Database Linkage Research Articles

Heredity of pregnancy-related pelvic girdle pain in Sweden.

Pelvic girdle pain during and after pregnancy is a major public health problem with significant daily problems for affected women and their families. There is now accumulating evidence that pregnancy-related pelvic girdle pain originates from the sacroiliac joints and the pubic symphysis as well as their extra-articular ligaments. However, the heritability of the disease remains to be determined. We hypothesized that there is an increased familial risk of pregnancy-related pelvic girdle pain. A population-based national database linkage registry study of approximately 9.3 million individuals within 4.2 million families in Sweden with a recruitment period from 1997 to 2018. The Swedish Multi-generation register was used to find female pairs of twins, full siblings, half-siblings and first cousins where both in the pairs had a completed pregnancy. The outcome measure was diagnosis of pregnancy-related pelvic girdle pain (International Classification of Diseases-10 O26.7 [1997-2018]) in the first pregnancy. Data was obtained from the Swedish Hospital Discharge Register, the Swedish Outpatient Care Register, the Swedish Medical Birth Register, the Primary Healthcare Register, and Medical Treatment Register. Cox regression analysis was used to calculate adjusted estimated effect of the exposure variable familial history of pregnancy-related pelvic girdle pain on the outcome variable pregnancy-related pelvic girdle pain at first birth. From the registers, 1 010 064 women pregnant with their first child within 795 654 families were collected. In total, 109 147 women were diagnosed with pregnancy-related pelvic girdle pain. The adjusted hazard ratio for a familial risk of pregnancy-related pelvic girdle pain was 2.09 (95% CI 1.85-2.37) among twins (monozygotic and dizygotic), 1.78 (95% CI 1.74-1.82) in full siblings, 1.16 (95% CI 1.06-1.28) in half-siblings from the mother, 1.09 (95% CI 1.024-1.16) in half-siblings from the father and 1.09 (95% CI 1.07-1.12) in first cousins. This nationwide observational study showed a familial clustering of pregnancy-related pelvic girdle pain. The hazard ratio for the condition was associated with the degree of relatedness, suggesting that heredity factors contribute to the development of pregnancy-related pelvic girdle pain. There is no causal treatment available for pregnancy-related pelvic girdle pain and further studies are now encouraged to clarify the specific genetic factors that contribute to the disease and for future targeted interventions.

Examining the treatment journey of patients diagnosed with hepatocellular carcinoma (HCC) using retrospective real-world data in British Columbia, Canada.

e16111 Background: Hepatocellular carcinoma (HCC) is a complex disease with treatments that may include locoregional therapy (LRT), systemic therapy and liver transplant. Using retrospective real-world data from a tertiary care hospital in British Columbia (BC). We examined characteristics, treatment pathways, survival and relevant healthcare costs of HCC patients. Methods: This was a retrospective, real-world data study using data from two existing chart review studies (called MOTION and HCC) and the BC Transplant database. The Canadian province of British Columbia has a population of 5.0 million who receive medical coverage under the Medical Services Plan. “MOTION” collected data on patients who received LRTs between 2004 and 2017, while “HCC” collected data on patients who received systemic therapies between 2000 and 2019. MOTION and HCC and Transplant data were linked and patients were assigned a unique study number and personal health information (PHI) was removed. Start and stop dates for LRT and systemic therapies were used to determine treatment patterns. Treatment costs were provided by BC Cancer’s Pharmacy department, medical oncology visits were assumed to be monthly while patients were on systemic treatment, LRT costs were based on departmental expenditures, and the cost of a liver transplant was provided by the regional health authority. Analyses used descriptive statistics for baseline characteristics, duration and costs (in 2021 Canadian dollars) and Kaplan-Meier curves for survival. Results: There were 417 and 413 patients from the HCC and MOTION databases respectively, in which 63 patients were in both (referred to as the “common” group). Median age (years) at diagnosis was 64, 62 and 62 for HCC, MOTION and common groups respectively, while the proportion who were female was 17.5%, 18.2% and 15.9%. Median overall survival (mOS) from diagnosis was 21, 33 and 28 months for the HCC, MOTION and common groups, respectively. From start of systemic treatment, mOS was 9 months for the HCC group, 29 months from first LRT for patients in the MOTION group, and 26 months for the common group from first treatment. Treatment and cost results were based only on the common group: mean length of follow-up from diagnosis was 34.6 ± 26.2 months and mean time from diagnosis to first treatment was 4.7 ± 4.2 months. DEB-TACE was most frequently the first LRT, often received 3 rounds of LRT before starting systemic treatment, in which sorafenib was given mostly as first line systemic therapy. When including only costs related to LRTs, systemic therapies, medical oncology visits and liver transplants, the mean cost per patient was $94,419, resulting in $51,649 mean cost per patient per year. Conclusions: The linkage of HCC-related LRT, systemic and transplant databases provided detailed results related to the treatment patterns, survival and costs of HCC patients in Canada.

