Systematic review and meta-analysis of validation studies on a diabetes case definition from health administrative records

A Capture-Recapture-based Ascertainment Probability Weighting Method for Effect Estimation With Under-ascertained Outcomes

Outcome under-ascertainment, characterized by the incomplete identification or reporting of cases, poses a substantial challenge in epidemiologic research. While capture-recapture methods can estimate unknown case numbers, their role in estimating exposure effects in observational studies is not well established. This paper presents an ascertainment probability weighting framework that integrates capture-recapture and propensity score weighting. We propose a nonparametric estimator of effects on binary outcomes that combines exposure propensity scores with data from two conditionally independent outcome measurements to simultaneously adjust for confounding and under-ascertainment. Demonstrating its practical application, we apply the method to estimate the relationship between health care work and coronavirus disease 2019 testing in a Swedish region. We find that ascertainment probability weighting greatly influences the estimated association compared to conventional inverse probability weighting, underscoring the importance of accounting for under-ascertainment in studies with limited outcome data coverage. We conclude with practical guidelines for the method's implementation, discussing its strengths, limitations, and suitable scenarios for application.

Incident Diabetes in Women With Patterns of Gestational Diabetes Occurrences Across 2 Pregnancies

Importance

Gestational diabetes is a type 2 diabetes risk indicator, and recurrence further augments risk. In women with a single occurrence across 2 pregnancies, it is unclear whether first- vs second-pregnancy gestational diabetes differ in terms of risk.

Objective

To compare the hazards of incident diabetes among those with gestational diabetes in the first, in the second, and in both pregnancies with women without gestational diabetes in either.

Design, Setting, and Participants

This was a retrospective cohort study with cohort inception from April 1, 1990, to December 31, 2012. Follow-up was April 1, 1990, to April 1, 2019. Participants were mothers with 2 singleton deliveries between April 1, 1990, and December 31, 2012, without diabetes before or between pregnancies, who were listed in public health care insurance administrative databases and birth, stillbirth, and death registries in Quebec, Canada. Data were analyzed from July to December 2023.

Exposure

Gestational diabetes occurrence(s) across 2 pregnancies.

Main outcomes and measures

Incident diabetes from the second delivery until a third pregnancy, death, or the end of the follow-up period, whichever occurred first.

Results

The 431 980 women with 2 singleton deliveries studied had a mean (SD) age of 30.1 (4.5) years at second delivery, with a mean (SD) of 2.8 (1.5) years elapsed between deliveries; 373 415 (86.4%) were of European background, and 78 770 (18.2%) were at the highest quintile of material deprivation. Overall, 10 920 women (2.5%) had gestational diabetes in their first pregnancy, 16 145 (3.7%) in their second, and 8255 (1.9%) in both (12 205 incident diabetes events; median [IQR] follow-up 11.5 [5.3-19.4] years). First pregnancy-only gestational diabetes increased hazards 4.35-fold (95% CI, 4.06-4.67), second pregnancy-only increased hazards 7.68-fold (95% CI, 7.31-8.07), and gestational diabetes in both pregnancies increased hazards 15.8-fold (95% CI, 15.0-16.6). Compared with first pregnancy-only gestational diabetes, second pregnancy-only gestational diabetes increased hazards by 76% (95% CI, 1.63-1.91), while gestational diabetes in both pregnancies increased it 3.63-fold (95% CI, 3.36-3.93).

Conclusions and relevance

In this retrospective cohort study of nearly half a million women with 2 singleton pregnancies, both the number and ordinal pregnancy of any gestational diabetes occurrence increased diabetes risk. These considerations offer greater nuance than an ever or never gestational diabetes dichotomy.

Influence of diabetes and other risk factors on in-hospital mortality following kidney transplantation: an analysis of the Spanish National Hospital Discharge Database from 2016 to 2020

INTRODUCTION

To assess time trends in incidence, clinical characteristics, complications, and hospital outcomes among patients with type 1 diabetes (T1D), with type 2 diabetes (T2D), and patients without diabetes who underwent kidney transplant (KT); to identify variables associated with in-hospital mortality (IHM); and to determine the impact of the COVID-19 pandemic.

RESEARCH DESIGN AND METHODS

We used a nationwide discharge database to select KT recipients admitted to Spanish hospitals from 2016 to 2020. We stratified patients according to diabetes status. We used multivariable logistic regression to identify the variables associated with IHM.

RESULTS

A total of 14 594 KTs were performed in Spain (T2D, 22.28%; T1D, 3.72%). The number of KTs rose between 2016 and 2019 and and decreased from 2019 to 2020 in all groups. In patients with T2D, the frequency of KT complications increased from 21.08% in 2016 to 34.17% in 2020 (p<0.001). Patients with T2D had significantly more comorbidity than patients with T1D and patients without diabetes (p<0.001). Patients with T1D experienced KT rejection significantly more frequently (8.09%) than patients with T2D (5.57%).COVID-19 was recorded in 26 out of the 2444 KTs performed in 2020, being found in 6 of the 39 patients deceased that year (15.38%) and in 0.83% of the survivors.The variables associated with IHM were comorbidity and complications of KT. The presence of T1D was associated with IHM (OR 2.6; 95% CI 1.36 to 5.16) when patients without diabetes were the reference category. However, T2D was not associated with a higher IHM (OR 0.86; 95% CI 0.61 to 1.2).

CONCLUSIONS

The COVID-19 pandemic led to a decrease in the number of transplants. Patients with T1D have more rejection of the transplanted organ than patients with T2D. Fewer women with T2D undergo KT. The presence of T1D is a risk factor for IHM.

Risk Factors for Type 2 Diabetes in the Multiethnic Cohort

OBJECTIVES

In this report, we investigated the association between established risk factors and type 2 diabetes (T2D) across 5 distinct ethnic groups and explored differences according to T2D definition within the Multiethnic Cohort (MEC) Study.

METHODS

Using the full MEC, with participants in Hawaii and Los Angeles (N=172,230), we applied Cox regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). All participants completed questionnaires asking about demographics, anthropometrics, lifestyle factors, and regular diet. T2D status was determined from self-reported diagnosis/medication and Medicare claims. We assessed the associations between well-established risk factors and T2D in the full cohort, after stratification by ethnic group, according to the T2D definition, and in a biorepository subset. Effect modification by ethnicity was evaluated using Wald's tests.

RESULTS

Overall, 46,500 (27%) participants had an incident T2D diagnosis after a mean follow-up of 17.1±6.9 years. All predictors were significantly associated with T2D: overweight (HR=1.74), obesity (HR=2.90), red meat intake (HR=1.15), short (HR=1.04) and long (HR=1.08) sleep duration, and smoking (HR=1.26) predicted a significantly higher T2D incidence, whereas coffee (HR=0.90) and alcohol (HR=0.78) consumption, physical activity (HR=0.89), and diet quality (HR=0.96) were associated with lower T2D incidence. The strength of these associations was similar across ethnic groups with noteworthy disparities for overweight/obesity, physical activity, alcohol intake, coffee consumption, and diet quality.

CONCLUSIONS

These findings confirm the importance of known risk factors for T2D across ethnic groups, but small differences were detected that may contribute to disparate incidence rates in some ethnic groups, especially for obesity and physical activity.

Estimating the completeness of physician billing claims for diabetes case ascertainment: a multiprovince investigation

INTRODUCTION

Previous research has suggested that how physicians are paid may affect the completeness of billing claims for estimating chronic disease. The purpose of this study is to estimate the completeness of physician billings for diabetes case ascertainment.

METHODS

We used administrative data from eight Canadian provinces covering the period 1 April 2014 to 31 March 2016. The patient cohort was stratified into two mutually exclusive groups based on their physician remuneration type: fee-for-service (FFS), for those paid only on that basis; and non-fee-for-service (NFFS). Using diabetes prescription drug data as our reference data source, we evaluated whether completeness of disease case ascertainment varied with payment type. Diabetes incidence rates were then adjusted for completeness of ascertainment.

RESULTS

The cohort comprised 86 110 patients. Overall, equal proportions received their diabetes medications from FFS and NFFS physicians. Overall, physician payment method had little impact upon the percentage of missed diabetes cases (FFS, 14.8%; NFFS, 12.2%). However, the difference in missed cases between FFS and NFFS varied widely by province, ranging from -1.0% in Nova Scotia to 29.9% in Newfoundland and Labrador. The difference between the observed and adjusted disease incidence rates also varied by province, ranging from 22% in Prince Edward Island to 4% in Nova Scotia.

CONCLUSION

The difference in the loss of cases by physician remuneration method varied across jurisdictions. This loss may contribute to an underestimation of disease incidence. The method we used could be applied to other chronic diseases for which drug therapy could serve as reference data source.

Considering gestational diabetes and gestational hypertension history across two pregnancies in relationship to cardiovascular disease development: A retrospective cohort study

AIMS

Gestational diabetes (GDM) and hypertension (GHTN) occurrences signal elevated cardiovascular disease (CVD) risk. There is little study of occurrence and recurrence of these conditions in relationship to CVD. Among women with two singleton pregnancies, we aimed to quantify CVD risk in relationship to the number of GDM/GHTN occurrences.

METHODS

In this Quebec-based retrospective cohort study (n = 431,980), we ascertained the number of GDM/GHTN occurrences over two pregnancies and assessed for CVD over a median of 16.4 years (cohort inception 1990-2012, outcomes 1990-2019). We defined CVD as a composite of myocardial infarction, stroke, and angina, requiring hospitalization and/or causing death. We adjusted Cox proportional hazards models for offspring size, preterm/term birth status, maternal age group, time between deliveries, ethnicity, deprivation level, and co-morbid conditions.

RESULTS

Compared to absence of GDM/GHTN in either pregnancy, one GDM/GHTN occurrence increased CVD hazards by 47% (hazard ratio [HR] = 1.47, 95% confidence interval [CI] 1.35-1.61), two occurrences nearly doubled hazards (HR = 1.91, 95% CI 1.68-2.17), and three or more approximately tripled CVD hazards (HR = 2.93, 95% CI 2.20-3.90). Individual components of the composite demonstrated similar findings.

