary heart disease (CHD) and evaluated the association between severity and adverse outcomes.

Methods

Primary and secondary care data for 213 088 people with CHD in 398 practices in England between 2007 and 2017 were used. The cohort was randomly divided into training and validation datasets (80%/20%) for the severity model. Using 20 clinical severity indicators (each assigned a weight=1), baseline and longitudinal CVD severity scores were calculated as the sum of indicators. Adjusted Cox and competing-risk regression models were used to estimate risks for all-cause and cause-specific hospitalisation and mortality.

Results

Mean age was 64.5±12.7 years, 46% women, 16% from deprived areas, baseline severity score 1.5±1.2, with higher scores indicating a higher burden of disease. In the training dataset, 138 510 (81%) patients were hospitalised at least once, and 39 944 (23%) patients died. Each 1-unit increase in baseline severity was associated with 41% (95% CI 37% to 45%, area under the receiver operating characteristics (AUROC) curve=0.79) risk for 1 year for all-cause mortality; 59% (95% CI 52% to 67%, AUROC=0.80) for cardiovascular (CV)/diabetes mortality; 27% (95% CI 26% to 28%) for any-cause hospitalisation and 37% (95% CI 36% to 38%) for CV/diabetes hospitalisation. Findings were consistent in the validation dataset.

Conclusions

Higher CVD severity score is associated with higher risks for any-cause and cause-specific hospital admissions and mortality in people with CHD. Our reproducible score based on routinely collected data can help practitioners better prioritise management of people with CHD in primary care.

Electronic Medical Record-Based Case Phenotyping for the Charlson Conditions: Scoping Review

Background

Electronic medical records (EMRs) contain large amounts of rich clinical information. Developing EMR-based case definitions, also known as EMR phenotyping, is an active area of research that has implications for epidemiology, clinical care, and health services research.

Objective

This review aims to describe and assess the present landscape of EMR-based case phenotyping for the Charlson conditions.

Methods

A scoping review of EMR-based algorithms for defining the Charlson comorbidity index conditions was completed. This study covered articles published between January 2000 and April 2020, both inclusive. Embase (Excerpta Medica database) and MEDLINE (Medical Literature Analysis and Retrieval System Online) were searched using keywords developed in the following 3 domains: terms related to EMR, terms related to case finding, and disease-specific terms. The manuscript follows the Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA) guidelines.

Results

A total of 274 articles representing 299 algorithms were assessed and summarized. Most studies were undertaken in the United States (181/299, 60.5%), followed by the United Kingdom (42/299, 14.0%) and Canada (15/299, 5.0%). These algorithms were mostly developed either in primary care (103/299, 34.4%) or inpatient (168/299, 56.2%) settings. Diabetes, congestive heart failure, myocardial infarction, and rheumatology had the highest number of developed algorithms. Data-driven and clinical rule–based approaches have been identified. EMR-based phenotype and algorithm development reflect the data access allowed by respective health systems, and algorithms vary in their performance.

Conclusions

Recognizing similarities and differences in health systems, data collection strategies, extraction, data release protocols, and existing clinical pathways is critical to algorithm development strategies. Several strategies to assist with phenotype-based case definitions have been proposed.

Validity of Acute Cardiovascular Outcome Diagnoses Recorded in European Electronic Health Records: A Systematic Review

BACKGROUND

Electronic health records are widely used in cardiovascular disease research. We appraised the validity of stroke, acute coronary syndrome and heart failure diagnoses in studies conducted using European electronic health records.

METHODS

Using a prespecified strategy, we systematically searched seven databases from dates of inception to April 2019. Two reviewers independently completed study selection, followed by partial parallel data extraction and risk of bias assessment. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value estimates were narratively synthesized and heterogeneity between sensitivity and PPV estimates were assessed using I2.

RESULTS

We identified 81 studies, of which 20 validated heart failure diagnoses, 31 validated acute coronary syndrome diagnoses with 29 specifically recording estimates for myocardial infarction, and 41 validated stroke diagnoses. Few studies reported specificity or negative predictive value estimates. Sensitivity was ≤66% in all but one heart failure study, ≥80% for 91% of myocardial infarction studies, and ≥70% for 73% of stroke studies. PPV was ≥80% in 74% of heart failure, 88% of myocardial infarction, and 70% of stroke studies. PPV by stroke subtype was variable, at ≥80% for 80% of ischaemic stroke but only 44% of haemorrhagic stroke. There was considerable heterogeneity (I2 >75%) between sensitivity and PPV estimates for all diagnoses.