Lethal use of force surveillance: practical considerations for open-source database linkage

Currently, there remains no reliable and timely government tracking in the USA of deaths caused by law enforcement. Federal efforts to track these events are insufficient, generally missing as many...

Feasibility and validity of using healthcare databases to conduct cross-national comparative studies of opioid use, its determinants and consequences.

A cross-national comparative (CNC) study about opioid utilization would allow the identification of strategies to improve pain management and mitigate risk. However, little is known about the accessibility and validity of information in healthcare databases internationally. This study aimed to identify the feasibility of using healthcare databases to conduct a CNC study of opioid utilization and its associated consequences. A cross-sectional survey was launched in March 2018, including experts interested in CNC studies comparing opioid utilization by purposeful sampling. An electronic survey was used to collect database characteristics, medicine information, and linkage information of each aggregate-level dataset (AD) and individual patient-level dataset (IPD). Overall, participants from 21 geographical regions reported 18 ADs and 19 IPDs. Information on dispensed medications is available from 17 ADs and 17 IPDs. Of the 16 ADs that include primary care settings, only 9 ADs can obtain information from secondary care settings. Fourteen IPDs included patients' characteristics or could be retrieved from linkage databases. Although most ADs are publicly accessible (n = 13), only five IPDs can be accessed without extra cost. Most ADs could be used to report opioid utilization in a primary care setting. IPDs with linkage databases should be applied to identify potential determinants, clinical outcomes, and policy impact. Data access restrictions and governance policies across jurisdictions can be challenging for timely analysis and require further collaboration.

The matrix revolutions: How databases and database linkages will transform epidemiologic research.

The matrix revolutions: How databases and database linkages will transform epidemiologic research.

National data linkage assessment of live births and deaths in Mexico: Estimating under‐five mortality rate ratios for vulnerable newborns and trends from 2008 to 2019

Linked datasets that enable longitudinal assessments are scarce in low and middle-income countries. We aimed to assess the linkage of administrative databases of live births and under-five child deaths to explore mortality and trends for preterm, small (SGA) and large for gestational age (LGA) in Mexico. We linked individual-level datasets collected by National statistics from 2008 to 2019. Linkage was performed based on agreement on birthday, sex, residential address. We used the Centre for Data and Knowledge Integration for Health software to identify the best candidate pairs based on similarity. Accuracy was assessed by calculating the area under the receiver operating characteristic curve. We evaluated completeness by comparing the number of linked records with reported deaths. We described the percentage of linked records by baseline characteristics to identify potential bias. Using the linked dataset, we calculated mortality rate ratios (RR) in neonatal, infants, and children under-five according to gestational age, birthweight, and size. For the period 2008-2019, a total of 24,955,172 live births and 321,165 under-five deaths were available for linkage. We excluded 1,539,046 records (6.2%) with missing or implausible values. We succesfully linked 231,765 deaths (72.2%: range 57.1% in 2009 and 84.3% in 2011). The rate of neonatal mortality was higher for preterm compared with term (RR 3.83, 95% confidence interval, [CI] 3.78, 3.88) and for SGA compared with appropriate for gestational age (AGA) (RR 1.22 95% CI, 1.19, 1.24). Births at <28 weeks had the highest mortality (RR 35.92, 95% CI, 34.97, 36.88). LGA had no additional risk vs AGA among children under five (RR 0.92, 95% CI, 0.90, 0.93). We demonstrated the utility of linked data to understand neonatal vulnerability and child mortality. We created a linked dataset that would be a valuable resource for future population-based research.