CONCLUSIONS

Health care providers and mothers should consider a complete history of GDM/GHTN occurrences to ascertain the importance and urgency of preventive action.

Prevalence of Dementia among Patients Hospitalized with Type 2 Diabetes Mellitus in Spain, 2011-2020: Sex-Related Disparities and Impact of the COVID-19 Pandemic

(1) Background: To assess changes in the prevalence of dementia among patients hospitalized with type 2 diabetes (T2DM), to analyze the effects of dementia on in-hospital mortality (IHM) in this population, to evaluate sex differences, and to determine the impact of the COVID-19 pandemic on these parameters. (2) Methods: We used a nationwide discharge database to select all patients with T2DM aged 60 years or over admitted to Spanish hospitals from 2011 to 2020. We identified those with all-cause dementia, Alzheimer's disease (AD), and vascular dementia (VaD). The effect of sex, age, comorbidity, and COVID-19 on the prevalence of dementia subtypes and on IHM was assessed using multivariable logistic regression. (3) Results: We identified 5,250,810 hospitalizations with T2DM. All-cause dementia was detected in 8.31%, AD in 3.00%, and VaD in 1.55%. The prevalence of all subtypes of dementia increased significantly over time. After multivariable adjustment, higher values were observed in women for all-cause dementia (OR 1.34; 95% CI 1.33-1.35), AD (OR 1.6; 95% CI 1.58-1.62), and VaD (OR 1.12; 95% CI 1.11-1.14). However, female sex was a protective factor for IHM in patients with all-cause dementia (OR 0.90; 95% CI 0.89-0.91), AD (OR 0.89; 95% CI 0.86-0.91), and VaD (OR 0.95; 95% CI 0.91-0.99). IHM among patients with dementia remained stable over time, until 2020, when it increased significantly. Higher age, greater comorbidity, and COVID-19 were associated with IHM in all dementia subtypes. (4) Conclusions: The prevalence of dementia (all-cause, AD, and VaD) in men and women with T2DM increased over time; however, the IHM remained stable until 2020, when it increased significantly, probably because of the COVID-19 pandemic. The prevalence of dementia is higher in women than in men, although female sex is a protective factor for IHM.

Prevalence and associated factors of diabetes mellitus among patients with tuberculosis in South Korea from 2011 to 2018: a nationwide cohort study

OBJECTIVES

This study aimed to identify the prevalence of diabetes mellitus (DM) among patients with tuberculosis (TB) using a nationwide cohort in South Korea.

DESIGN

A retrospective cohort study.

SETTING

This study used the Korean Tuberculosis and Post-Tuberculosis cohort, which was constructed by linking the Korean National Tuberculosis Surveillance, National Health Information Database (NHID) and Statistics Korea data for the causes of death.

PARTICIPANTS

During the study period, all notified patients with TB with at least one claim in the NHID were included. Exclusion criteria were age less than 20 years, drug resistance, initiation of TB treatment before the study period and missing values in covariates.

OUTCOME MEASURES

DM was defined as having at least two claims of the International Classification of Diseases (ICD) code for DM or at least one claim of the ICD code for DM and prescription of any antidiabetic drugs. Newly diagnosed DM (nDM) and previously diagnosed DM (pDM) were defined as DM diagnosed after and before TB diagnosis, respectively.

RESULTS

A total of 26.8% (70 119) of patients were diagnosed with DM. The age-standardised prevalence increased as age increased or income decreased. Patients with DM were more likely to be men, older, had the lowest income group, had more acid-fast bacilli smear and culture positivity, had a higher Charlson Comorbidity Index score and had more comorbidities compared with patients without DM. Approximately 12.5% (8823) patients had nDM and 87.4% (61 296) had pDM among those with TB-DM.

CONCLUSIONS

The prevalence of DM among patients with TB was considerably high in Korea. To achieve the goal of TB control and improve the health outcomes of both TB and DM, integrated screening of TB and DM and care delivery in clinical practice are necessary.

PhenoDEF: a corpus for annotating sentences with information of phenotype definitions in biomedical literature

Background

Adverse events induced by drug-drug interactions are a major concern in the United States. Current research is moving toward using electronic health record (EHR) data, including for adverse drug events discovery. One of the first steps in EHR-based studies is to define a phenotype for establishing a cohort of patients. However, phenotype definitions are not readily available for all phenotypes. One of the first steps of developing automated text mining tools is building a corpus. Therefore, this study aimed to develop annotation guidelines and a gold standard corpus to facilitate building future automated approaches for mining phenotype definitions contained in the literature. Furthermore, our aim is to improve the understanding of how these published phenotype definitions are presented in the literature and how we annotate them for future text mining tasks.

Results

Two annotators manually annotated the corpus on a sentence-level for the presence of evidence for phenotype definitions. Three major categories (inclusion, intermediate, and exclusion) with a total of ten dimensions were proposed characterizing major contextual patterns and cues for presenting phenotype definitions in published literature. The developed annotation guidelines were used to annotate the corpus that contained 3971 sentences: 1923 out of 3971 (48.4%) for the inclusion category, 1851 out of 3971 (46.6%) for the intermediate category, and 2273 out of 3971 (57.2%) for exclusion category. The highest number of annotated sentences was 1449 out of 3971 (36.5%) for the “Biomedical & Procedure” dimension. The lowest number of annotated sentences was 49 out of 3971 (1.2%) for “The use of NLP”. The overall percent inter-annotator agreement was 97.8%. Percent and Kappa statistics also showed high inter-annotator agreement across all dimensions.

Conclusions

The corpus and annotation guidelines can serve as a foundational informatics approach for annotating and mining phenotype definitions in literature, and can be used later for text mining applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13326-022-00272-6.

Interpreting global trends in type 2 diabetes complications and mortality

International trends in traditional diabetes complications (cardiovascular, renal, peripheral vascular, ophthalmic, hepatic or neurological diseases) and mortality rates are poorly characterised. An earlier review of studies published up to 2015 demonstrated that most data come from a dozen high-income countries (HICs) in North America, Europe or the Asia-Pacific region and that, in these countries at least, rates of acute glycaemic fluctuations needing medical attention and amputations, myocardial infarction and mortality were all declining over the period. Here, we provide an updated review of published literature on trends in type 2 diabetes complications and mortality in adults since 2015. We also discuss issues related to data collection, analysis and reporting that have influenced global trends in type 2 diabetes and its complications. We found that most data on trends in type 2 diabetes, its complications and mortality come from a small number of HICs with comprehensive surveillance systems, though at least some low- and middle-income countries (LMICs) from Africa and Latin America are represented in this review. The published data suggest that HICs have experienced declines in cardiovascular complication rates and all-cause mortality in people with diabetes. In parallel, cardiovascular complications and mortality rates in people with diabetes have increased over time in LMICs. However, caution is warranted in interpreting trends from LMICs due to extremely sparse data or data that are not comparable across countries. We noted that approaches to case ascertainment and definitions of complications and mortality (numerators) and type 2 diabetes (the denominator) vary widely and influence the interpretation of international data. We offer four key recommendations to more rigorously document trends in rates of type 2 diabetes complications and mortality, over time and worldwide: (1) increasing investments in data collection systems; (2) standardising case definitions and approaches to ascertainment; (3) strengthening analytical capacity; and (4) developing and implementing structured guidelines for reporting of data.

Analysis of the relationship between surgeon procedure volume and complications after total knee arthroplasty using a propensity-matched cohort study

Objectives

This study aimed to identify a threshold in annual surgeon volume associated with increased risk of revision (for any cause) and deep infection requiring surgery following primary elective total knee arthroplasty (TKA).

Design

A propensity score matched cohort study.

Setting

Ontario, Canada.

Participants

169 713 persons who received a primary TKA between 2002 and 2016, with 3-year postoperative follow-up.

Main outcome measures

Revision arthroplasty (for any cause), and the occurrence of deep surgical infection requiring surgery.

Results

Based on restricted cubic spline analysis, the threshold for increased probability of revision and deep infection requiring surgery was <70 cases/year. After matching of 51 658 TKA recipients from surgeons performing <70 cases/year to TKA recipients from surgeons with greater than 70 cases/year, patients in the former group had a higher rate of revision (for any cause, 2.23% (95% Confidence Interval (CI) 1.39 to 3.07) vs 1.70% (95% CI 0.85 to 2.55); Hazard Ratio (HR) 1.33, 95% CI 1.21 to 1.47, p<0.0001) and deep infection requiring surgery (1.29% (95% CI 0.44 to 2.14) vs 1.09% (95% CI 0.24 to 1.94); HR 1.33, 95% CI 1.17 to 1.51, p<0.0001).

Conclusions

For primary TKA recipients, cases performed by surgeons who had performed fewer than 70 TKAs in the year prior to the index TKA were at 31% increased relative risk of revision (for any cause), and 18% increased relative risk for deep surgical infection requiring surgery, at 3-year follow-up.

Research using population-based administration data integrated with longitudinal data in child protection settings: A systematic review

INTRODUCTION

Over the past decade there has been a marked growth in the use of linked population administrative data for child protection research. This is the first systematic review of studies to report on research design and statistical methods used where population-based administrative data is integrated with longitudinal data in child protection settings.

METHODS

The systematic review was conducted according to Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) statement. The electronic databases Medline (Ovid), PsycINFO, Embase, ERIC, and CINAHL were systematically searched in November 2019 to identify all the relevant studies. The protocol for this review was registered and published with Open Science Framework (Registration DOI: 10.17605/OSF.IO/96PX8).

RESULTS

The review identified 30 studies reporting on child maltreatment, mental health, drug and alcohol abuse and education. The quality of almost all studies was strong, however the studies rated poorly on the reporting of data linkage methods. The statistical analysis methods described failed to take into account mediating factors which may have an indirect effect on the outcomes of interest and there was lack of utilisation of multi-level analysis.

CONCLUSION

We recommend reporting of data linkage processes through following recommended and standardised data linkage processes, which can be achieved through greater co-ordination among data providers and researchers.