CONCLUSION

Overall, European electronic health record stroke, acute coronary syndrome and heart failure diagnoses are accurate for use in research, although validity estimates for heart failure and individual stroke subtypes were lower. Where possible, researchers should validate data before use or carefully interpret the results of previous validation studies for their own study purposes.

Development and validation of the DIabetes Severity SCOre (DISSCO) in 139 626 individuals with type 2 diabetes: a retrospective cohort study

OBJECTIVE

Clinically applicable diabetes severity measures are lacking, with no previous studies comparing their predictive value with glycated hemoglobin (HbA_1c). We developed and validated a type 2 diabetes severity score (the DIabetes Severity SCOre, DISSCO) and evaluated its association with risks of hospitalization and mortality, assessing its additional risk information to sociodemographic factors and HbA_1c.

RESEARCH DESIGN AND METHODS

We used UK primary and secondary care data for 139 626 individuals with type 2 diabetes between 2007 and 2017, aged ≥35 years, and registered in general practices in England. The study cohort was randomly divided into a training cohort (n=111 748, 80%) to develop the severity tool and a validation cohort (n=27 878). We developed baseline and longitudinal severity scores using 34 diabetes-related domains. Cox regression models (adjusted for age, gender, ethnicity, deprivation, and HbA_1c) were used for primary (all-cause mortality) and secondary (hospitalization due to any cause, diabetes, hypoglycemia, or cardiovascular disease or procedures) outcomes. Likelihood ratio (LR) tests were fitted to assess the significance of adding DISSCO to the sociodemographics and HbA_1c models.

RESULTS

A total of 139 626 patients registered in 400 general practices, aged 63±12 years were included, 45% of whom were women, 83% were White, and 18% were from deprived areas. The mean baseline severity score was 1.3±2.0. Overall, 27 362 (20%) people died and 99 951 (72%) had ≥1 hospitalization. In the training cohort, a one-unit increase in baseline DISSCO was associated with higher hazard of mortality (HR: 1.14, 95% CI 1.13 to 1.15, area under the receiver operating characteristics curve (AUROC)=0.76) and cardiovascular hospitalization (HR: 1.45, 95% CI 1.43 to 1.46, AUROC=0.73). The LR tests showed that adding DISSCO to sociodemographic variables significantly improved the predictive value of survival models, outperforming the added value of HbA_1c for all outcomes. Findings were consistent in the validation cohort.

CONCLUSIONS

Higher levels of DISSCO are associated with higher risks for hospital admissions and mortality. The new severity score had higher predictive value than the proxy used in clinical practice, HbA_1c. This reproducible algorithm can help practitioners stratify clinical care of patients with type 2 diabetes.

Quantifying outcome misclassification in multi-database studies: The case study of pertussis in the ADVANCE project

BACKGROUND

The Accelerated Development of VAccine beNefit-risk Collaboration in Europe (ADVANCE) is a public-private collaboration aiming to develop and test a system for rapid benefit-risk (B/R) monitoring of vaccines using European healthcare databases. Event misclassification can result in biased estimates. Using different algorithms for identifying cases of Bordetella pertussis (BorPer) infection as a test case, we aimed to describe a strategy to quantify event misclassification, when manual chart review is not feasible.

METHODS

Four participating databases retrieved data from primary care (PC) setting: BIFAP: (Spain), THIN and RCGP RSC (UK) and PEDIANET (Italy); SIDIAP (Spain) retrieved data from both PC and hospital settings. BorPer algorithms were defined by healthcare setting, data domain (diagnoses, drugs, or laboratory tests) and concept sets (specific or unspecified pertussis). Algorithm- and database-specific BorPer incidence rates (IRs) were estimated in children aged 0-14 years enrolled in 2012 and 2014 and followed up until the end of each calendar year and compared with IRs of confirmed pertussis from the ECDC surveillance system (TESSy). Novel formulas were used to approximate validity indices, based on a small set of assumptions. They were applied to approximately estimate positive predictive value (PPV) and sensitivity in SIDIAP.