The Birth of Digital Epidemiology in South Korea

The COVID-19 Epidemiological Investigation Support System (EISS) is a digital epidemiological tool, which utilizes location data from cellular base stations, credit card transactions records, and QR codes. It is a mass surveillance system that uses big data to track the entire infected population, featuring an extensive, automated, and speedy processing of data on personal location and the linkage of multiple databases from various governmental agencies. Based on interviews with people who have developed Korean digital epidemiology systems, this paper explores the technical, infrastructural, social, and institutional factors that have shaped Korean digital epidemiology since the 2014 avian flu crisis and examines the essential conditions of big data for digital epidemiology. The main findings are as follows: The feasibility of EISS goes beyond the matter of privacy; it is closely connected to technological infrastructures such as a high density of cellular base stations and private cloud systems; people’s behavior such as a high rate of smartphone and credit card usage; and new forms of governance and institutions for speedy data processing. Multiple database linkage would develop EISS into a big data surveillance system that enables the prediction of risk-prone groups in a more preemptive manner.

Abstract P1-07-04: Claims-derived comorbidities associations with real-world overall survival (rwOS) among patients with metastatic breast cancer in an electronic health record-claims linked dataset

Abstract Introduction: Comorbid conditions can significantly impact various aspects of cancer care and therefore are often utilized as covariates to control for selection bias in observational studies. Oncology based electronic health record (EHR) data is often used as a source for comorbidity information but may be subject to missingness due to differences in clinician documentation. Utilizing additional data sources, such as health care claims data, may allow for greater capture of comorbidities. This study used a claims data linked to oncology EHR to identify comorbidities and related patient characteristics, as well as the association between a claims-derived comorbidity index and real-world overall survival (rwOS) among outpatient metastatic breast cancer (mBC) patients, to characterize the comorbidities data from this data source. Methods: We selected patients diagnosed with mBC between 2013 and 2021 from a linkage of mBC de-identified EHR-derived Flatiron Health Research Database (FHRD) and Komodo Health claims data. Charlson Comorbidity Index (CCI) were identified based on the presence of a pre-specified International Classification of Diseases (ICD) codes within 12 months preceding cancer diagnosis. To understand comorbidity data within this linkage, patient characteristics were assessed after the study cohort was stratified by the presence of any comorbidity assessed as part of CCI. We calculated the CCI for the study cohort as well as the prevalence of comorbid conditions that are used as part of the CCI. Kaplan-Meier estimates and Cox proportional hazards model were used to estimate median survival time and hazard ratios (HRs) of rwOS among patients with a CCI score of 1 and 2+. Results: The study cohort had 3,213 patients with a median age of 63 [IQR(52, 76)] and 98.7% were female. White patients accounted for 60.1% of the study population while Black and Hispanic/Latino patients accounted for 12.5% and 7.4% of the study cohort, respectively. Patients with comorbidities were older (mean age of 65 vs 61), had a higher (2+) ECOG performance status (PS) score (14.0% vs 9.7%), and were more likely to be Black (16.0% vs 11.0%) or Hispanic/Latino 10.8% vs 6.3%). Twenty percent of the patients had diabetes, 8.2% had peripheral vascular disease, 7.5% had mild liver disease, and 6.3% had congestive heart failure. According to index values, 21.0% of the study cohort had a CCI score of 1, and 18.0% had a score of 2 and above. The median survival times in years for patients with a CCI score of 1 and 2+ were 2.51 [95% CI 2.29,2.81] and 2.05 [95% CI 1.83,2.42], respectively. Based on univariate analysis, patients with a CCI score greater than zero had a higher risk of mortality (CCI 1: HR = 1.25 [95%CI 1.10, 1.41, p&amp;lt; 0.01], CCI 2+: HR = 1.54 [95%CI 1.35, 1.76, p&amp;lt; 0.01]). In the multivariate analysis, adjusted for age, ethnicity/race, and baseline ECOG PS score, the HR for the CCI 1 group was not statistically significant (HR=1.12 [95%CI 0.99, 1.27, p=0.08]). However, the risk of mortality was statistically significant for the CCI 2+ group (HR=1.31 [95%CI 1.14, 1.50], p &amp;lt; 0.01) in the adjusted analysis. Conclusion: Claims data linked to EHR can be used to identify comorbidities and describe patient characteristics. As anticipated, generally, the presence of more comorbid conditions was associated with worse rwOS for patients with mBC. This finding supports use of this linked dataset for similar future studies. Citation Format: Mustafa S. Ascha, Alemseged A. Asfaw, Prakirthi Yerram, Samantha N. Reiss, Sarah D. Brake, Niquelle B. Wadé. Claims-derived comorbidities associations with real-world overall survival (rwOS) among patients with metastatic breast cancer in an electronic health record-claims linked dataset [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P1-07-04.