Self-reported infertility diagnoses and treatment history approximately 20 years after fertility treatment initiation

Background

Infertility history may have important implications for clinical practice and scientific discovery. Previous research on the validity of self-reported infertility measurements has been limited in scope and duration (< 5 years). In this study, we validated self-reported infertility history measures 15–23 years after fertility treatment initiation among women who utilized assisted reproductive technology (ART).

Methods

Women who received ART treatments from three Boston infertility clinics and who enrolled in a prior study (1994–2003) were re-contacted in 2018 for the AfteR Treatment Follow-up Study (ART-FS). Infertility history was collected from clinical records and two self-report questionnaires (at ART initiation and at ART-FS enrollment). Treatment history included specific details (fresh or frozen embryo transfers, number of cycles) and treatment recall prior to ART initiation. Self-reported infertility diagnoses included polycystic ovary syndrome (PCOS), endometriosis, uterine factor infertility, tubal factor infertility, diminished ovarian reserve/advanced maternal age, male factor infertility, and other/unknown. We compared self-reported measures from 2018 to self-reported and clinical data from prior study initiation, using Cohen’s kappa, sensitivity, specificity, and 95% confidence intervals.

Results

Of 2644 women we attempted to recontact, 808 completed the ART-FS, with an average follow-up of 19.6 years (standard deviation: 2.7). Recall of fertility treatment usage had moderate sensitivity (IVF = 0.85, Clomiphene/Gonadotropin = 0.81) but low specificity across different infertility treatment modalities (IVF = 0.63, Clomiphene/Gonadotropin = 0.55). Specific IVF details had low to moderate validity and reliability with clinical records. Reliability of recalled infertility diagnosis was higher when compared to self-report at ART initiation (PCOS K = 0.66, Endometriosis K = 0.76, Tubal K = 0.73) than when compared to clinical records (PCOS K = 0.31, Endometriosis K = 0.48, Tubal K = 0.62) and varied by diagnosis.

Conclusions

The ability of women to recall specific IVF treatment details was moderately accurate and recall of self-reported infertility diagnosis varied by diagnosis and measurement method.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40738-021-00099-2.

A primer on quantitative bias analysis with positive predictive values in research using electronic health data

OBJECTIVE

In health informatics, there have been concerns with reuse of electronic health data for research, including potential bias from incorrect or incomplete outcome ascertainment. In this tutorial, we provide a concise review of predictive value-based quantitative bias analysis (QBA), which comprises epidemiologic methods that use estimates of data quality accuracy to quantify the bias caused by outcome misclassification.

TARGET AUDIENCE

Health informaticians and investigators reusing large, electronic health data sources for research.

SCOPE

When electronic health data are reused for research, validation of outcome case definitions is recommended, and positive predictive values (PPVs) are the most commonly reported measure. Typically, case definitions with high PPVs are considered to be appropriate for use in research. However, in some studies, even small amounts of misclassification can cause bias. In this tutorial, we introduce methods for quantifying this bias that use predictive values as inputs. Using epidemiologic principles and examples, we first describe how multiple factors influence misclassification bias, including outcome misclassification levels, outcome prevalence, and whether outcome misclassification levels are the same or different by exposure. We then review 2 predictive value-based QBA methods and why outcome PPVs should be stratified by exposure for bias assessment. Using simulations, we apply and evaluate the methods in hypothetical electronic health record-based immunization schedule safety studies. By providing an overview of predictive value-based QBA, we hope to bridge the disciplines of health informatics and epidemiology to inform how the impact of data quality issues can be quantified in research using electronic health data sources.

Incidence and prevalence of hypoglycaemia in type 1 and type 2 diabetes individuals: A systematic review and meta-analysis

BACKGROUND

Previous meta-analysis investigating the incidence and prevalence of hypoglycaemia in both types of diabetes is limited. The purpose of this review is to conduct a systematic review and meta-analysis of the existing literature which investigates the incidence and prevalence of hypoglycaemia in individuals with diabetes.

METHODS

PubMed, Embase and Cochrane library databases were searched up to October 2018. Observational studies including individuals with diabetes of all ages and reporting incidence and/or prevalence of hypoglycaemia were included. Two reviewers independently screened articles, extracted data and assessed the quality of included studies. Meta-analysis was performed using a random effects model with 95% confidence interval (CI) to estimate the pooled incidence and prevalence of hypoglycaemia in individuals with diabetes.

RESULTS

Our search strategy generated 35,007 articles, of which 72 studies matched the inclusion criteria and were included in the meta-analysis. The prevalence of hypoglycaemia ranged from 0.074% to 73.0%, comprising a total of 2,462,810 individuals with diabetes. The incidence rate of hypoglycaemia ranged from 0.072 to 42,890 episodes per 1,000 person-years: stratified by type of diabetes, it ranged from 14.5 to 42,890 episodes per 1,000 person-years and from 0.072 to 16,360 episodes per 1,000-person years in type 1 and type 2 diabetes, respectively.

CONCLUSION

Hypoglycaemia is very common among individuals with diabetes. Further studies are needed to investigate hypoglycaemia-associated risk factors.

Geospatial patterns of comorbidity prevalence among people with osteoarthritis in Alberta Canada

BACKGROUND

Knowledge of geospatial pattern in comorbidities prevalence is critical to an understanding of the local health needs among people with osteoarthritis (OA). It provides valuable information for targeting optimal OA treatment and management at the local level. However, there is, at present, limited evidence about the geospatial pattern of comorbidity prevalence in Alberta, Canada.

METHODS

Five administrative health datasets were linked to identify OA cases and comorbidities using validated case definitions. We explored the geospatial pattern in comorbidity prevalence at two standard geographic areas levels defined by the Alberta Health Services: descriptive analysis at rural-urban continuum level; spatial analysis (global Moran's I, hot spot analysis, cluster and outlier analysis) at the local geographic area (LGA) level. We compared area-level indicators in comorbidities hotspots to those in the rest of Alberta (non-hotspots).

RESULTS

Among 359,638 OA cases in 2013, approximately 60% of people resided in Metro and Urban areas, compared to 2% in Rural Remote areas. All comorbidity groups exhibited statistically significant spatial autocorrelation (hypertension: Moran's I index 0.24, z score 4.61). Comorbidity hotspots, except depression, were located primarily in Rural and Rural Remote areas. Depression was more prevalent in Metro (Edmonton-Abbottsfield: 194 cases per 1000 population, 95%CI 192-195) and Urban LGAs (Lethbridge-North: 169, 95%CI 168-171) compared to Rural areas (Fox Creek: 65, 95%CI 63-68). Comorbidities hotspots included a higher percentage of First Nations or Inuit people. People with OA living in hotspots had lower socioeconomic status and less access to care compared to non-hotspots.

CONCLUSIONS

The findings highlight notable rural-urban disparities in comorbidities prevalence among people with OA in Alberta, Canada. Our study provides valuable evidence for policy and decision makers to design programs that ensure patients with OA receive optimal health management tailored to their local needs and a reduction in current OA health disparities.

A weighting method for simultaneous adjustment for confounding and joint exposure-outcome misclassifications

Joint misclassification of exposure and outcome variables can lead to considerable bias in epidemiological studies of causal exposure-outcome effects. In this paper, we present a new maximum likelihood based estimator for marginal causal effects that simultaneously adjusts for confounding and several forms of joint misclassification of the exposure and outcome variables. The proposed method relies on validation data for the construction of weights that account for both sources of bias. The weighting estimator, which is an extension of the outcome misclassification weighting estimator proposed by Gravel and Platt (Weighted estimation for confounded binary outcomes subject to misclassification. Stat Med 2018; 37: 425–436), is applied to reinfarction data. Simulation studies were carried out to study its finite sample properties and compare it with methods that do not account for confounding or misclassification. The new estimator showed favourable large sample properties in the simulations. Further research is needed to study the sensitivity of the proposed method and that of alternatives to violations of their assumptions. The implementation of the estimator is facilitated by a new R function (ipwm) in an existing R package (mecor).

Validation of acute exacerbation of chronic obstructive pulmonary disease (COPD) recording in electronic health records: a systematic review protocol

INTRODUCTION

Many patients with chronic obstructive pulmonary disease (COPD) experience a sustained worsening in symptoms termed an acute exacerbation (AECOPD). AECOPDs impact on patients' quality of life and lung function, are costly to health services and are an important topic for research. Electronic health records (EHR) are increasingly being used to study AECOPD, requiring accurate detection of AECOPD in EHRs to ensure generalisable results. The aim of this protocol is to provide an overview of studies that validate AECOPD definitions used in EHRs and administrative claims databases.

METHODS AND ANALYSIS

Medline and Embase will be searched for terms related to COPD exacerbation, EHRs and validation. All studies published between 1 January 1990 and 30 September 2019 written in English that validate AECOPD in EHRs and administrative claims databases will be considered.

INCLUSION CRITERIA

EHR data must be routinely collected; the AECOPD detection algorithm must be compared against a reference standard; and a measure of validity must be calculable. Two independent reviewers will screen articles for inclusion, extract study details and assess risk of bias using QUADAS-2. Disagreements will be resolved by consensus or arbitration by a third reviewer. This protocol has been developed in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols checklist.

ETHICS AND DISSEMINATION

This will be a review of previously published literature therefore no ethical approval is required. Results from this review will be published in a peer-reviewed journal. The results can be used in future research to identify occurrences of AECOPD.

PROSPERO REGISTRATION NUMBER

CRD42019130863.

Improving measurement of binary covariates in claims data: A simulation study

PURPOSE

When investigators have two claims-based definitions for a binary confounder, it is unclear whether to prefer the more sensitive or more specific definition. Our objective was to compare adjusting for the sensitive or specific definition alone vs two novel approaches combining both definitions: a "two-algorithm indicator" and a "two-algorithm restriction" approach.

METHODS

Each simulated patient had a binary exposure, outcome, and confounder. We created two nested, misclassified versions of the confounder using validated heart failure definitions. The sensitive definition had a sensitivity/specificity of 0.98/0.83, while the specific definition had a sensitivity/specificity of 0.77/0.99. Patients were classified into 3 groups: group 0 did not meet either definition, group 1 met the sensitive but not specific definition, and group 2 met both. The two-algorithm indicator approach adjusted using indicators for groups 1 and 2, while the two-algorithm restriction approach excluded patients in group 1 and adjusted using an indicator for group 2. Adjusted exposure odds ratios (ORs) were estimated for each approach using logistic regression.