RESULTS

The number of cases and the estimated BorPer IRs per 100,000 person-years in PC, using data representing 3,173,268 person-years, were 0 (IR = 0.0), 21 (IR = 4.3), 21 (IR = 5.1), 79 (IR = 5.7), and 2 (IR = 2.3) in BIFAP, SIDIAP, THIN, RCGP RSC and PEDIANET respectively. The IRs for combined specific/unspecified pertussis were higher than TESSy, suggesting that some false positives had been included. In SIDIAP the estimated IR was 45.0 when discharge diagnoses were included. The sensitivity and PPV of combined PC specific and unspecific diagnoses for BorPer cases in SIDIAP were approximately 85% and 72%, respectively.

CONCLUSION

Retrieving BorPer cases using only specific concepts has low sensitivity in PC databases, while including cases retrieved by unspecified concepts introduces false positives, which were approximately estimated to be 28% in one database. The share of cases that cannot be retrieved from a PC database because they are only seen in hospital was approximately estimated to be 15% in one database. This study demonstrated that quantifying the impact of different event-finding algorithms across databases and benchmarking with disease surveillance data can provide approximate estimates of algorithm validity.

Assessing the severity of Type 2 diabetes using clinical data-based measures: a systematic review

AIMS

To identify and critically appraise measures that use clinical data to grade the severity of Type 2 diabetes.

METHODS

We searched MEDLINE, Embase and PubMed between inception and June 2018. Studies reporting on clinical data-based diabetes-specific severity measures in adults with Type 2 diabetes were included. We excluded studies conducted solely in participants with other types of diabetes. After independent screening, the characteristics of the eligible measures including design and severity domains, the clinical utility of developed measures, and the relationship between severity levels and health-related outcomes were assessed.

RESULTS

We identified 6798 studies, of which 17 studies reporting 18 different severity measures (32 314 participants in 17 countries) were included: a diabetes severity index (eight studies, 44%); severity categories (seven studies, 39%); complication count (two studies, 11%); and a severity checklist (one study, 6%). Nearly 89% of the measures included diabetes-related complications and/or glycaemic control indicators. Two of the severity measures were validated in a separate study population. More severe diabetes was associated with increased healthcare costs, poorer cognitive function and significantly greater risks of hospitalization and mortality. The identified measures differed greatly in terms of the included domains. One study reported on the use of a severity measure prospectively.

CONCLUSIONS

Health records are suitable for assessment of diabetes severity; however, the clinical uptake of existing measures is limited. The need to advance this research area is fundamental as higher levels of diabetes severity are associated with greater risks of adverse outcomes. Diabetes severity assessment could help identify people requiring targeted and intensive therapies and provide a major benchmark for efficient healthcare services.

Correlation of miR-21 and BNP with pregnancy-induced hypertension complicated with heart failure and the diagnostic value

Correlation of miR-21 and B-type natriuretic peptide (BNP) with pregnancy-induced hypertension (PIH) complicated with heart failure and the diagnostic value was investigated. Sixty patients with PIH complicated with heart failure admitted to Affiliated Hospital of Chengde Medical University from July 2016 to July 2017 were enrolled as the experimental group, and 35 normal pregnant women as the control group. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) method was used to determine the expression level of plasma miR-21 expression level. An automatic biochemical analyzer was used to determine plasma BNP expression level. Spearmans correlation analysis was used for the correlation analysis of miR-21 and BNP. ROC curve was used for evaluating the diagnostic values of miR-21 and BNP for PIH complicated with heart failure. miR-21 and BNP expression levels were higher in patients with PIH complicated with heart failure than those in the normal individuals, and were increased in line with the heart failure grade (P<0.001). The plasma miR-21 expression was positively correlated with BNP in patients with PIH complicated with heart failure (r=0.685, P<0.001). Both miR-21 and BNP had higher diagnostic values for PIH complicated with heart failure, in the diagnosis, the best cut-off value [odds ratio (OR)] of miR-21 was 1.113, with an area under curve (AUC) of 0.889 and a 95% confidence interval (CI) of 82.05-95.76%; the OR of BNP was 123, with an AUC of 0.747 and a 95% CI of 64.95-84.38%. Blood pressure, six-minute walk test (6MWT), left ventricular ejection fraction (LVEF) and left ventricular end diastolic diameter (LVEDD) were independent risk factors for the occurrence of PIH complicated with heart failure (P<0.05). In conclusion, miR-21 and BNP, highly expressed in patients with PIH complicated with heart failure, are expected to become important biomarkers for diagnosing PIH complicated with heart failure and judging the degree of heart failure in the patients, and worthy of clinical popularization and application.