Childhood Abuse, Intimate Partner Violence in Young Adulthood, and Welfare Receipt by Midlife.

To investigate prospective associations between type of child abuse (physical, sexual, both), timing (childhood, young adulthood, both), and welfare receipt into middle-age. Database linkage study using the Quebec Longitudinal Study of Kindergarten Children cohort born in 1980 and government administrative databases (N = 3020). We assessed parental tax returns, family and personal background characteristics (1982-1987). At age 22 years, participants answered retrospective questionnaires on experienced childhood abuse (physical, sexual abuse < age 18 years) and intimate partner violence (IPV) (ages 18-22). Main outcome was years on social assistance, on the basis of participant tax returns (ages 23-37 years). Analysis included weights for population representativeness. Of 1690 participants (54.4% females) with available data, 22.4% reported childhood abuse only, 14.5% IPV only, and 18.5% both. Prevalence of childhood physical, sexual, and both was 20.4%, 12.2%, and 8.3%, respectively. Adjusting for socioeconomic background and individual characteristics, we found that childhood physical abuse alone and physical or sexual abuse combined were associated with a two-fold risk of welfare receipt, as compared to never-abused (adjusted incidence risk ratio 2.43, 95% confidence interval [CI], 1.65-3.58; and adjusted incidence risk ratio 2.04, 95% CI, 1.29-3.23, respectively). Repeated abuse (childhood abuse combined with adult IPV) had a three-fold risk (adjusted incidence ratio 3.59, 95% CI, 2.39-5.37). Abuse across several developmental periods (childhood and young adulthood) is associated with increased risks of long-term welfare receipt, independently of socioeconomic background. Results indicate a dose-response association. Early prevention and targeted identification are crucial to preventing economic adversity that may potentially lead to intergenerational poverty.

P127 Inflammatory Bowel Disease-associated colorectal cancers: retrospective cohort study from a tertiary centre surveillance programme over 10 years

Abstract Background Patients with IBD have increased risk for developing colorectal cancer (CRC), with poorer survival, compared to those without IBD. Despite surveillance colonoscopy, IBD post-colonoscopy CRC (PCCRC) rates are unacceptably high at 28-45%. We define surveillance colonoscopy quality, IBD-dysplasia and IBD-CRC prevalence, IBD-CRC characteristics, and IBD PCCRC rates at a tertiary IBD centre. Methods The GI Reporting Tool (Unisoft Medical Systems) was searched for IBD surveillance colonoscopy reports from April 2008–December 2018. Electronic medical records were reviewed. IBD phenotype and diagnosis date was confirmed on the Lothian IBD Registry. Follow-up was until September 2022. Results 1892 colonoscopies from 1262 patients were included. Median follow-up was 7.1 (4.4, 9.5) years. 885/1262 (70.1%) had UC, 344/1262 (27.3%) had Crohn’s, and 33/1262 (2.6%) had IBD-U. Median disease duration at colonoscopy was 16.8 (11, 24.6) years. 1825/1892 (96.5%) colonoscopies achieved caecal intubation, 1735/1840 (94.3%) adhered to biopsy guidelines and 704/1892 (37%) used chromoendoscopy. Histological inflammation was identified in 873/1823 (47.9%) cases; 57.2% in the right colon. Bowel preparation was poor in 194/1892 (10.3%) procedures and the most documented reason for not performing chromoendoscopy. Poor bowel preparation was associated with 23/60 (38.3%) poorly tolerated and 29/52 (55.8%) incomplete colonoscopies. Moviprep was superior to Picolax (p&amp;lt;0.001). GI physicians were more likely to use chromoendoscopy (p&amp;lt;0.001), biopsy per guideline (p&amp;lt;0.001), and identify dysplasia (p=0.012) compared to surgical colleagues. Dysplasia was identified in 145/1892 (7.7%) colonoscopies: 84/1892 (4.4%) sporadic, 22/1892 (1.2%) IBD-associated, and 39/1892 (2.1%) unspecified. IBD-CRC is rare, at 0.8 cases per 1000 patient years. Within the study period, 20/1262 (1.6%) patients had CRC diagnosed. 17/20 (85%) had UC and 3/20 (15%) had Crohn’s. 15/17 (88%) UC patients had E3 disease. 11/20 (55%) presented with lymph node or distant metastasis. 8/20 (40%) tumours had mucinous differentiation; only 1/15 (6.7%) was well differentiated. IBD-CRC patients had longer disease duration (28.3 (19.7, 37.8) years) compared to patients without CRC (22.25 (16.42, 29.85) years) at follow-up (p=0.028). The 3-year PCCRC rate was 15%. Conclusion IBD-CRC is rare however IBD PCCRC rates are unacceptably high. While this may reflect disease biology, a new approach to IBD surveillance is urgently needed. Interventions could include dedicated IBD surveillance lists, patient education around bowel preparation, and pre-screening for inflammation (e.g. faecal calprotectin). Cancer registry database linkage is now essential for us to identify IBD-CRC cases not identified by surveillance.