RESULTS

The crude OR was 1.33 (95% CI, 1.07-1.63). Adjusting for the specific or sensitive definitions resulted in ORs of 1.09 (95% CI, 0.87-1.35) and 1.14 (95% CI, 0.91-1.40). The two-algorithm indicator method returned an OR of 1.07 (95% CI, 0.86-1.33). The two-algorithm restriction approach returned an OR of 1.02 (95% CI, 0.79-1.29) but excluded 20% of the cohort.

CONCLUSIONS

The two-algorithm indicator approach may improve adjustment for claims-based confounders by returning a point estimate at least as unbiased as the better of the two component definitions.

Gestational diabetes associated with incident diabetes in childhood and youth: a retrospective cohort study

BACKGROUND

Indicators of childhood- and youth-onset diabetes may be useful for early detection of diabetes; there is a known association between composite exposure of parental type 2 diabetes and gestational diabetes mellitus with childhood- and youth-onset diabetes. We examined associations between gestational diabetes mellitus and incidence of childhood- and youth-onset diabetes in offspring.

METHODS

Using public health insurance administrative databases from Quebec, Canada, we randomly selected singleton live births with maternal gestational diabetes mellitus (1990-2007) and matched them 1:1 with singleton live births without gestational diabetes mellitus. Follow-up was to Mar. 31, 2012. We examined associations of diabetes in offspring with maternal gestational diabetes mellitus through unadjusted and adjusted Cox proportional hazards models. In secondary analyses, we separately considered age groups ranging from birth to age 12 years, and age 12 to 22 years.

RESULTS

Incidence of pediatric diabetes (per 10 000 person-years) was higher in offspring born to mothers with gestational diabetes mellitus (4.52, 95% confidence interval [CI] 4.47-4.57) than in mothers without gestational diabetes mellitus (2.4, 95% CI 2.37-2.46). In an adjusted Cox proportional hazards model, maternal gestational diabetes mellitus was associated with development of pediatric diabetes overall (birth to age 22 yr: hazard ratio [HR] 1.77, 95% CI 1.41-2.22), during childhood (birth to age 12 yr: HR 1.43, 95% CI 1.09-1.89), and in youth (age 12 to 22 yr: HR 2.53, 95% CI 1.67-3.85).

INTERPRETATION

Gestational diabetes mellitus is associated with incident diabetes in offspring during childhood and adolescence. Future studies are needed to examine longer-term outcomes in patients with pediatric diabetes with a maternal history of gestational diabetes mellitus, to ascertain how they compare with other patients with childhood- or youth-onset diabetes, in terms of disease severity and outcomes.

Diabetes diagnosis from administrative claims and estimation of the true prevalence of diabetes among 4.2 million individuals of the Veneto region (North East Italy)

BACKGROUND AND AIMS

Diabetes can often remain undiagnosed or unregistered in administrative databases long after its onset, even when laboratory test results meet diagnostic criteria. In the present work, we analyse healthcare data of the Veneto Region, North East Italy, with the aims of: (i) developing an algorithm for the identification of diabetes from administrative claims (4,236,007 citizens), (ii) assessing its reliability by comparing its performance with the gold standard clinical diagnosis from a clinical database (7525 patients), (iii) combining the algorithm and the laboratory data of the regional Health Information Exchange (rHIE) system (543,520 subjects) to identify undiagnosed diabetes, and (iv) providing a credible estimate of the true prevalence of diabetes in Veneto.

METHODS AND RESULTS

The proposed algorithm for the identification of diabetes was fed by administrative data related to drug dispensations, outpatient visits, and hospitalisations. Evaluated against a clinical database, the algorithm achieved 95.7% sensitivity, 87.9% specificity, and 97.6% precision. To identify possible cases of undiagnosed diabetes, we applied standard diagnostic criteria to the laboratory test results of the subjects who, according to the algorithm, had no diabetes-related claims. Using a simplified probabilistic model, we corrected our claims-based estimate of known diabetes (6.17% prevalence; 261,303 cases) to account for undiagnosed cases, yielding an estimated total prevalence of 7.50%.

CONCLUSION

We herein validated an algorithm for the diagnosis of diabetes using administrative claims against the clinical diagnosis. Together with rHIE laboratory data, this allowed to identify possibly undiagnosed diabetes and estimate the true prevalence of diabetes in Veneto.

Existing comorbidities in people with osteoarthritis: a retrospective analysis of a population-based cohort in Alberta, Canada

OBJECTIVES

The purpose of this study is to estimate the prevalence of comorbidities among people with osteoarthritis (OA) using administrative health data.

DESIGN

Retrospective cohort analysis.

SETTING

All residents in the province of Alberta, Canada registered with the Alberta Health Care Insurance Plan population registry.

PARTICIPANTS

497 362 people with OA as defined by 'having at least one OA-related hospitalization, or at least two OA-related physician visits or two ambulatory care visits within two years'.

PRIMARY OUTCOME MEASURES

We selected eight comorbidities based on literature review, clinical consultation and the availability of validated case definitions to estimate their frequencies at the time of diagnosis of OA. Sex-stratified age-standardised prevalence rates per 1000 population of eight clinically relevant comorbidities were calculated using direct standardisation with 95% CIs. We applied χ² tests of independence with a Bonferroni correction to compare the percentage of comorbid conditions in each age group.

RESULTS

54.6% (n=2 71 794) of people meeting the OA case definition had at least one of the eight selected comorbidities. Females had a significantly higher rate of comorbidities compared with males (standardised rates ratio=1.26, 95% CI 1.25 to 1.28). Depression, chronic obstructive pulmonary disease (COPD) and hypertension were the most prevalent in both females and males after age-standardisation, with 40% of all cases having any combination of these comorbidities. We observed a significant difference in the percentage of comorbidities among age groups, illustrated by the youngest age group (<45 years) having the highest percentage of cases with depression (24.6%), compared with a frequency of 16.1% in those >65 years.

CONCLUSIONS

Our findings highlight the high frequency of comorbidity in people with OA, with depression having the highest age-standardised prevalence rate. Comorbidities differentially affect females, and vary by age. These factors should inform healthcare programme and delivery.

Potentials, challenges, and limitations of the analysis of administrative data on vascular limb amputations in health care

Although more and more data on lower limb amputations are becoming available by leveraging the widening access to health care administrative databases, the applicability of these data for public health decisions is still limited. Problems can be traced back to methodological issues, how data are generated and to conceptual issues, namely, how data are interpreted in a multidimensional environment. The present review summarised all of the steps from converting the claims data of administrative databases into the analytical data and reviewed the wide array of sources of potential biases in the analysis of such data. The origins of uncertainty of administrative data analysis include uncontrolled confounding due to a lack of clinical data, the left- and right-censored nature of data collection, the non-standardized diagnosis/procedure-based data extraction methods (i.e., numerator/denominator problems) and additional methodological problems associated with temporal and spatial analyses. The existence of these methodological challenges in the administrative data-based analysis should not deter the analysts from using these data as a powerful tool in the armamentarium of clinical research. However, it must be done with caution and a thorough understanding and respect of the methodological limitations. In addition to this requirement, there is a profound need for pursuing further research on methodology and widening the search for other indicators (structural, process or outcome) that allow a deeper insight how the quality of vascular care may be assessed. Effective research using administrative data is based on strong collaboration in three domains, namely expertise in claims data handling and processing, the clinical field, and statistical analysis. The final interpretations of results and the countermeasures on the level of vascular care ought to be grounded on the integrity of research, open discussions and institutionalized mechanisms of science arbitration and honest brokering.

An Area-Level Indicator of Latent Soda Demand: Spatial Statistical Modeling of Grocery Store Transaction Data to Characterize the Nutritional Landscape in Montreal, Canada

Measurement of neighborhood dietary patterns at high spatial resolution allows public health agencies to identify and monitor communities with an elevated risk of nutrition-related chronic diseases. Currently, data on diet are obtained primarily through nutrition surveys, which produce measurements at low spatial resolutions. The availability of store-level grocery transaction data provides an opportunity to refine the measurement of neighborhood dietary patterns. We used these data to develop an indicator of area-level latent demand for soda in the Census Metropolitan Area of Montreal in 2012 by applying a hierarchical Bayesian spatial model to data on soda sales from 1,097 chain retail food outlets. The utility of the indicator of latent soda demand was evaluated by assessing its association with the neighborhood relative risk of prevalent type 2 diabetes mellitus. The indicator improved the fit of the disease-mapping model (deviance information criterion: 2,140 with the indicator and 2,148 without) and enables a novel approach to nutrition surveillance.

Association Between Prediabetes/Diabetes and Asthma Exacerbations in a Claims-Based Obese Asthma Cohort

BACKGROUND

Metabolic dysfunction may contribute to worsened asthma in obesity. The relationship between prediabetes and diabetes, metabolic conditions more common in obesity, and asthma outcomes is not well characterized.

OBJECTIVE

We sought to determine the association between prediabetes/diabetes and asthma exacerbations in an obese asthma cohort.

METHODS

A retrospective cohort of US obese adults with asthma, aged 18-64, was created from a claims-based health services database spanning 2010 to 2015. Individuals with a hemoglobin A1c (HbA1c) measurement were identified, categorized as within normal (<5.6%), prediabetes (5.7% to 6.4%), and diabetes (≥6.5%) ranges. Exacerbations, defined as asthma-related hospitalization, emergency department visit, or corticosteroid prescription ±14 days of an asthma-related outpatient visit, were ascertained. Associations were fit with zero-inflated negative binomial models, adjusted for age, sex, region, smoking, medication use, and comorbidities.

RESULTS

A total of 5722 individuals were identified. Higher HgbA1c was associated with higher asthma exacerbation rates. In the fully adjusted model, compared with individuals with normal HbA1c, those in the prediabetes range had a 27% higher rate (95% confidence interval [CI], 5% to 52%), and those in the diabetes range had a 33% higher rate (95% CI, 2% to 73%).