Patterns and trends of utilization of incretin-based medicines between 2008 and 2014 in three Italian geographic areas

BACKGROUND

The incretin-based medicines GLP1 analogues (GLP1a) and dipeptidyl peptidase-4 inhibitors (DPP4i) are hypoglycaemic agents licensed for the treatment of type 2 diabetes mellitus (T2DM). Although these drugs possess comparable efficacy and low risk of hypoglycaemia, differences in terms of route of administration (subcutaneous versus oral), effect on body weight and gastrointestinal tolerabily can impact their actual use in clinical practice. This study aimed to describe the real-world utilization of incretin-based medicines in the Italian clinical practice.

METHODS

A multi-database, population-based, descriptive, cohort study was performed using administrative data collected between 2008 and 2014 from three Italian geographic areas. Subjects aged ≥18 were selected. New users were defined as those with ≥1 dispensing of GLP1a or DPP4i during the year of interest and none in the past. Trends of cumulative annual incidence of use in the general adult population were observed. New users of GLP1a or DPP4i were respectively described in terms of demographic characteristics and use of antidiabetic drugs during 1 year before and after the first incretin dispensing.

RESULTS

The overall study population included 4,943,952 subjects. A total of 7357 new users of GLP1a and 41,907 of DPP4i were identified during the study period. Incidence of use increased between 2008 (0.2‰ for both GLP1a and DPP4i) and 2011 (GLP1a = 0.6‰; DPP4i = 2.5‰) and slightly decreased thereafter. In 2014, 61% of new GLP1a users received once-daily liraglutide while 52% of new DPP4i users received metformin/DPP4i in fixed-dose. The percentage of new DPP4i users older than 65 years of age increased from 30.9 to 62.6% during the study period. Around 12% of new users had not received any antidiabetic before starting an incretin.

CONCLUSIONS

During the study period, DPP4i rapidly became the most prescribed incretin-based medicine, particularly among older new user. The choice of the specific incretin-based medicine at first prescription appeared to be directed towards those with higher convenience of use (e.g. oral DPP4i rather than subcutaneous GLP1a, once-daily liraglutide rather than twice-daily exenatide). The non-negligibile use of incretin-based medicines as first-line pharmacotherapy for T2DM warrants further effectiveness and safety evaluations to better define their place in therapy.

Measurement Error and Misclassification in Electronic Medical Records: Methods to Mitigate Bias

PURPOSE OF REVIEW

We sought to: 1) examine common sources of measurement error in research using data from electronic medical records (EMR), 2) discuss methods to assess the extent and type of measurement error, and 3) describe recent developments in methods to address this source of bias.

RECENT FINDINGS

We identified eight sources of measurement error frequently encountered in EMR studies, the most prominent being that EMR data usually reflect only the health services and medications delivered within the specific health facility/system contributing to the EMR data. Methods for assessing measurement error in EMR data usually require gold standard or validation data, which may be possible using data linkage. Recent methodological developments to address the impact of measurement error in EMR analyses were particularly rich in the multiple imputation literature.

SUMMARY

Presently, sources of measurement error impacting EMR studies are still being elucidated, as are methods for assessing and addressing them. Given the magnitude of measurement error that has been reported, investigators are urged to carefully evaluate and rigorously address this potential source of bias in studies based in EMR data.

Epidemiology of acromegaly in Italy: analysis from a large longitudinal primary care database

PURPOSE

Epidemiological data are pivotal for the estimation of disease burden in populations.

AIM

Of the study was to estimate the incidence and prevalence of acromegaly in Italy along with the impact of comorbidities and hospitalization rates as compared to the general population.