Commentary on Lim et al.: Mortality database linkages-A critical methodology to understand the structural drivers of the overdose crisis.

Commentary on Lim et al.: Mortality database linkages-A critical methodology to understand the structural drivers of the overdose crisis.

Linkage of administrative and compensation databases for work-related asthma surveillance in Ontario: A proof of concept study

RATIONALE: Approximately 15% of all asthma cases are work-related and eligible for workers’ compensation in Ontario. However, compensation rates of work-related asthma (WRA) are far less than predicted, making it difficult to estimate the prevalence of the disease. OBJECTIVES We aimed to estimate prevalence of compensated WRA in Ontario; profile the pattern of compensated WRA by demographic, temporal and geographic factors; and demonstrate the potential for database linkage to monitor rates of compensated WRA cases. METHODS Compensated WRA claims data were linked to asthma cases in the Institute of Clinical Evaluative Sciences (ICES) asthma database via encrypted health card numbers. WRA claims between April 1998 and March 2002 were accessed from: i) the Ontario Workplace Safety and Insurance Board (WSIB) Occupational Disease Information Surveillance System (ODISS); and ii) a University of Toronto research database (RD) created by abstracting the same WSIB ODISS claim files. MAIN RESULTS: The estimated prevalence of WRA among individuals with asthma in the asthma database was less than 1% compared to an expected prevalence of 15-20%. Sensitivity of the Asthma database for including individuals with asthma with WRA was very good but differed significantly based on claims category (p < 0.001) compared to the RD as the gold standard. CONCLUSIONS Our findings suggest WRA is severely under-reported. Approximately 11-15% of compensated WRA claims are not captured by the asthma database. Factors accounting for discordance between databases should be explored in order for administrative data linkage to be used to monitor the rates of compensated WRA cases in Ontario.

Registration with Primary Health Care and COVID-19 mortality: cohort of diabetics from five administrative health regions in the city of Rio de Janeiro, Brazil, 2020-2021.

The present study carried out an analysis of survival according to the status of registration with Primary Health Care (PHC) and of factors associated with death from COVID-19, in cases residing in Programmatic Area 3.1 (PA3.1) with a diagnosis of diabetes (in the notification form or in the electronic medical record), of the Municipality of Rio de Janeiro (RJ), Brazil, in 2020-2021. A probabilistic linkage of databases was performed based on information on cases notified as COVID-19 and data from the electronic medical records of people living with diabetes. A survival analysis was carried out, using the Cox regression model stratified by age group and adjusted for confounding variables. Individuals registered with the PHC of PA3.1 had almost twice the risk of death from COVID-19 (adjusted hazard ratio [HRadj]=1.91) when compared to those unregistered. This association was stronger in individuals aged 18 to 59 years registered with the PHC (HRadj=2.82) than in individuals aged 60 years or over (HRadj=1.56). Surveillance strategies for identifying and adequately monitoring higher-risk groups, among individuals living with diabetes, within the scope of Primary Health Care, can contribute to reducing mortality from COVID-19.