CONCLUSIONS

Prediabetes and diabetes were associated with higher rates of asthma exacerbation among obese adults with asthma. Results support evidence that insulin resistance and metabolic syndrome, metabolic features common in prediabetes/diabetes, can influence asthma morbidity.

A systematic review of database validation studies among fertility populations

STUDY QUESTION

Are routinely collected data from fertility populations adequately validated?

SUMMARY ANSWER

Of the 19 studies included, only one validated a national fertility registry and none reported their results in accordance with recommended reporting guidelines for validation studies.

WHAT IS KNOWN ALREADY

Routinely collected data, including administrative databases and registries, are excellent sources of data, particularly for reporting, quality assurance, and research. However, these data are subject to misclassification bias due to misdiagnosis or errors in data entry and therefore need to be validated prior to using for clinical or research purposes.

STUDY DESIGN SIZE DURATION

We conducted a systematic review by searching Medline, Embase, and CINAHL from inception to 6 October 2016 to identify validation studies of databases that contain routinely collected data in an ART setting. Webpages of international ART centers were also searched.

PARTICIPANTS/MATERIALS SETTING METHODS

We included studies that compared at least two data sources to validate ART population data. Key words and MeSH terms were adapted from previous systematic reviews investigating routinely collected data (e.g. administrative databases and registries), measures of validity (including sensitivity, specificity, and predictive value), and ART (including infertility, IVF, advanced reproductive age, and diminished ovarian reserve). Only full-text studies in English were considered. Results were synthesized qualitatively. The electronic search yielded 1074 citations, of which 19 met the inclusion criteria.

MAIN RESULTS AND THE ROLE OF CHANCE

Two studies validated a fertility database using medical records; seven studies used an IVF registry to validate vital records or maternal questionnaires, and two studies failed to adequately describe their reference standard. Four studies investigated the validity of mode of conception from birth registries; two studies validated diagnoses or treatments in a fertility database; four studies validated a linkage algorithm between a fertility registry and another administrative database; one study created an algorithm in a single database to identify a patient population. Sensitivity was the most commonly reported measure of validity (12 studies), followed by specificity (9 studies). Only three studies reported four or more measures of validation, and five studies presented CIs for their estimates. The prevalence of the variable in the target population (pre-test prevalence) was reported in seven studies; however, only four of the studies had prevalence estimates from the study population (post-test prevalence) within a 2% range of the pre-test estimate. The post-test estimate was largely discrepant from the pre-test value in two studies.

LIMITATIONS REASONS FOR CAUTION

The search strategy was limited to the studies and reports published in English, which may not capture validation studies from countries that do not speak English. Furthermore, only three specific fertility-based diagnostic variables (advanced reproductive age, diminished ovarian reserve, and chorionicity) were searched in Medline, Embase, and CINAHL. Consequently, published studies with other diagnoses or conditions relevant to infertility may not have been captured in our review.

WIDER IMPLICATIONS OF THE FINDINGS

There is a paucity of literature on validation of routinely collected data from a fertility population. Furthermore, the prevalence of the markers that have been validated are not being presented, which can lead to biased estimates. Stakeholders rely on these data for monitoring outcomes of treatments and adverse events; therefore, it is essential to ascertain the accuracy of these databases and make the reports publicly available.

STUDY FUNDING/COMPETING INTERESTS

This study was supported by Canadian Institutes of Health Research (CIHR) (FDN-148438). There are no competing interests for any of the authors.

REGISTRATION NUMBER

International Prospective Register of Systematic Reviews ID: CRD42016048466.

Considering parents as a unit: Associations of gestational diabetes and gestational hypertension with postpartum diabetes and hypertension in couples

OBJECTIVES

To evaluate the associations of a combined indicator of gestational diabetes mellitus (GDM) and gestational hypertension (GH) with diabetes and with hypertension in parental couples as a 'unit'.

STUDY DESIGN

Utilizing administrative health data, GH was identified in matched pairs (GDM vs. no GDM) of mothers with singleton live births (Quebec, Canada; cohort inception 1990-2007). Couples were categorized based on GDM/GH status s ('neither'; 'either'; or 'both'). Using validated administrative health database definitions, associations of this indicator with diabetes and with hypertension in both members of the couple (12 weeks postpartum to March 31, 2012) were evaluated through adjusted Cox proportional hazard models.

RESULTS

In 63,438 couples over a mean of 12.8 years (SD 5.4), diabetes risk was 9-fold higher (HR: 8.9; 95% CI 6.4, 12.2) in couples with either GDM/GH and 16-fold higher (HR 16.0; 95% CI 10.9, 23.5) in couples with both conditions compared to those with neither. Hypertension risk was 2-fold higher (HR: 1.8; 95% CI 1.5, 2.0) in couples either GDM/GH and 6-fold higher (HR 5.8; 95% CI 4.9, 7.0) in couples with both conditions compared to those with neither condition.

CONCLUSIONS

GDM/GH predict diabetes and hypertension in couples as a unit underscoring the concept of shared couple risk. These findings may help foster couple collaboration for cardiovascular risk reduction in the household.

Estimating the prevalence of diabetes mellitus and thyroid disorders using medication data in Flanders, Belgium

Background

Various methods exist to estimate disease prevalences. The aim of this study was to determine whether dispensed, self-reported and prescribed medication data could be used to estimate the prevalence of diabetes mellitus and thyroid disorders. Second, these pharmaco-epidemiological estimates were compared with prevalences based on self-reported diagnoses and doctor-registered diagnoses.

Methods

Data on medication for diabetes and thyroid disorders were obtained from three different sources in Flanders (Belgium) for 2008: a purely administrative database containing data on dispensed medication, the Belgian National Health Interview Survey for self-reported medication and diagnoses, and a patient record database for prescribed medication and doctor-registered diagnoses. Prevalences were estimated based on medication data and compared with each other. Cross-tabulations of dispensed medication and self-reported diagnoses, and prescribed medication and doctor-registered diagnoses, were investigated.

Results

Prevalences based on dispensed medication were the highest (4.39 and 2.98% for diabetes and thyroid disorders, respectively). The lowest prevalences were found using prescribed medication (2.39 and 1.72%, respectively). Cross-tabulating dispensed medication and self-reported diagnoses yielded a moderate to high sensitivity for diabetes (90.4%) and thyroid disorders (77.5%), while prescribed medication showed a low sensitivity for doctor-registered diagnoses (56.5 and 43.6%, respectively). The specificity remained above 99% in all cases.

Conclusions

This study was the first to perform cross-tabulations for disease prevalence estimates between different databases and within (sub)populations. Purely administrative database was shown to be a reliable source to estimate disease prevalence based on dispensed medication. Prevalence estimates based on prescribed or self-reported medication were shown to have important limitations.

Comparing external and internal validation methods in correcting outcome misclassification bias in logistic regression: A simulation study and application to the case of postsurgical venous thromboembolism following total hip and knee arthroplasty

PURPOSE

We assessed the validity of postsurgery venous thromboembolism (VTE) diagnoses identified from administrative databases and compared Bayesian and multiple imputation (MI) approaches in correcting for outcome misclassification in logistic regression models.

METHODS

Sensitivity and specificity of postsurgery VTE among patients undergoing total hip or knee replacement (THR/TKR) were assessed against chart review in six Montreal hospitals in 2009 to 2010. Administrative data on all THR/TKR Quebec patients in 2009 to 2010 were obtained. The performance of Bayesian external, Bayesian internal, and MI approaches to correct the odds ratio (OR) of postsurgery VTE in tertiary versus community hospitals was assessed using simulations. Bayesian external approach used prior information from external sources, while Bayesian internal and MI approaches used chart review.

RESULTS

In total, 17 319 patients were included, 2136 in participating hospitals, among whom 75 had VTE in administrative data versus 81 in chart review. VTE sensitivity was 0.59 (95% confidence interval, 0.48-0.69) and specificity was 0.99 (0.98-0.99), overall. The adjusted OR of VTE in tertiary versus community hospitals was 1.35 (1.12-1.64) using administrative data, 1.45 (0.97-2.19) when MI was used for misclassification correction, and 1.53 (0.83-2.87) and 1.57 (0.39-5.24) when Bayesian internal and external approaches were used, respectively. In simulations, all three approaches reduced the OR bias and had appropriate coverage for both nondifferential and differential misclassification.

CONCLUSION

VTE identified from administrative data had low sensitivity and high specificity. The Bayesian external approach was useful to reduce outcome misclassification bias in logistic regression; however, it required accurate specification of the misclassification properties and should be used with caution.

Identifying diabetes cases in health administrative databases: a validation study based on a large French cohort

OBJECTIVES

In the French national health insurance information system (SNDS) three diabetes case definition algorithms are applied to identify diabetic patients. The objective of this study was to validate those using data from a large cohort.

METHODS

The CONSTANCES cohort (Cohorte des consultants des Centres d'examens de santé) comprises a randomly selected sample of adults living in France. Between 2012 and 2014, data from 45,739 participants recorded in a self-administrated questionnaire and in a medical examination were linked to the SNDS. Two gold standards were defined: known diabetes and pharmacologically treated diabetes. Sensitivity, specificity, positive and negative predictive values (PPV, NPV) and kappa coefficients (k) were estimated.

RESULTS

All three algorithms had specificities and NPV over 99%. Their sensitivities ranged from 73 to 77% in algorithm A, to 86 and 97% in algorithm B and to 93 and 99% in algorithm C, when identifying known and pharmacologically treated diabetes, respectively. Algorithm C had the highest k when using known diabetes as the gold standard (0.95). Algorithm B had the highest k (0.98) when testing for pharmacologically treated diabetes.

CONCLUSIONS

The SNDS is an excellent source for diabetes surveillance and studies on diabetes since the case definition algorithms applied have very good test performances.

The Use of International Classification of Diseases Codes to Identify Patients with Pancreatitis: A Systematic Review and Meta-analysis of Diagnostic Accuracy Studies

Background

Hospital discharge codes are increasingly used in gastroenterology research, but their accuracy in the setting of acute pancreatitis (AP) and chronic pancreatitis (CP), one of the most frequent digestive diseases, has never been assessed systematically. The aim was to conduct a systematic literature review and determine accuracy of diagnostic codes for AP and CP, as well as the effect of covariates.