METHODS

Retrospective epidemiological study (from 2000 to 2014) and case control-study. Data were extracted from the Health Search Database (HSD). HSD contains patient records from about 1000 general practitioners (GPs) throughout Italy, covering a population of more than 1 million patients. It includes information about patient demographics and medical data including clinical diagnoses and diagnostic tests.

RESULTS

At the end of the study period, 74 acromegaly patients (out of 1,066,871 people) were identified, resulting in a prevalence of 6.9 per 100,000 inhabitants [95% CI 5.4-8.5]. Prevalence was higher in females than men (p = 0.004), and showed a statistically significant trend of increase over time (p < 0.0001). Overall, incidence during the study period was 0.31 per 100,000 person-years. Hypertension and type II diabetes mellitus were the comorbidities more frequently associated with acromegaly (31.3 and 14.6%, respectively) and patients were more likely to undergo a high frequency of yearly hospitalization (≥3 accesses/year, p < 0.001) compared to sex-age matched controls.

CONCLUSIONS

This epidemiological study on acromegaly carried out using a large GP-based database, documented a disease prevalence of about 7 cases per 100,000 inhabitants. As expected, acromegaly was associated with a number of comorbidities (mainly hypertension and type II diabetes mellitus) and a high rate of patients' hospitalization.

Using electronic health records to quantify and stratify the severity of type 2 diabetes in primary care in England: rationale and cohort study design

INTRODUCTION

The increasing prevalence of type 2 diabetes mellitus (T2DM) presents a significant burden on affected individuals and healthcare systems internationally. There is, however, no agreed validated measure to infer diabetes severity from electronic health records (EHRs). We aim to quantify T2DM severity and validate it using clinical adverse outcomes.

METHODS AND ANALYSIS

Primary care data from the Clinical Practice Research Datalink, linked hospitalisation and mortality records between April 2007 and March 2017 for patients with T2DM in England will be used to develop a clinical algorithm to grade T2DM severity. The EHR-based algorithm will incorporate main risk factors (severity domains) for adverse outcomes to stratify T2DM cohorts by baseline and longitudinal severity scores. Provisionally, T2DM severity domains, identified through a systematic review and expert opinion, are: diabetes duration, glycated haemoglobin, microvascular complications, comorbidities and coprescribed treatments. Severity scores will be developed by two approaches: (1) calculating a count score of severity domains; (2) through hierarchical stratification of complications. Regression models estimates will be used to calculate domains weights. Survival analyses for the association between weighted severity scores and future outcomes-cardiovascular events, hospitalisation (diabetes-related, cardiovascular) and mortality (diabetes-related, cardiovascular, all-cause mortality)-will be performed as statistical validation. The proposed EHR-based approach will quantify the T2DM severity for primary care performance management and inform the methodology for measuring severity of other primary care-managed chronic conditions. We anticipate that the developed algorithm will be a practical tool for practitioners, aid clinical management decision-making, inform stratified medicine, support future clinical trials and contribute to more effective service planning and policy-making.

ETHICS AND DISSEMINATION

The study protocol was approved by the Independent Scientific Advisory Committee. Some data were presented at the National Institute for Health Research School for Primary Care Research Showcase, September 2017, Oxford, UK and the Diabetes UK Professional Conference March 2018, London, UK. The study findings will be disseminated in relevant academic conferences and peer-reviewed journals.

Diabetes management in the primary care setting: a comparison of physicians' performance by gender

BACKGROUND

A major shift in the gender of the medical-doctor workforce is now underway, and all over the world it is expected that an average 65% of the medical workforce will be women by 2030. In addition, an aging population means that chronic diseases, such as diabetes, are becoming more prevalent and the demand for care is rising. There is growing evidence of female physicians performing better than male physicians.AimOur study aimed to investigate whether any differences in diabetes process indicators are associated with gender, and/or the interaction between gender and different organizational models.Design and settingA population-based cross-sectional analysis was conducted on a large data set obtained by processing the public health administration databases of seven Italian local health units (LHUs). The seven LHUs, distributed all over the Italian peninsula in seven different regions, took part in a national project called MEDINA, with the focus on chronic disease management in primary care (PC).