Federal Funding and Citation Metrics of US Biomedical Researchers, 1996 to 2022

Both citation and funding metrics converge in shaping current perceptions of academic success. To evaluate what proportion of the most-cited US-based scientists are funded by biomedical federal agencies and whether funded scientists are more cited than nonfunded ones. This survey study used linkage of a Scopus-based database on top-cited US researchers (according to a composite citation metric) and the National Institutes of Health RePORTER database of federal funding (33 biomedical federal agencies). Matching was based on name and institution. US-based top-cited scientists who were allocated to any of 69 scientific subfields highly related to biomedicine were considered in the main analysis. Data were downloaded on June 11, 2022. Proportion of US-based top-cited biomedical scientists who had any (1996-2022), recent (2015-2022), and current (2021-2022) funding. Comparisons of funded and nonfunded scientists assessed total citations and a composite citation index. There were 204 603 records in RePORTER (1996-2022) and 75 316 US-based top-cited scientists in the career-long citation database; 40 887 scientists were included in the main analysis. The proportion of US-based top-cited biomedical scientists (according to career-long citation impact) who had received any federal funding from biomedical research agencies was 62.7% (25 650 of 40 887) for any funding (1996-2022), 23.1% (9427 of 40 887) for recent funding (2015-2022), and 14.1% (5778 of 40 887) for current funding (2021-2022). Respective proportions were 64.8%, 31.4%, and 20.9%, for top-cited scientists according to recent single-year citation impact. There was large variability across scientific subfields (eg, current funding: 31% of career-long impact top-cited scientists in geriatrics, 30% in bioinformatics and 29% in developmental biology, but 0% in legal and forensic medicine, general psychology and cognitive sciences, and gender studies). Funded top-cited researchers were overall more cited than nonfunded top-cited scientists (median [IQR], 9594 [5650-1703] vs 5352 [3057-9890] citations; P < .001) and substantial difference remained after adjusting for subfield and years since first publication. Differences were more prominent in some specific biomedical subfields. In this survey study, biomedical federal funding had offered support to approximately two-thirds of the top-cited biomedical scientists at some point during the last quarter century, but only a small minority of top-cited scientists had current federal biomedical funding. The large unevenness across subfields needs to be addressed with ways that improve equity, efficiency, excellence, and translational potential.

Dying to Get There: Patients Who Reside at Increased Distance from Tertiary Center Experience Increased Mortality Following Abdominal Aortic Aneurysm Rupture

Centralization of vascular surgery care for Ruptured Abdominal Aortic Aneurysms (RAAAs) to high-volume tertiary centers may hinder access to timely surgical intervention for patients in remote areas. The objective of this study was to determine the association between distance from vascular care and mortality from RAAAs in the province of Nova Scotia, Canada. A retrospective cohort study of all RAAAs in Nova Scotia between 2005 and 2015 was performed through linkage of administrative databases. Patients were divided into groups by estimated travel time from their place of residence to the tertiary center (<1hr and ≥1hr) using geographic information software. Baseline and operative characteristics were identified for all patients through available databases and completed through chart review. Mortality at home, during transfer to the vascular center, and overall 30-day mortality were compared between groups using t-test and chi-squared test, as appropriate. Multivariable logistic regression analysis was used to calculate the independent effect of travel time on survival outcomes. A total of 567 patients with RAAA were identified from 2005-2015, of which 250 (44%) resided <1hr travel time to the tertiary center and 317 (56%) resided ≥1hr. On multivariable analysis, travel time ≥1hr from vascular care was an independent predictor of mortality at home (odds ratio [OR] 1.68, 95% confidence interval [CI] 1.07-2.63, P=0.02), mortality prior to operation (OR 2.64, 95% CI 1.81-3.83, P<0.001), and overall 30-day mortality (OR 1.61, 95% CI 1.10-2.37, P=0.02). In patients who received an operation (n=294), there was no association between increased travel time and mortality (OR 1.02, 95% CI 0.60-1.73, P=0.94). Travel time ≥1hr to the tertiary center is associated with significantly higher mortality from ruptured abdominal aortic aneurysm (AAA). However, there was no difference in overall chance of survival between groups for patients that underwent AAA repair. Therefore, strategies to facilitate early detection, and timely transfer to a vascular surgery center may improve outcomes for patients with RAAA.

RWD126 Introducing the German MSDS-AOK PLUS Database: Linkage of Administrative Claims Data and a Patient Registry for Advanced RWE Capabilities in Multiple Sclerosis

With varying availability of data among real-world data (RWD) sources, the linkage of databases can expand real-world evidence opportunities in multiple sclerosis (MS). We introduce a novel linked database via a use case combining the strengths of claims and MS registries.