Methods

Three databases (Pubmed, EMBASE and Scopus) were searched by two independent reviewers for relevant studies that used International Classification of Disease (ICD) codes. Summary estimates of sensitivity, specificity and positive predictive value were obtained from bivariate random-effects regression models. Sensitivity and subgroup analyses according to recurrence of AP and age of the study population were performed.

Results

A total of 24 cohorts encompassing 18,106 patients were included. The pooled estimates of sensitivity and specificity of ICD codes for AP were 0.85 and 0.96, respectively. The pooled estimates of sensitivity and specificity of ICD codes for CP were 0.75 and 0.94, respectively. The positive predictive value of ICD codes was 0.71 for either AP or CP. It increased to 0.78 when applied to incident episode of AP only. The positive predictive value decreased to 0.68 when the ICD codes were applied to paediatric patients.

Conclusion

Nearly three out of ten patients are misidentified as having either AP or CP with the indiscriminate use of ICD codes. Limiting the use of ICD codes to adult patients with incident episode of AP may improve identification of patients with pancreatitis in administrative databases.

Primary versus secondary source of data in observational studies and heterogeneity in meta-analyses of drug effects: a survey of major medical journals

Background

The data from individual observational studies included in meta-analyses of drug effects are collected either from ad hoc methods (i.e. “primary data”) or databases that were established for non-research purposes (i.e. “secondary data”). The use of secondary sources may be prone to measurement bias and confounding due to over-the-counter and out-of-pocket drug consumption, or non-adherence to treatment. In fact, it has been noted that failing to consider the origin of the data as a potential cause of heterogeneity may change the conclusions of a meta-analysis. We aimed to assess to what extent the origin of data is explored as a source of heterogeneity in meta-analyses of observational studies.

Methods

We searched for meta-analyses of drugs effects published between 2012 and 2018 in general and internal medicine journals with an impact factor > 15. We evaluated, when reported, the type of data source (primary vs secondary) used in the individual observational studies included in each meta-analysis, and the exposure- and outcome-related variables included in sensitivity, subgroup or meta-regression analyses.

Results

We found 217 articles, 23 of which fulfilled our eligibility criteria. Eight meta-analyses (8/23, 34.8%) reported the source of data. Three meta-analyses (3/23, 13.0%) included the method of outcome assessment as a variable in the analysis of heterogeneity, and only one compared and discussed the results considering the different sources of data (primary vs secondary).

Conclusions

In meta-analyses of drug effects published in seven high impact general medicine journals, the origin of the data, either primary or secondary, is underexplored as a source of heterogeneity.

Electronic supplementary material

The online version of this article (10.1186/s12874-018-0561-3) contains supplementary material, which is available to authorized users.

Maternal diabetes mellitus and timing of pubertal development in daughters and sons: a nationwide cohort study

OBJECTIVE

To study the association between maternal diabetes and timing of pubertal development in daughters and sons.

DESIGN

Prospective cohort study.

SETTING

Not applicable.

PATIENT(S)

A total of 15,822 mother-child pairs included in the Danish National Birth Cohort and the Puberty Cohort with prospectively collected, register-based and self-reported information on maternal diabetes and self-reported information on pubertal development.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Adjusted mean monthly difference in age at attaining pubertal milestones in children born of mothers with diabetes compared with children born of mothers without diabetes.

RESULT(S)

A total of 502 children were born of mothers with diabetes during pregnancy. In daughters exposed to gestational diabetes mellitus, we observed advanced onset in all pubertal milestones. The associations were statistically significant with regard to pubic hair Tanner stage 2 (-4.8 months) (95% confidence interval [CI] -7.7, -2.0), pubic hair Tanner stage 3 (-2.2 months) (95 % CI -4.4, 0.0), pubic hair Tanner stage 5 (-6.0 months) (95% CI -10.8, -1.2), and menarche (-2.5 months) (95 % CI -4.9, 0.0). We observed no tendencies between maternal type 1 or type 2 diabetes mellitus and pubertal development in daughters. We observed no associations between maternal diabetes and pubertal development in sons.

CONCLUSION(S)

Our findings suggest that gestational diabetes mellitus may accelerate the pubertal development in daughters. Our results did not support an association between type 1 or type 2 diabetes mellitus and daughters' pubertal development, as well as between any type of maternal diabetes and sons' pubertal development.

The impact of depression medications on oral antidiabetic drug adherence in patients with diabetes and depression

AIMS

To compare adherence and persistence to oral antidiabetic drugs (OAD) between patients who are new users of second generation antipsychotics (SGA) versus new users of other depression therapies in adults with type 2 diabetes mellitus (T2DM) and major depressive disorder (MDD).

METHODS

Adults 18-64 years with previously-treated T2DM and MDD (past OAD and SSRI/SNRI use) who are new users of SGA or non-SGA therapies (bupropion, lithium, mirtazapine, thyroid hormone, tricyclic antidepressant) were identified in the 2009-2015 MarketScan® Commercial Claims and Encounters database. Multivariate regression models were used to determine the odds of a ≥10% decline in OAD adherence over 180- and 365-days, and time to OAD discontinuation, adjusting for differences between groups.

RESULTS

A total of 8664 (21.5% SGA), 8311 (22.1% SGA), and 17,524 (21.3% SGA) patients met inclusion criteria for the 180-day adherence, 365-day adherence, and persistence cohorts, respectively. Over 180-days, 16.6% of SGA and 13.3% of non-SGA initiators had a ≥10% decline in OAD adherence (adjusted odds ratio [OR] = 1.41, 95% CI 1.21-1.63). Over 365-days, 22.3% of SGA and 18.9% of non-SGA initiators had a ≥ 10% decline (OR = 1.34, 95% CI 1.17-1.53). Time to OAD discontinuation was similar between groups (adjusted hazard ratio = 1.03, 95% CI 0.94-1.12).

CONCLUSION

Use of SGA was associated with a 1.3-1.4 times higher odds of a ≥10% decline in OAD adherence. Adherence to OAD is critical for optimal diabetes control and reductions in this magnitude may impact A1C. Close monitoring of OAD adherence after SGA initiation is warranted.

Can administrative health utilisation data provide an accurate diabetes prevalence estimate for a geographical region?

AIM

To validate the New Zealand Ministry of Health (MoH) Virtual Diabetes Register (VDR) using longitudinal laboratory results and to develop an improved algorithm for estimating diabetes prevalence at a population level.

METHODS

The assigned diabetes status of individuals based on the 2014 version of the MoH VDR is compared to the diabetes status based on the laboratory results stored in the Auckland regional laboratory result repository (TestSafe) using the New Zealand diabetes diagnostic criteria. The existing VDR algorithm is refined by reviewing the sensitivity and positive predictive value of the each of the VDR algorithm rules individually and as a combination.

RESULTS

The diabetes prevalence estimate based on the original 2014 MoH VDR was 17% higher (n = 108,505) than the corresponding TestSafe prevalence estimate (n = 92,707). Compared to the diabetes prevalence based on TestSafe, the original VDR has a sensitivity of 89%, specificity of 96%, positive predictive value of 76% and negative predictive value of 98%. The modified VDR algorithm has improved the positive predictive value by 6.1% and the specificity by 1.4% with modest reductions in sensitivity of 2.2% and negative predictive value of 0.3%. At an aggregated level the overall diabetes prevalence estimated by the modified VDR is 5.7% higher than the corresponding estimate based on TestSafe.

CONCLUSION

The Ministry of Health Virtual Diabetes Register algorithm has been refined to provide a more accurate diabetes prevalence estimate at a population level. The comparison highlights the potential value of a national population long term condition register constructed from both laboratory results and administrative data.

Systematic review and meta-analysis of algorithms used to identify drug-induced liver injury (DILI) in health record databases

BACKGROUND & AIMS

Drug induced liver injury (DILI) is largely underreported, leading to underestimation of its burden. Electronic detection of DILI in healthcare databases shows promise to overcome the issues of spontaneous reporting. The performance of detection algorithms may vary because of inconsistent DILI definition and detection criteria. We performed a systematic review and meta-analysis to identify the DILI detection criteria used in health record databases and determine the performance characteristics of the detection algorithms.

METHODS

We searched PubMed, EMBASE and Scopus for studies that utilized laboratory threshold criteria to identify DILI cases. Validation studies were included in the meta-analysis. Data were abstracted using standardized forms and quality was assessed using modified Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) criteria. We evaluate the performance characteristics of the detection algorithm by obtaining the pooled estimate of the positive predictive value (PPV) assuming a random effects model.

RESULTS

A total of 29 studies met the inclusion criteria for the systematic review; 25 of these studies (n = 35 948) had PPV estimates for performing the meta-analysis. The PPV of DILI detection algorithms was low, ranging from 1.0% to 40.2%, with a pooled estimate of 14.6% (95% CI 10.7-18.9). Algorithms that performed better had prespecified exclusion diagnoses as well as drugs of interest to minimize false positives.

CONCLUSION

Algorithm performance varied with different case definitions of DILI attributed to different laboratory threshold criteria, diagnosis codes, and study drugs.

Hospital Admissions in People With Alzheimer's Disease or Senile Dementia According to Type 2 Diabetes Status: An Observational 10-Year Study

We aimed to describe admission rates and outcomes of hospitalized people with Alzheimer's disease (AD) or senile dementia stratified by type 2 diabetes (T2D) in Spain, 2004 to 2013 (N = 541 858, 143 501 [26.5%] of whom had T2D). We excluded alternative causes of dementia. Hospitalization rates were higher in people with T2D (130.5 vs 91.5 cases/105 people). People older than 84 years and comorbidity increased over time. Crude inhospital mortality (IHM) decreased over time both in people with and without T2D (all P values <.001). Time trend analyses, 2004 to 2013, showed an overall adjusted incidence rate ratio (95% confidence interval [95% CI]) of hospitalization of 1.41 (1.40-1.42) for T2D (men, 1.32 [1.30-1.33]; women, 1.46 [1.45-1.48]). In logistic regression analyses, IHM decreased over time (odds ratio, OR [95% CI] = 0.97 [0.96-0.98]) and T2D was not associated with a higher IHM (OR [95% CI] = 0.99 [0.98-1.01]). In conclusion, admission rates were higher in patients with T2D. Inhospital mortality decreased over time. Diabetes did not predict IHM in patients with AD or senile dementia.