METHODS

A total score was calculated for the average performance in the previously listed five indicators, representing global adherence to a quality management of patients with diabetes. A multilevel analysis was applied to see how LHUs affected the outcome. A quantile regression model was also fitted.

RESULTS

Our study included 2287 Italian general practitioners (586 of them female) caring for a total of 2 646 059 patients. Analyzing the performance scores confirmed that female general practitioners obtained better results than males. The differences between males and females were stronger on the 25th and 75th percentiles of the score than on the median values. The interaction between gender and LHU was not significant.

CONCLUSION

Our study evidenced that female physicians perform better than males in providing PC for diabetes independently by the different organizational models. Further research to understand the reasons for these gender differences is needed.

Monitoring compliance with standards of care for chronic diseases using healthcare administrative databases in Italy: Strengths and limitations

Background

A recent comprehensive report on healthcare quality in Italy published by the Organization of Economic Co-operation and Development (OECD) recommended that regular monitoring of quality of primary care by means of compliance with standards of care for chronic diseases is performed. A previous ecological study demonstrated that compliance with standards of care could be reliably estimated on regional level using administrative databases. This study compares estimates based on administrative data with estimates based on GP records for the same persons, to understand whether ecological fallacy played a role in the results of the previous study.

Methods

We compared estimates of compliance with diagnostic and therapeutic standards of care for type 2 diabetes (T2DM), hypertension and ischaemic heart disease (IHD) from administrative data (IAD) with estimates from medical records (MR) for the same persons registered with 24 GP’s in 2012. Data were linked at an individual level.

Results

32,688 persons entered the study, 12,673 having at least one of the three diseases according to at least one data source. Patients not detected by IAD were many, for all three conditions: adding MR increased the number of cases of T2DM, hypertension, and IHD by +40%, +42%, and +104%, respectively. IAD had imperfect sensitivity in detecting population compliance with therapies (adding MR increased the estimate, from +11.5% for statins to +14.7% for antithrombotics), and, more substantially, with diagnostic recommendations (adding MR increased the estimate, from +23.7% in glycated hemoglobin tests, to +50.5% in electrocardiogram). Patients not detected by IAD were less compliant with respect to those that IAD correctly identified (from -4.8 percentage points in proportion of IHD patients compliant with a yearly glycated hemoglobin test, to -40.1 points in the proportion of T2DM patients compliant with the same recommendation). IAD overestimated indicators of compliance with therapeutic standards (significant differences ranged from 3.3. to 3.6 percentage points) and underestimated indicators of compliance with diagnostic standards (significant differences ranged from -2.3 to -14.1 percentage points).

Conclusion

IAD overestimated the percentage of patients compliant with therapeutic standards by less than 6 percentage points, and underestimated the percentage of patients compliant with diagnostic standards by a maximum of 14 percentage points. Therefore, both discussions at local level between GP's and local health unit managers and discussions at central level between national and regional policy makers can be informed by indicators of compliance estimated by IAD, which, based on those results, have the ability of signalling critical or excellent clusters. However, this study found that estimates are partly flawed, because a high number of patients with chronic diseases are not detected by IAD, patients detected are not representative of the whole population of patients, and some categories of diagnostic tests are markedly underrecorded in IAD (up to 50% in the case of electrocardiograms). Those results call to caution when interpreting IAD estimates. Audits based on medical records, on the local level, and an interpretation taking into account information external to IAD, on the central level, are needed to assess a more comprehensive compliance with standards.

Health Data for Public Health: Towards New Ways of Combining Data Sources to Support Research Efforts in Europe

Objectives: To present the European landscape regarding the re-use of health administrative data for research. Methods: We present some collaborative projects and solutions that have been developed by Nordic countries, Italy, Spain, France, Germany, and the UK, to facilitate access to their health data for research purposes. Results: Research in public health is transitioning from siloed systems to more accessible and re-usable data resources. Following the example of the Nordic countries, several European countries aim at facilitating the re-use of their health administrative databases for research purposes. However, the ecosystem is still a complex patchwork, with different rules, policies, and processes for data provision. Conclusion: The challenges are such that with the abundance of health administrative data, only a European, overarching public health research infrastructure, is able to efficiently facilitate access to this data and accelerate research based on these highly valuable resources.