Effect of after-hours presentation in ruptured abdominal aortic aneurysm

ObjectiveRuptured abdominal aortic aneurysms (RAAAs) are surgical emergencies that require immediate and expert treatment. It has been unclear whether presentation during evenings and weekends, when “on call” teams are primarily responsible for patient care, is associated with worse outcomes. Our objective was to evaluate the outcomes of patients presenting with RAAAs after-hours vs during the workday. MethodsA retrospective cohort study of all RAAAs in Nova Scotia between 2005 and 2015 was performed through linkage of administrative databases. Patients who had presented to the hospital with RAAAs during the workday (Monday through Friday, 6 am to 6 pm) were compared with those who had presented after-hours (6 pm to 6 am during the week and on weekends). The baseline and operative characteristics were identified for all patients through the available databases and a review of the medical records. Mortality before surgery, 30-day mortality, and operative mortality were compared between groups using multivariable logistic regression, adjusting for factors clinically significant on univariable analysis. ResultsA total of 390 patients with RAAAs were identified from 2005 to 2015, of whom 205 (53%) had presented during the workday and 185 (47%) after-hours. The overall chance of survival (OCS) was 45% overall, 49% if admitted to hospital, and 64% if surgery had been performed. During the workday, the OCS was 43% overall, 48% if admitted to hospital, and 67% if surgery had been performed. After-hours, the OCS was 46% overall, 49% if admitted to hospital, and 61% if surgery had been performed. Mortality before surgery was increased for patients who had presented to the hospital during the workday compared with after-hours (36% vs 26%; P = .04). The 30-day mortality (57% vs 54%; P = .62), rates of operative management (63% vs 72%; P = .06), and operative mortality (33% vs 39%; P = .33) were similar between the workday and after-hours groups (57% vs 54%; P = .06). After adjusting for significant clinical variables, the patients who had presented with RAAAs after-hours had had a similar odds of dying before surgery (odds ratio [OR], 0.64; 95% confidence interval [CI], 0.41-1.03), operative management (OR, 1.47; 95% CI, 0.93-2.31), 30-day mortality (OR, 0.98; 95% CI, 0.63-1.51), and operative mortality (OR, 1.33; 95% CI, 0.78-2.26). In the subgroup of patients presenting to a hospital with endovascular capabilities, patients presenting after-hours had had similar odds of 30-day mortality (OR, 1.07; 95% CI, 0.57-2.02), and operative mortality (OR, 1.14; 95% CI, 0.58-2.23). ConclusionsWe found that patients presenting to the hospital with RAAAs after-hours did not have increased adjusted odds of mortality before surgery, operative management, 30-day mortality, or operative mortality.

9New metrics for assessing linkage quality in deterministic record linkage of health databases

Linkage among healthcare databases, claims, and external data collections such as registries, cohorts or clinical trials is becoming increasingly important for conducting epidemiological research. For example, the French National Health Insurance Information System (SNDS) contains claims-based data on dispensed drugs, on out-of-hospital laboratory tests and on outpatient medical consultations, but without tests results, diagnosis or even the symptoms motivating the patient's consultation. On the other hand, registries, cohorts or clinical trials collect medically validated data, which is time-consuming and costly. The linkage of the two sources could therefore overcome lack of information and limit confounding biases. However, processing record linkage on huge health databases and assessing linkage quality can be challenging. This paper introduces a deterministic record linkage strategy that focuses on assessing linkage quality using new quality metrics. We developed a deterministic linkage strategy that systematically considers all combinations of individual identifiers. An exhaustive exploration of all variable combinations makes it possible to compute a new metric, referred to as robustness, and to generate a linkage cartography that precisely summarizes the linked pair characteristics. This cartography is central to our approach and makes it possible for the expert to easily accept/reject groups of linked pairs. The approach was tested on synthetic datasets staging a variety of possible linkage scenarios (datasets size/ratio, overlap, and errors), and on two real-world studies (a registry database and a clinical trial). Dataset simulations demonstrated very good accuracy with a limited impact of different factors tested, scalability, and encouraging runtimes. Minima were greater than 0.95 for recall and greater than 0.99 for precision, whatever the scenario. Feasibility on real datasets was verified with good results: among 3985 patients from the registry, the algorithm found 3850 single linked pairs and 135 proposals with multiple candidates out of 504,795 candidates. After reviewing the linkage cartography, the expert validated 3783 linked pairs and a manual review of multiple candidates added 20 pairs, reaching a linkage rate of 95.4%. For the trial, only 2 records out of 129 were not linked among 22,426 candidates, as a result of early withdrawal (no information in the trial database) giving a linkage rate of 98.4%. In both cases, unlinked records did not seem to show any potential bias. Performance is good since linking a synthetic database holding 30,000 patients versus a synthetic database holding 3,000,000 patients takes a few seconds only. The approach is resilient by design to missing information and therefore well suited for linking the SNDS database to cohorts/registries when information overlap between the two databases cannot be perfect. Finally, our implementation is fast enough to interactively improve the linkage results through successive improved runs. The novelty of our approach is twofold: first, the linkage cartography provides a new way of classifying and comparing deterministic rules from the set of all possible rules and second, the approach is by design resilient to data corruption and can reach better recall than standard deterministic linkage strategies. Finally, good performance and scalability open the door to the linkage of very large datasets. Les auteurs n'ont pas précisé leurs éventuels liens d'intérêts.