Diabetes risk in women with gestational diabetes mellitus and a history of polycystic ovary syndrome: a retrospective cohort study

AIMS

To investigate whether polycystic ovary syndrome further increases postpartum diabetes risk in women with gestational diabetes mellitus and to explore relationships between polycystic ovary syndrome and incident diabetes in women who do not develop gestational diabetes.

METHODS

This retrospective cohort study (Quebec Physician Services Claims; Hospitalization Discharge Databases; Birth and Death registries) included 34 686 women with gestational diabetes during pregnancy (live birth), matched 1:1 to women without gestational diabetes by age group, year of delivery and health region. Diagnostic codes were used to define polycystic ovary syndrome and incident diabetes. Cox regression models were used to examine associations between polycystic ovary syndrome and incident diabetes.

RESULTS

Polycystic ovary syndrome was present in 1.5% of women with gestational diabetes and 1.2% of women without gestational diabetes. There were more younger mothers and mothers who were not of white European ancestry among those with polycystic ovary syndrome. Those with polycystic ovary syndrome more often had a comorbidity and a lower proportion had a previous pregnancy. Polycystic ovary syndrome was associated with incident diabetes (hazard ratio 1.52; 95% CI 1.27, 1.82) among women with gestational diabetes. No conclusive associations between polycystic ovary syndrome and diabetes were identified (hazard ratio 0.94; 95% CI 0.39, 2.27) in women without gestational diabetes.

CONCLUSION

In women with gestational diabetes, polycystic ovary syndrome confers additional risk for incident diabetes postpartum. In women without gestational diabetes, an association between PCOS and incident diabetes was not observed. Given the already elevated risk of diabetes in women with a history of gestational diabetes, a history of both polycystic ovary syndrome and gestational diabetes signal a critical need for diabetes surveillance and prevention.

Conjoint Associations of Gestational Diabetes and Hypertension With Diabetes, Hypertension, and Cardiovascular Disease in Parents: A Retrospective Cohort Study

The conjoint association of gestational diabetes mellitus (GDM) and gestational hypertension (GH) with cardiometabolic disease has not been well studied. We evaluated a combined GDM/GH risk indicator in both mothers and fathers because of shared spousal behaviors and environments. In the present population-based retrospective cohort study, GH was identified in matched pairs of mothers with GDM or without GDM (matched on age group, health region, and year of delivery) who had singleton live births in Quebec, Canada (1990–2007). A total of 64,232 couples were categorized based on GDM/GH status (neither, either, or both). Associations with diabetes, hypertension, and a composite of cardiovascular disease (CVD) and mortality were evaluated using Cox proportional hazard models (from 12 weeks postpartum to March 2012). Compared with having neither GDM nor GH, having either was associated with incident diabetes (hazard ratio (HR) = 14.7, 95% confidence interval (CI): 12.9, 16.6), hypertension (HR = 1.9, 95% CI: 1.8, 2.0), and CVD/mortality (HR = 1.4, 95% CI: 1.2, 1.7). We found associations of greater magnitude among participants who had both (for diabetes, HR = 36.9, 95% CI: 26.0, 52.3; for hypertension, HR = 5.7, 95% CI: 4.9, 6.7; and for CVD/mortality, HR = 2.4, 95% CI: 1.6, 3.5). Associations with diabetes were also observed in fathers (for either, HR = 1.2, 95% CI: 1.1, 1.3; for both, HR = 1.8, 95% CI: 1.4, 2.3). In conclusion, we found associations of a combined GDM/GH indicator with cardiometabolic disease in mothers and with diabetes in fathers, with stronger associations when both GDM and GH were present.

Novel Methods and Data Sources for Surveillance of State-Level Diabetes and Prediabetes Prevalence

States bear substantial responsibility for addressing the rising rates of diabetes and prediabetes in the United States. However, accurate state-level estimates of diabetes and prediabetes prevalence that include undiagnosed cases have been impossible to produce with traditional sources of state-level data. Various new and nontraditional sources for estimating state-level prevalence are now available. These include surveys with expanded samples that can support state-level estimation in some states and administrative and clinical data from insurance claims and electronic health records. These sources pose methodologic challenges because they typically cover partial, sometimes nonrandom subpopulations; they do not always use the same measurements for all individuals; and they use different and limited sets of variables for case finding and adjustment. We present an approach for adjusting new and nontraditional data sources for diabetes surveillance that addresses these limitations, and we present the results of our proposed approach for 2 states (Alabama and California) as a proof of concept. The method reweights surveys and other data sources with population undercoverage to make them more representative of state populations, and it adjusts for nonrandom use of laboratory testing in clinically generated data sets. These enhanced diabetes and prediabetes prevalence estimates can be used to better understand the total burden of diabetes and prediabetes at the state level and to guide policies and programs designed to prevent and control these chronic diseases.

Impact of type 2 diabetes mellitus on in-hospital-mortality after major cardiovascular events in Spain (2002-2014)

BACKGROUND

Diabetes mellitus has long been associated with cardiovascular events. Nevertheless, the higher burden of traditional cardiovascular risk factors reported in high-income countries is offset by a more widespread use of preventive measures and revascularization or other invasive procedures. The aim of this investigation is to describe trends in number of cases and outcomes, in-hospital mortality (IHM) and length of hospital stay (LHS), of hospital admissions for major cardiovascular events between type 2 diabetes (T2DM) and matched non-diabetes patients.

METHODS

Retrospective study using National Hospital Discharge Database, analyzed in 4 years 2002, 2006, 2010, 2014, in Spain. We included patients (≥ 40 years old) with a primary diagnosis of myocardial infarction, ischemic and hemorrhagic stroke, aortic aneurysm and dissection and acute lower limb ischemia in people with T2DM. Cases were matched with controls (without T2DM) by ICD-9-CM codes, sex, age, province of residence and year.

RESULTS

We selected 130,011 matched couples (50,427 with myocardial infarction, 60,236 with stroke, 2599 with aortic aneurysm and dissection and 16,749 with acute lower limb ischemia. Among T2DM patients we found increasing numbers of admissions overtime for stroke (10,794 in 2002 vs 17,559 in 2014), aortic aneurysm and dissection (390 vs 841) and acute lower limb ischemia (3854 vs. 4548). People were progressively older (except for myocardial infarction), had more comorbidities (especially T2DM patients), and were more frequently coded overtime for cardiovascular risk factors (smoking, obesity, hypertension, lipid disorders) and renal diseases. LHS and IHM declined overtime, though IHM only did it significantly in T2DM patients. Multivariable adjustment showed that T2DM patients had a significantly 15% higher mortality rate during admission for myocardial infarction, a 6% higher mortality for stroke, and a 6% higher mortality rate for "all cardiovascular events combined", than non-diabetic matched controls.

CONCLUSIONS

The number of hospital admissions for stroke, aortic aneurysm and dissection and acute lower limb ischemia increased overtime, but remained stable for myocardial infarction. T2DM is associated to higher IHM after major cardiovascular events. Further research is needed to help us understand the reasons for an apparently increased mortality in T2DM patients when admitted to hospital for some major cardiovascular events.

Sleep duration and incidence of type 2 diabetes: the Multiethnic Cohort

OBJECTIVES

As an emerging risk factor for the rising incidence of type 2 diabetes, we examined sleep duration in relation to type 2 diabetes and several biomarkers.

DESIGN

Prospective cohort recruited 1993-1996.

SETTING

The Multiethnic Cohort in Hawaii and California.

PARTICIPANTS

A cohort of 151,691 White, African American, Japanese American, Native Hawaiian, and Latino participants; 9695 cohort members had biomarker measurements.

MEASUREMENTS

Sleep duration was self-reported at cohort entry. Diabetes status was obtained from 3 questionnaires and confirmed by 3 administrative data sources. Biomarkers were measured by standard assays 9.6±2.1 years after cohort entry. We estimated diabetes risk as a time-varying outcome using Cox regression adjusted for body mass index assessed at 3 time points and other known confounders and computed adjusted means of biomarkers by sleep hours.

RESULTS

During 7.9±3.5 years of follow-up, 8487 new diabetes cases were diagnosed. Long sleep duration (≥9 hours), as compared with 7-8 hours, was significantly associated with higher incidence (hazard ratio, 1.12; 95% confidence interval 1.04-1.21), but the 4% elevated incidence for short sleep duration (≤6 hours) did not reach significance (95% confidence interval 0.99-1.09). After stratification, the associations appeared stronger in Japanese American than other ethnic groups and in participants without comorbidity. Hours of sleep were positively associated with C-reactive protein and triglycerides and inversely related to high-density lipoprotein cholesterol and adiponectin but not with leptin levels and homeostatic model assessment of insulin resistance.

CONCLUSION

In this multiethnic population, the 12% higher diabetes risk for long sleep hours may be mediated through inflammation, a poor lipid profile, and lower adiponectin levels.

Observational study of vascular dementia in the Spanish elderly population according to type 2 diabetes status: trends in incidence, characteristics and outcomes (2004-2013)

OBJECTIVES

To examine trends overtime in the incidence and in-hospital outcomes of vascular dementia (VaD) hospitalisations in patients aged 70 years or over suffering and not suffering from type 2 diabetes mellitus (T2DM) between 2004 and 2013 in Spain.

DESIGN

Retrospective study.

SETTING

Spain.

PARTICIPANTS

National hospital discharge data were used; patients aged ≥70, discharged from a hospital with VaD as a primary diagnosis, were selected.

MAIN OUTCOME MEASURES

Overall incidence, therapeutic and diagnostic procedures, comorbidities, infectious complications, duration of hospital stays and in-hospital mortality (IHM).