Post-Endoscopy Upper Gastrointestinal Cancer Rate in a Tertiary Referral Centre: An Australian Data Linkage Analysis.

High-quality gastroscopy is critical for early diagnosis of upper gastrointestinal cancers (UGCs), and assessment of missed cancers may serve as a key quality metric. Using a prospective gastroscopy database and data linkage with the Queensland Cancer Registry, we assessed the risk of developing UGC within 3years of a cancer-negative gastroscopy at an Australian tertiary centre. Additional aims were to identify factors predictive of missed cancer, perform root cause analyses for missed cancers and assess overall survival. We identified patients who were diagnosed with UGC within 3years of undergoing gastroscopy between 2011 and 2016. Non-mucosal cancers, cancers distal to duodenum and patients undergoing surveillance were excluded. Cases diagnosed within 6months of gastroscopy were defined as detected cancers, while those developing within 6-36months were defined as missed cancers. Post-endoscopy UGC rate (PEUGIC-3Y) was calculated as ratio of missed over total cancers detected. Demographic, clinical, endoscopic and histologic variables were analysed. A total of 17,131 gastroscopies were performed for 10,393 patients during the study period. One hundred and twenty-six UGCs were diagnosed, including 120 detected UGCs and 6 missed UGCs. The overall PEUGIC-3Y rate was 4.8% (95% CI 2.1-10.4). The missed UGCs included 3 gastric adenocarcinomas, 2 gastro-oesophageal junction adenocarcinomas and 1 oesophageal squamous cell carcinoma. At the preceding 'cancer-negative gastroscopy', no macroscopic abnormalities were detected at the site of future UGC in 5/6 patients. A UGC developed in 2/6 patients despite an apparent adequate examination at index gastroscopy. Age, sex, indication for endoscopy and cancer location or histology were not predictive of missed cases, and survival was comparable between groups. We demonstrate that the PEUGIC-3Y rate was 4.8% (95% CI 2.1-10.4). The majority of missed cases were adenocarcinomas of the gastro-oesophageal junction or stomach and developed in segments which were found to be normal at index gastroscopy, highlighting the challenges in detecting subtle mucosal lesions in the upper gastrointestinal tract. While overall survival between patients with detected and post-gastroscopy cancers was comparable, these ultimately represent potential missed opportunities for diagnosing an early cancer and underscore the need for quality improvement in gastroscopy.

Impact of body mass index on survival in women receiving chemotherapy for early breast cancer

The impact of elevated body mass index (BMI) on overall survival (OS) in patients receiving modern anthracycline-taxane chemotherapy for early breast cancer (EBC) has not yet been well established. The purpose of our study was to examine overall survival (OS) by BMI category in women with EBC receiving either doxorubicin (A), cyclophosphamide (C) + paclitaxel (P) or fluorouracil (F), epirubicin (E), cyclophosphamide (C) + docetaxel (D). This was a retrospective cohort study in patients ≥ 18years with resected stage I-III BC diagnosed between 2007 and 2017 in Ontario, identified through linkage of administrative databases. Patients were classified according to baseline BMI into underweight (< 18.5kg/m2), normal (18.5-24.9kg/m2), overweight (25-29.9kg/m2), and obese (≥ 30kg/m2) World Health Organization (WHO) categories. The primary outcome was OS. Univariable and multivariable analyses were used to examine the association between clinico-pathologic characteristics and OS among BMI categories. Our cohort included 11,601 women, of whom 3890 (33.5%) were normal weight, 3696 (31.9%) overweight, and 3847 (33.1%) obese. Median OS was 7.9years. There were no statistically significant differences in OS according to BMI (p = 0.66) in the overall study cohort or among the BMI categories after adjusting for age, nodal status, stage, grade, ER and HER2 status for either AC-P or FEC-D- treated patients (p = 0.45 and p = 0.97, respectively). Our large population-based retrospective cohort analysis of EBC patients receiving adjuvant anthracycline-taxane chemotherapy found no significant impact of BMI on OS. Further investigation is warranted to confirm these findings in prospective patient cohorts.