RESULTS

In total, 170 607 admissions for VaD (34.3% with T2DM) were identified. We found a significant upward linear trend in the incidence of VaD for men and women with and without diabetes between 2004 and 2013. The adjusted incidence was higher among people with T2DM over the study period. We found a higher incidence in men than women in all years under study. A positive association between T2DM and VaD hospitalisation was found among both men (IRR 2.14, 95% CI 2.11 to 2.16) and women (incidence rate ratio (IRR) 2.22; 95% CI 2.19 to 2.25). Pneumonia was significantly associated with a higher mortality (OR 2.59, 95% CI 2.52 to 2.67). We found that percutaneous endoscopic gastrostomy was associated with lower IHM (OR 0.37, 95% CI 0.31 to 0.45), while parenteral nutrition had the opposite effect (OR 1.29, 95% CI 1.18 to 1.41). There was no association between diabetes and higher IHM (OR 0.99, 95% CI 0.93 to 1.06). The time-trend analyses of the entire sample showed a significant reduction in mortality in patients with VaD (OR 0.98, 95% CI 0.97 to 0.99).

CONCLUSIONS

Incidence rates for VaD hospitalisations were twice as high in patients with diabetes compared with those without. Men had significantly higher incidence rates than women, regardless of diabetes status. In both groups studied, pneumonia and parenteral nutrition were associated with mortality while percutaneous endoscopic gastrostomy was associated with survival. Having diabetes was not associated with higher IHM after hospitalisation with VaD.

Clinically Apparent Arterial Thrombosis in Persons with Systemic Vasculitis

Objective

To estimate the incidence rate of clinically apparent arterial thrombotic events and associated comorbidities in patients with primary systemic vasculitis.

Methods

Using large cohort administrative data from Quebec, Canada, we identified patients with vasculitis, including polyarteritis nodosa (PAN) and granulomatosis with polyangiitis (GPA). Incident acute myocardial infarctions (AMIs) and cerebrovascular accidents (CVAs) after the diagnosis of vasculitis were ascertained in the PAN and GPA group via billing and hospitalization data. These were compared to rates of a general population comparator group. The incidences of comorbidities (type 2 diabetes mellitus, dyslipidemia, and hypertension) were also collected.

Results

Among the 626 patients identified with vasculitis, 19.7% had PAN, 2.9% had Kawasaki disease, 23.8% had GPA, 52.4% had GCA, and 1.3% had Takayasu arteritis. The AMI rate was substantially higher in males aged 18–44 with PAN, with rates up to 268.1 events per 10,000 patient years [95% CI 67.1–1070.2], approximately 30 times that in the age- and sex-matched control group. The CVA rate was also substantially higher, particularly in adults aged 45–65. Patients with vasculitis had elevated incidences of diabetes, dyslipidemia, and hypertension versus the general population.

Conclusion

Atherothrombotic rates were elevated in patients identified as having primary systemic vasculitis. While incident rates of cardiovascular comorbidities were also increased, the substantial elevation in AMIs seen in young adults suggests a disease-specific component which requires further investigation.

Trajectory of systolic blood pressure in a low-income, racial-ethnic minority cohort with diabetes and baseline uncontrolled hypertension

In two primary care clinics in Texas serving low-income patients, systolic blood pressure (SBP) trajectory was examined during 2 years in patients with diabetes mellitus (mean SBP ≥140 mm Hg: 152 mm Hg±11.2 in the baseline year). Among 860 eligible patients, 62.0% were women, 78.8% were Hispanic, and 41.2% were uninsured. Overall, SBP dropped 0.56 mm Hg per month or 13.4 mm Hg by 24 months. For patients with mean glycated hemoglobin ≥9% in year 1, SBP declined 4.8 mm Hg less by 24 months vs those with glycated hemoglobin <7% (P=.03). Compared with white women, SPB declined 7.2 mm Hg less by 24 months in Hispanic women (P=.03) and 9.6 mm Hg less by 24 months in black men (P=.04). SBP also declined 9.1 mm Hg less by 24 months for patients taking four or more blood pressure drug classes at baseline vs one drug class. In this low-income cohort, clinically complex patients and racial-ethnic minorities had clinically significantly smaller declines in SBP.

Validity of Health Administrative Database Definitions for Hypertension: A Systematic Review

BACKGROUND

Health administrative data are frequently used for hypertension surveillance. The aim of this systematic review was to determine the sensitivity and specificity of the commonly used hypertension case definition of 2 physician outpatient claims within a 2-year period or 1 hospital discharge abstract record.

METHODS

Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we searched MEDLINE (from 1946) and EMBASE (from 1947) for relevant studies through September 2016 (keywords: "hypertension," "administrative databases," "validation studies"). Data with standardized forms and assessed quality using Quality Assessment of Diagnostic Accuracy Studies criteria were reviewed by 2 reviewers. Pooled sensitivity and specificity were estimated using a generalized linear-model approach to random-effects bivariate regression meta-analysis.

RESULTS

The search strategy identified 1732 abstracts, among which 3 articles were deemed relevant. One of the articles incorporated 2 studies with differing reference standards and study populations; thus, we considered each separately. The quality scores of the retained studies ranged from 10-12 of a maximum 14. The sensitivity of the definition investigated to identify hypertension using administrative health databases was 71.2% (95% confidence interval [CI], 68.3-73.7) and the specificity was 94.5% (95% CI, 93.2-95.6) when compared with surveys or medical records.

CONCLUSIONS

The 2 physician outpatient claims within a 2-year period or 1 hospital discharge abstract record hypertension case definition accurately classifies individuals as hypertensive in approximately 70% of cases and correctly identifies persons as nonhypertensive in approximately 95% of cases. This is likely sufficiently sensitive and specific for most research and surveillance purposes.

Correcting hazard ratio estimates for outcome misclassification using multiple imputation with internal validation data

OBJECTIVE

Outcome misclassification may occur in observational studies using administrative databases. We evaluated a two-step multiple imputation approach based on complementary internal validation data obtained from two subsamples of study participants to reduce bias in hazard ratio (HR) estimates in Cox regressions.

METHODS

We illustrated this approach using data from a surveyed sample of 6247 individuals in a study of statin-diabetes association in Quebec. We corrected diabetes status and onset assessed from health administrative data against self-reported diabetes and/or elevated fasting blood glucose (FBG) assessed in subsamples. The association between statin use and new onset diabetes was evaluated using administrative data and the corrected data. By simulation, we assessed the performance of this method varying the true HR, sensitivity, specificity, and the size of validation subsamples.

RESULTS

The adjusted HR of new onset diabetes among statin users versus non-users was 1.61 (95% confidence interval: 1.09-2.38) using administrative data only, 1.49 (0.95-2.34) when diabetes status and onset were corrected based on self-report and undiagnosed diabetes (FBG ≥ 7 mmol/L), and 1.36 (0.92-2.01) when corrected for self-report and undiagnosed diabetes/impaired FBG (≥ 6 mmol/L). In simulations, the multiple imputation approach yielded less biased HR estimates and appropriate coverage for both non-differential and differential misclassification. Large variations in the corrected HR estimates were observed using validation subsamples with low participation proportion. The bias correction was sometimes outweighed by the uncertainty introduced by the unknown time of event occurrence.

CONCLUSION

Multiple imputation is useful to correct for outcome misclassification in time-to-event analyses if complementary validation data are available from subsamples. Copyright © 2017 John Wiley & Sons, Ltd.

Age- and sex-specific prevalence of diabetes associated with diseases of the exocrine pancreas: A population-based study

BACKGROUND AND AIMS

Diabetes associated with diseases of the exocrine pancreas (DP) is a recognized clinical condition but data on its prevalence are limited to a few single centre studies. Relative contribution of the three major diseases of the exocrine pancreas (acute pancreatitis, chronic pancreatitis, pancreatic cancer) to prevalence of DP as well as the effect of age and sex is largely unknown. The study aimed to determine age- and sex-specific prevalence of DP overall and after acute pancreatitis, chronic pancreatitis, and pancreatic cancer alone at the population level.

METHODS

Nationwide population database covering nearly 3 million residents in New Zealand over a 10-year study period was used. DP was identified based on International Classification of Diseases-10 codes. Data were reported as prevalence per 1000 population and corresponding 95% confidence intervals.

RESULTS

The crude prevalence of DP was 1.13 [1.12, 1.14] per 1000, with 70-79 years age group having the highest prevalence at 3.94 [3.92, 3.97] per 1000. Men had an overall prevalence of 1.32 [1.31, 1.33] per 1000 and women-0.93 [0.92, 0.94] (p<0.05). Acute pancreatitis contributed 61% to overall prevalence of DP.

CONCLUSIONS

Prevalence of DP in the general population is close to that of type 1 diabetes. Three out of five DP cases develop after acute pancreatitis. There is a variation in age of onset of DP, with the working and ageing population most affected. Men have a 40% higher risk of developing DP than women.

Identifying Local Hot Spots of Pediatric Chronic Diseases Using Emergency Department Surveillance

OBJECTIVE

To use novel geographic methods and large-scale claims data to identify the local distribution of pediatric chronic diseases in New York City.

METHODS

Using a 2009 all-payer emergency claims database, we identified the proportion of unique children aged 0 to 17 with diagnosis codes for specific medical and psychiatric conditions. As a proof of concept, we compared these prevalence estimates to traditional health surveys and registry data using the most geographically granular data available. In addition, we used home addresses to map local variation in pediatric disease burden.

RESULTS

We identified 549,547 New York City children who visited an emergency department at least once in 2009. Though our sample included more publicly insured and uninsured children, we found moderate to strong correlations of prevalence estimates when compared to health surveys and registry data at prespecified geographic levels. Strongest correlations were found for asthma and mental health conditions by county among younger children (0.88, P = .05 and 0.99, P < .01, respectively). Moderate correlations by neighborhood were identified for obesity and cancer (0.53 and 0.54, P < .01). Among adolescents, correlations by health districts were strong for obesity (0.95, P = .05), and depression estimates had a nonsignificant, but strong negative correlation with suicide attempts (-0.88, P = .12). Using SaTScan, we also identified local hot spots of pediatric chronic disease.

CONCLUSIONS

For conditions easily identified in claims data, emergency department surveillance may help estimate pediatric chronic disease prevalence with higher geographic resolution. More studies are needed to investigate limitations of these methods and assess reliability of local disease estimates.